The assertion here was to make sure the newly constructed and
typechecked expression selected the same receiver-qualified method,
but in the case of anonymous receiver types we can actually end up
with separate types.Field instances corresponding to each types.Type
instance. In that case, the assertion spuriously failed.
The fix here is to relax and assertion and just compare the method's
name and type (including receiver type).
Fixes#58776.
Change-Id: I67d51ddb020e6ed52671473c93fc08f283a40886
Reviewed-on: https://go-review.googlesource.com/c/go/+/471676
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
(cherry picked from commit 37a2004b43)
Reviewed-on: https://go-review.googlesource.com/c/go/+/472620
Auto-Submit: Dmitri Shuralyov <dmitshur@google.com>
TryBot-Bypass: Dmitri Shuralyov <dmitshur@google.com>
This test is flaky, apparently due to a typo'd operator in CL 21447
that causes it to compare “same port OR IP” instead of
“same port AND IP”.
If we merely fixed the comparison, the test would hopefully stop being
flaky itself, but we would still be left with another problem:
repeatedly dialing a port that we believe to be unused can interfere
with other tests, which may open the previously-unused port and then
attempt a single Dial and expect it to succeed. Arbitrary other Dial
calls for that port may cause the wrong connection to be accepted,
leading to spurious test failures.
Moreover, the test can be extremely expensive for the amount of data
we hope to get from it, depending on the system's port-reuse
algorithms and dial implementations. It is already scaled back by up
to 1000x on a huge number of platforms due to latency, and may even be
ineffective on those platforms because of the arbitrary 1ms Dial
timeout. And the incremental value from it is quite low, too: it tests
the workaround for what is arguably a bug in the Linux kernel, which
ought to be fixed (and tested) upstream instead of worked around in
every open-source project that dials local ports.
Instead of trying to deflake this test, let's just get rid of it.
Updates #18290.
Fixes#58717.
Change-Id: I8a58b93d67916a33741c9ab29ef99c49c46b32c4
Reviewed-on: https://go-review.googlesource.com/c/go/+/460657
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
(cherry picked from commit e08642cae1)
Reviewed-on: https://go-review.googlesource.com/c/go/+/471155
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
Auto-Submit: Dmitri Shuralyov <dmitshur@google.com>
CL 416115 added using faccessat2(2) from syscall.Faccessat on Linux
(which is the only true way to implement AT_EACCESS flag handing),
if available. If not available, it uses some heuristics to mimic the
kernel behavior, mostly taken from glibc (see CL 126415).
Next, CL 414824 added using the above call (via unix.Eaccess) to
exec.LookPath in order to check if the binary can really be executed.
As a result, in a very specific scenario, described below,
syscall.Faccessat (and thus exec.LookPath) mistakenly tells that the
binary can not be executed, while in reality it can be. This makes
this bug a regression in Go 1.20.
This scenario involves all these conditions:
- no faccessat2 support available (i.e. either Linux kernel < 5.8,
or a seccomp set up to disable faccessat2);
- the current user is not root (i.e. geteuid() != 0);
- CAP_DAC_OVERRIDE capability is set for the current process;
- the file to be executed does not have executable permission
bit set for either the current EUID or EGID;
- the file to be executed have at least one executable bit set.
Unfortunately, this set of conditions was observed in the wild -- a
container run as a non-root user with the binary file owned by root with
executable permission set for a user only [1]. Essentially it means it
is not as rare as it may seem.
Now, CAP_DAC_OVERRIDE essentially makes the kernel bypass most of the
checks, so execve(2) and friends work the same was as for root user,
i.e. if at least one executable bit it set, the permission to execute
is granted (see generic_permission() function in the Linux kernel).
Modify the code to check for CAP_DAC_OVERRIDE and mimic the kernel
behavior for permission checks.
[1] https://github.com/opencontainers/runc/issues/3715
For #58552.
Fixes#58624.
Change-Id: I82a7e757ab3fd3d0193690a65c3b48fee46ff067
Reviewed-on: https://go-review.googlesource.com/c/go/+/468735
Reviewed-by: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
(cherry picked from commit 031401a790)
Reviewed-on: https://go-review.googlesource.com/c/go/+/469956
Auto-Submit: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
When inlining functions that contain function literals, we need to be
careful about position information. The OCLOSURE node should use the
inline-adjusted position, but the ODCLFUNC and its body should use the
original positions.
However, the same problem can arise with certain generic constructs,
which require the compiler to synthesize function literals to insert
dictionary arguments.
go.dev/cl/425395 fixed the issue with user-written function literals
in a somewhat kludgy way; this CL extends the same solution to
synthetic function literals.
This is all quite subtle and the solutions aren't terribly robust, so
longer term it's probably desirable to revisit how we track inlining
context for positions. But for now, this seems to be the least bad
solution, esp. for backporting to 1.20.
Updates #54625.
Fixes#58531.
Change-Id: Icc43a70dbb11a0e665cbc9e6a64ef274ad8253d1
Reviewed-on: https://go-review.googlesource.com/c/go/+/468415
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
(cherry picked from commit 873c14cec730ee278848f7cc58d2b4d89ab52288)
Reviewed-on: https://go-review.googlesource.com/c/go/+/471677
Auto-Submit: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
This patch provides a fix for a problem linking large arm32 binaries
with external linking, specifically R_CALLARM relocations against
runtime.duff* routines being flagged by the external linker as not
reaching.
What appears to be happening in the bug in question is that the Go
linker and the external linker are using slightly different recipes to
decide whether a given R_CALLARM relocation will "fit" (e.g. will not
require a trampoline). The Go linker is taking into account the addend
on the call reloc (which for calls to runtime.duffcopy or
runtime.duffzero is nonzero), whereas the external linker appears to
be ignoring the addend.
Example to illustrate:
Addr Size Func
----- ----- -----
...
XYZ 1024 runtime.duffcopy
...
ABC ... mypackge.MyFunc
+ R0: R_CALLARM o=8 a=848 tgt=runtime.duffcopy<0>
Let's say that the distance between ABC (start address of
runtime.duffcopy) and XYZ (start of MyFunc) is just over the
architected 24-bit maximum displacement for an R_CALLARM (let's say
that ABC-XYZ is just over the architected limit by some small value,
say 36). Because we're calling into runtime.duffcopy at offset 848,
however, the relocation does in fact fit, but if the external linker
isn't taking into account the addend (assuming that all calls target
the first instruction of the called routine), then we'll get a
"doesn't fit" error from the linker.
To work around this problem, revise the ARM trampoline generation code
in the Go linker that computes the trampoline threshold to ignore the
addend on R_CALLARM relocations, so as to harmonize the two linkers.
Fixes#58503.
Updates #58428.
Updates #58425.
Change-Id: I56e580c05b7b47bbe8edf5532a1770bbd700fbe5
Reviewed-on: https://go-review.googlesource.com/c/go/+/469275
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
(cherry picked from commit 0b5affb193)
Reviewed-on: https://go-review.googlesource.com/c/go/+/471597
Unlike the rest of nistec, the P-256 assembly doesn't use complete
addition formulas, meaning that p256PointAdd[Affine]Asm won't return the
correct value if the two inputs are equal.
This was (undocumentedly) ignored in the scalar multiplication loops
because as long as the input point is not the identity and the scalar is
lower than the order of the group, the addition inputs can't be the same.
As part of the math/big rewrite, we went however from always reducing
the scalar to only checking its length, under the incorrect assumption
that the scalar multiplication loop didn't require reduction.
Added a reduction, and while at it added it in P256OrdInverse, too, to
enforce a universal reduction invariant on p256OrdElement values.
Note that if the input point is the infinity, the code currently still
relies on undefined behavior, but that's easily tested to behave
acceptably, and will be addressed in a future CL.
Updates #58647Fixes#58720
Fixes CVE-2023-24532
(Filed with the "safe APIs like complete addition formulas are good" dept.)
Change-Id: I7b2c75238440e6852be2710fad66ff1fdc4e2b24
Reviewed-on: https://go-review.googlesource.com/c/go/+/471255
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
(cherry picked from commit 203e59ad41)
Reviewed-on: https://go-review.googlesource.com/c/go/+/471695
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Run-TryBot: Roland Shoemaker <roland@golang.org>
Fix a misuse of bufio.Reader.Read in the helper class
cmd/internal/cov.MReader; the MReader method in question should have
been using io.ReadFull (passing the bufio.Reader) instead of directly
calling Read.
Using the Read method instead of io.ReadFull will result in a "short"
read when processing a specific subset of counter data files, e.g.
those that are short enough to not trigger the mmap-based scheme we
use for larger files, but also with a large args section (something
large enough to exceed the default 4k buffer size used by
bufio.Reader).
Along the way, add some additional defered Close() calls for files
opened by the CovDataReader.visitPod, to enure we don't leave any open
file descriptor following a call to CovDataReader.Visit.
Fixes#58427.
Updates #58411.
Change-Id: Iea48dc25c0081be1ade29f3a633df02a681fd941
Reviewed-on: https://go-review.googlesource.com/c/go/+/466677
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
(cherry picked from commit a7fe9ada10)
Reviewed-on: https://go-review.googlesource.com/c/go/+/468536
Reviewed-by: Cherry Mui <cherryyz@google.com>
The sweep assist computation is intentionally racy for performance,
since the specifics of sweep assist aren't super sensitive to error.
However, if overflow occurs when computing the live heap delta, we can
end up with a massive sweep target that causes the sweep assist to sweep
until sweep termination, causing severe latency issues. In fact, because
heapLive doesn't always increase monotonically then anything that
flushes mcaches will cause _all_ allocating goroutines to inevitably get
stuck in sweeping.
Consider the following scenario:
1. SetGCPercent is called, updating sweepHeapLiveBasis to heapLive.
2. Very shortly after, ReadMemStats is called, flushing mcaches and
decreasing heapLive below the value sweepHeapLiveBasis was set to.
3. Every allocating goroutine goes to refill its mcache, calls into
deductSweepCredit for sweep assist, and gets stuck sweeping until
the sweep phase ends.
Fix this by just checking for overflow in the delta live heap calculation
and if it would overflow, pick a small delta live heap. This probably
means that no sweeping will happen at all, but that's OK. This is a
transient state and the runtime will recover as soon as heapLive
increases again.
Note that deductSweepCredit doesn't check overflow on other operations
but that's OK: those operations are signed and extremely unlikely to
overflow. The subtraction targeted by this CL is only a problem because
it's unsigned. An alternative fix would be to make the operation signed,
but being explicit about the overflow situation seems worthwhile.
For #57523.
Fixes#58536.
Change-Id: Ib18f71f53468e913548aac6e5358830c72ef0215
Reviewed-on: https://go-review.googlesource.com/c/go/+/468375
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Update golang.org/x/net to the tip of internal-branch.go1.20-vendor to
include CL 468336.
The contents of that CL were already merged into this branch in CL
468122, so this CL just brings go.mod back in line to matching the
actual vendored content.
For #58356
For #57855
Change-Id: I6ee9483077630c11c725927f38f6b69a784106db
Reviewed-on: https://go-review.googlesource.com/c/go/+/468302
Run-TryBot: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
Message marshalling makes use of BytesOrPanic a lot, under the
assumption that it will never panic. This assumption was incorrect, and
specifically crafted handshakes could trigger panics. Rather than just
surgically replacing the usages of BytesOrPanic in paths that could
panic, replace all usages of it with proper error returns in case there
are other ways of triggering panics which we didn't find.
In one specific case, the tree routed by expandLabel, we replace the
usage of BytesOrPanic, but retain a panic. This function already
explicitly panicked elsewhere, and returning an error from it becomes
rather painful because it requires changing a large number of APIs.
The marshalling is unlikely to ever panic, as the inputs are all either
fixed length, or already limited to the sizes required. If it were to
panic, it'd likely only be during development. A close inspection shows
no paths for a user to cause a panic currently.
This patches ends up being rather large, since it requires routing
errors back through functions which previously had no error returns.
Where possible I've tried to use helpers that reduce the verbosity
of frequently repeated stanzas, and to make the diffs as minimal as
possible.
Thanks to Marten Seemann for reporting this issue.
Updates #58001Fixes#58359
Fixes CVE-2022-41724
Change-Id: Ieb55867ef0a3e1e867b33f09421932510cb58851
Reviewed-on: https://team-review.git.corp.google.com/c/golang/go-private/+/1679436
Reviewed-by: Julie Qiu <julieqiu@google.com>
TryBot-Result: Security TryBots <security-trybots@go-security-trybots.iam.gserviceaccount.com>
Run-TryBot: Roland Shoemaker <bracewell@google.com>
Reviewed-by: Damien Neil <dneil@google.com>
(cherry picked from commit 1d4e6ca9454f6cf81d30c5361146fb5988f1b5f6)
Reviewed-on: https://team-review.git.corp.google.com/c/golang/go-private/+/1728205
Reviewed-by: Tatiana Bradley <tatianabradley@google.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/468121
Reviewed-by: Than McIntosh <thanm@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
TryBot-Bypass: Michael Pratt <mpratt@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
Reader.ReadForm is documented as storing "up to maxMemory bytes + 10MB"
in memory. Parsed forms can consume substantially more memory than
this limit, since ReadForm does not account for map entry overhead
and MIME headers.
In addition, while the amount of disk memory consumed by ReadForm can
be constrained by limiting the size of the parsed input, ReadForm will
create one temporary file per form part stored on disk, potentially
consuming a large number of inodes.
Update ReadForm's memory accounting to include part names,
MIME headers, and map entry overhead.
Update ReadForm to store all on-disk file parts in a single
temporary file.
Files returned by FileHeader.Open are documented as having a concrete
type of *os.File when a file is stored on disk. The change to use a
single temporary file for all parts means that this is no longer the
case when a form contains more than a single file part stored on disk.
The previous behavior of storing each file part in a separate disk
file may be reenabled with GODEBUG=multipartfiles=distinct.
Update Reader.NextPart and Reader.NextRawPart to set a 10MiB cap
on the size of MIME headers.
Thanks to Jakob Ackermann (@das7pad) for reporting this issue.
Updates #58006Fixes#58363
Fixes CVE-2022-41725
Change-Id: Ibd780a6c4c83ac8bcfd3cbe344f042e9940f2eab
Reviewed-on: https://team-review.git.corp.google.com/c/golang/go-private/+/1714276
Reviewed-by: Julie Qiu <julieqiu@google.com>
TryBot-Result: Security TryBots <security-trybots@go-security-trybots.iam.gserviceaccount.com>
Reviewed-by: Roland Shoemaker <bracewell@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
(cherry picked from commit 7d0da0029bfbe3228cc5216ced8c7b3184eb517d)
Reviewed-on: https://team-review.git.corp.google.com/c/golang/go-private/+/1728950
Reviewed-by: Damien Neil <dneil@google.com>
Run-TryBot: Roland Shoemaker <bracewell@google.com>
Reviewed-by: Tatiana Bradley <tatianabradley@google.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/468120
Auto-Submit: Michael Pratt <mpratt@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This patch detects at which index position profiling samples that have
the value-type samples count are, instead of the previously hard-coded
position of index 1. Runtime generated profiles always generate CPU
profiling data with the 0 index being CPU nanoseconds, and samples count
at index 1, which is why this previously hasn't come up.
This is a redo of CL 465135, now allowing empty profiles. Note that
preprocessProfileGraph will already cause pgo.New to return nil for
empty profiles.
For #58292
For #58309
Change-Id: Ia6c94f0793f6ca9b0882b5e2c4d34f38e600c1e3
Reviewed-on: https://go-review.googlesource.com/c/go/+/467375
Run-TryBot: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
In general it seems ok to assume that an open-source module that did
exist will continue to do so — after all, users of open-source modules
already do that all the time. However, we should not assume that those
modules do not publish new versions — that's really up to their
maintainers to decide.
Two existing tests did make that assumption for the module
gopkg.in/natefinch/lumberjack.v2. Let's remove those two tests.
If we need to replace them at some point, we can replace them with
hermetic test-only modules (#54503) or perhaps modules owned by the Go
project.
Updates #58445.
Fixes#58450.
Change-Id: Ica8fe587d86fc41f3d8445a4cd2b8820455ae45f
Reviewed-on: https://go-review.googlesource.com/c/go/+/466861
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: David Chase <drchase@google.com>
If you use an external linker with --gc-sections, nothing refers
to .go.buildinfo, so the section is deleted, which in turns makes
'go version' fail on the binary. It is important for vulnerability
scanning and the like to be able to run 'go version' on any binary.
Fix this by inserting a reference to .go.buildinfo from the rodata
section, which will not be GC'ed.
Fixes#58222.
Fixes#58224.
Change-Id: I1e13e9464acaf2f5cc5e0b70476fa52b43651123
Reviewed-on: https://go-review.googlesource.com/c/go/+/464435
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-on: https://go-review.googlesource.com/c/go/+/464796
In the types1 universe under the unified frontend, we never need to
worry about type parameter constraints, so we only see pure
interfaces. However, we might still see interfaces that contain union
types, because of interfaces like "interface{ any | int }" (equivalent
to just "any").
We can handle these without needing to actually represent type unions
within types1 by simply mapping any union to "any".
Fixes#58413.
Change-Id: I5e4efcf0339edbb01f4035c54fb6fb1f9ddc0c65
Reviewed-on: https://go-review.googlesource.com/c/go/+/458619
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
(cherry picked from commit a7de684e1b)
Reviewed-on: https://go-review.googlesource.com/c/go/+/466435
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: David Chase <drchase@google.com>
Blank node must be ignored when building arguments substitued tree.
Otherwise, it could be used to replace other blank node in left hand
side of an assignment, causing an invalid IR node.
Consider the following code:
type S1 struct {
s2 S2
}
type S2 struct{}
func (S2) Make() S2 {
return S2{}
}
func (S1) Make() S1 {
return S1{s2: S2{}.Make()}
}
var _ = S1{}.Make()
After staticAssignInlinedCall, the assignment becomes:
var _ = S1{s2: S2{}.Make()}
and the arg substitued tree is "map[*ir.Name]ir.Node{_: S1{}}". Now,
when doing static assignment, if there is any assignment to blank node,
for example:
_ := S2{}
That blank node will be replaced with "S1{}":
S1{} := S2{}
So constructing an invalid IR which causes the ICE.
Fixes#58335
Change-Id: I21b48357f669a7e02a7eb4325246aadc31f78fb9
Reviewed-on: https://go-review.googlesource.com/c/go/+/465098
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/466275
Reviewed-by: Than McIntosh <thanm@google.com>
CL 444277 fixed Time.UnmarshalText and Time.UnmarshalJSON to properly
unmarshal timestamps according to RFC 3339 instead of according
to Go's bespoke time syntax that is a superset of RFC 3339.
However, this change seems to have broken an AWS S3 unit test
that relies on parsing timestamps with single digit hours.
It is unclear whether S3 emits these timestamps in production or
whether this is simply a testing artifact that has been cargo culted
across many code bases. Either way, disable strict parsing for now
and re-enable later with better GODEBUG support.
Updates #54580
Change-Id: Icced2c7f9a6b2fc06bbd9c7e90f90edce24c2306
Reviewed-on: https://go-review.googlesource.com/c/go/+/462286
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-on: https://go-review.googlesource.com/c/go/+/462675
Reviewed-by: Joseph Tsai <joetsai@digital-static.net>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Unified IR added several new IR fields for holding *runtime._type
expressions. To avoid throwing off any frontend semantics
(particularly inlining cost heuristics), they were marked as
`mknode:"-"` so that code wouldn't visit them.
Unfortunately, this has a bad interaction with the static init
inlining optimization, because the latter relies on ir.EditChildren to
substitute all parameters. This potentially includes dictionary
parameters, which can appear within the new RType fields.
This CL adds a new ir.EditChildrenWithHidden function that also edits
these fields, and switches staticinit to use it. Longer term, we
should unhide the RType fields so that ir.EditChildren visits them
normally, but that's scarier so late in the release cycle.
Updates #57778.
Updates #57854.
Change-Id: I98c1e8cf366156dc0c81a0cb79029cc5e59c476f
Reviewed-on: https://go-review.googlesource.com/c/go/+/461686
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
(cherry picked from commit 9f2fbedf010d59c3ecaa8c25b07a5f68fcb2e3d5)
Reviewed-on: https://go-review.googlesource.com/c/go/+/462535
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
We need to avoid nospill registers at this point in regalloc.
Make sure that we don't restrict our register set to avoid registers
desired by other instructions, if the resulting set includes only
nospill registers.
Fixes#57846
Change-Id: I05478e4513c484755dc2e8621d73dac868e45a27
Reviewed-on: https://go-review.googlesource.com/c/go/+/461685
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This patch ensures that the go command's "list" subcommand accepts
coverage-related build options, which were incorrectly left out when
"go build -cover" was rolled out. This is needed in order to do things
like check the staleness of an installed cover-instrumented target.
Fixes#57785.
Change-Id: I140732ff1e6b83cd9c453701bb8199b333fc0f2e
Reviewed-on: https://go-review.googlesource.com/c/go/+/462116
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
In particular, CheckSignatureFrom just can't check the path length
limit, because it might be enforced above the parent.
We don't need to document the supported signature algorithms for
CheckSignatureFrom, since we document at the constants in what contexts
they are allowed and not. That does leave CheckSignature ambiguous,
though, because that function doesn't have an explicit context.
Change-Id: I4c107440a93f60bc0de07df2b7efeb1a4a766da0
Reviewed-on: https://go-review.googlesource.com/c/go/+/460537
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
At the moment the only documentation is the release notes,
but everything mentioned in the release notes should have
proper documentation separate from them.
Change-Id: I9885962f6c6d947039b0be59b608385781479271
Reviewed-on: https://go-review.googlesource.com/c/go/+/462196
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
The arena goexperiment contains code used inside Google in very
limited use cases that we will maintain, but the discussion on #51317
identified serious problems with the very idea of adding arenas to the
standard library. In particular the concept tends to infect many other
APIs in the name of efficiency, a bit like sync.Pool except more
publicly visible.
It is unclear when, if ever, we will pick up the idea and try to push
it forward into a public API, but it's not going to happen any time
soon, and we don't want users to start depending on it: it's a true
experiment and may be changed or deleted without warning.
The arena text in the release notes makes them seem more official
and supported than they really are, and we've already seen a couple
blog posts based on that erroneous belief. Delete the text to try to
set expectations better.
Change-Id: I4f6e328ac470a9cd410f5f722d0769ef62d5e5ba
Reviewed-on: https://go-review.googlesource.com/c/go/+/462355
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Eli Bendersky <eliben@google.com>
In the fix for 54332 the MOVD R1, R1 instruction was added to
morestack_noctxt function to set the SPWRITE bit. However, the
instruction MOVD R1, R1 results in or r1,r1,r1 which is a special
instruction on ppc64 architecture as it changes the thread priority
and can negatively impact performance in some cases.
More details on such similar nops can be found in Power ISA v3.1
Book II on Power ISA Virtual Environment architecture in the chapter
on Program Priority Registers and Or instructions.
Replacing this by OR R0, R1 has the same affect on setting SPWRITE as
needed by the first fix but does not affect thread priority and
hence does not cause the degradation in performance
Hash65536-64 2.81GB/s ±10% 16.69GB/s ± 0% +494.44%
Fixes#57741
Change-Id: Ib912e3716c6afd277994d6c1c5b2891f82225d50
Reviewed-on: https://go-review.googlesource.com/c/go/+/461597
Reviewed-by: Benny Siegert <bsiegert@gmail.com>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Auto-Submit: Benny Siegert <bsiegert@gmail.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
Merge List:
+ 2023-01-11 245e95dfab go/types, types2: don't look up fields or methods when expecting a type
+ 2023-01-11 18625d9bec runtime: fix incorrect comment
+ 2023-01-11 6ad27161f8 cmd/compile: better error message for when a type is in a constraint but not the type set
+ 2023-01-10 76d39ae349 cmd/link, runtime: Apple libc atfork workaround take 3
+ 2023-01-10 0a0de0fc42 runtime: revert use of __fork to work around Apple atfork bugs
+ 2023-01-10 82f09b75ca os/exec: avoid leaking an exec.Cmd in TestWaitInterrupt
+ 2023-01-09 0202ad0b3a cmd/compile: prevent IsNewObject from taking quadratic time
+ 2023-01-09 64519baf38 cmd/compile/internal/pgo: add hint to missing start_line error
+ 2023-01-09 376076f3c6 runtime: skip TestCgoPprofCallback in short mode, don't run in parallel
+ 2023-01-09 0bbd67e52f runtime/pprof: document possibility of empty stacks
+ 2023-01-09 d9f23cfe78 runtime/pprof: improve output of TestLabelSystemstack
+ 2023-01-09 8232a09e3e sync/atomic: fix the note of atomic.Store
+ 2023-01-09 841c3eb166 all: fix typos in go file comments
+ 2023-01-06 f721fa3be9 syscall: skip TestUseCgroupFD if cgroupfs not mounted
+ 2023-01-06 76ec919237 net: fix typo in hosts.go
+ 2023-01-06 660d4815ea cmd/compile: describe how Go maps to wasm implementation
+ 2023-01-05 119f679a3b crypto/tls: fix typo in cacheEntry godoc
+ 2023-01-05 d50ea217f6 cmd/cover: fix problems with "go test -covermode=atomic sync/atomic"
+ 2023-01-04 bae7d772e8 doc/go1.20: fix links to new strings functions
+ 2023-01-04 4e7c838483 crypto/internal/boring: add dev.boringcrypto README.md text
+ 2023-01-04 46e3d9d12a cmd/compile: use "satisfies" (not "implements") for constraint errors
+ 2023-01-04 79cdecc852 cmd/gofmt: fix a typo in a comment
+ 2023-01-03 9955a7e9bb README.vendor: minor updates
+ 2023-01-03 d03231d9ce doc/go1.20: fix http.ResponseController example
+ 2023-01-03 cdc73f0679 .github: suggest using private browsing in pkgsite template
Change-Id: I73be496aa4163ad1d3a6cc8114f1a612968d4b10
As we have seen many times, the type checker must be careful to avoid
accessing named type information before the type is fully set up. We
need a more systematic solution to this problem, but for now avoid one
case that causes a crash: checking a selector expression on an
incomplete type when a type expression is expected.
For golang/go#57522
Change-Id: I7ed31b859cca263276e3a0647d1f1b49670023a9
Reviewed-on: https://go-review.googlesource.com/c/go/+/461577
Run-TryBot: Robert Findley <rfindley@google.com>
Auto-Submit: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
CL 451735 worked around bugs in Apple's atfork handlers by calling
notify_is_valid_token and xpc_atfork_child at startup, so that init
code that wouldn't be safe in the child process would be warmed up in
the parent process instead, but xpc_atfork_child broke use of the xpc
library in Go programs, and xpc is internally used by various macOS
frameworks (#57263).
CL 459175 reverted that change, and then CL 459176 tried a new
approach: use __fork, which doesn't call any of the atfork handlers at all.
That worked, but an Apple engineer reviewing the change in private
email suggests that since __fork is not public API, it should be avoided.
The same engineer (with access to the source code for the xpc library)
suggests that the breakage in #57263 is caused by xpc_atfork_child
marking the library as unusable, expecting an imminent call to exec,
and that calling xpc_date_create_from_current instead would do the
necessary initialization without marking xpc as unusable.
CL 460475 reverted that change, to prepare for this one.
This CL goes back to the original “call functions to warm things up”
approach, replacing xpc_atfork_child with xpc_date_create_from_current.
The CL also updates cmd/link to use OS and SDK version 10.13.0 for
x86 macOS binaries, up from 10.9.0, also suggested by the Apple engineer.
Combined with the two warmup calls, this makes the fork hangs go away.
The minimum macOS version has been 10.13 High Sierra since Go 1.17,
so there should be no problem with writing that in the binaries too.
Fixes#33565.
Fixes#56784.
Fixes#57263.
Fixes#57577.
Change-Id: I20769d9daa1fe9ea930f8009481335f8a14dc21b
Reviewed-on: https://go-review.googlesource.com/c/go/+/460476
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
An Apple engineer suggests that since __fork is not public API,
it would be better to use a different fix. With the benefit of source code,
they suggest using xpc_date_create_from_current instead of
xpc_atfork_child. The latter sets some flags that disable certain
functionality for the remainder of the process lifetime (expecting exec),
while the former should do the necessary setup.
Reverting the __fork fix in order to prepare a clean fix based
on CL 451735 using xpc_date_create_from_current.
This reverts commit c61d322d5f.
Change-Id: I2da293ff537237ffd2d40ad756d827c95c84635b
Reviewed-on: https://go-review.googlesource.com/c/go/+/460475
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
In CL 436655 I added a GODEBUG setting to this test process to verify
that Wait is eventually called for every exec.Cmd before it becomes
unreachable. However, the cmdHang test helpers in
TestWaitInterrupt/Exit-hang and TestWaitInterrupt/SIGKILL-hang
intentially leak a subprocess in order to simulate a leaky third-party
program, as Go users might encounter in practical use.
To avoid tripping over the leak check, we call Wait on the leaked
subprocess in a background goroutine. Since we expect the process
running cmdHang to exit before its subprocess does, the call to Wait
should have no effect beyond suppressing the leak check.
Fixes#57596.
Updates #52580.
Updates #50436.
Change-Id: Ia4b88ea47fc6b605c27ca6d9d7669c874867a900
Reviewed-on: https://go-review.googlesource.com/c/go/+/460998
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
As part of IsNewObject, we need to go from the SelectN[0] use of
a call to the SelectN[1] use of a call. The current code does this
by just looking through the block. If the block is very large,
this ends up taking quadratic time.
Instead, prepopulate a map from call -> SelectN[1] user of that call.
That lets us find the SelectN[1] user in constant time.
Fixes#57657
Change-Id: Ie2e0b660af5c080314f4f17ba2838510a1147f9e
Reviewed-on: https://go-review.googlesource.com/c/go/+/461080
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Profiles only began adding Function.start_line in 1.20. If it is
missing, add a hint to the error message that they may need to profile a
build of the application built with a newer version of the toolchain.
Technically profiles are not required to come from Go itself (e.g., they
could be converted from perf), but in practice they most likely are.
Fixes#57674.
Change-Id: I87eca126d3fed0cff94bbb8dd748bd4652f88b12
Reviewed-on: https://go-review.googlesource.com/c/go/+/461195
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
I spent quite a while determining the cause of empty stacks in
profiles and reasoning out why this is okay. There isn't a great place
to record this knowledge, but a documentation comment on
appendLocsForStack is better than nothing.
Updates #51550.
Change-Id: I2eefc6ea31f1af885885c3d96199319f45edb4ce
Reviewed-on: https://go-review.googlesource.com/c/go/+/460695
Reviewed-by: Felix Geisendörfer <felix.geisendoerfer@datadoghq.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
The current output of TestLabelSystemstack is a bit cryptic. This CL
improves various messages and hopefully simplifies the logic in the
test.
Simplifying the logic leads to three changes in possible outcomes,
which I verified by running the logic before and after this change
through all 2^4 possibilities (https://go.dev/play/p/bnfb-OQCT4j):
1. If a sample both must be labeled and must not be labeled, the test
now reports that explicitly rather than giving other confusing output.
2. If a sample must not be labeled but is, the current logic will
print two identical error messages. The new logic prints only one.
3. If the test finds no frames at all that it recognizes, but the
sample is labeled, it will currently print a confusing "Sample labeled
got true want false" message. The new logic prints nothing. We've seen
this triggered by empty stacks in profiles.
Fixes#51550. This bug was caused by case 3 above, where it was
triggered by a profile label on an empty stack. It's valid for empty
stacks to appear in a profile if we sample a goroutine just as it's
exiting (and that goroutine may have a profile label), so the test
shouldn't fail in this case.
Change-Id: I1593ec4ac33eced5bb89572a3ba7623e56f2fb3d
Reviewed-on: https://go-review.googlesource.com/c/go/+/460516
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Felix Geisendörfer <felix.geisendoerfer@datadoghq.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This patch fixes an elderly bug with "go test -covermode=atomic
sync/atomic". Change the cover tool to avoid adding an import of
sync/atomic when processing "sync/atomic" itself in atomic mode;
instead make direct calls to AddUint32/StoreUint32. In addition,
change the go command to avoid injecting an artificial import of
"sync/atomic" for sync/atomic itself.
Fixes#57445.
Change-Id: I8c8fbd0bcf26c8a8607d4806046f826296508c74
Reviewed-on: https://go-review.googlesource.com/c/go/+/459335
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Per the latest spec, we distinguish between interface implementation
and constraint satisfaction. Use the verb "satisfy" when reporting
an error about failing constraint satisfaction.
This CL only changes error messages. It has no impact on correct code.
Fixes#57564.
Change-Id: I6dfb3b2093c2e04fe5566628315fb5f6bd709f17
Reviewed-on: https://go-review.googlesource.com/c/go/+/460396
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
I opened #54872 without considering my browser's extensions and one of
them ended up being the cause of my issue. This seems to be a common
error.
To help that, add a suggestion to use a private/incognito tab/window
to the pkgsite issue template when reproducing issues. This probably
would have been enough for me to figure things out before opening my
issue.
Updates #54872, #47213
Change-Id: Ic61a3462cb902c91554cf9432aaae1222c6a991e
Reviewed-on: https://go-review.googlesource.com/c/go/+/427962
Run-TryBot: Dan Peterson <danp@danp.net>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Hyang-Ah Hana Kim <hyangah@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Hyang-Ah Hana Kim <hyangah@gmail.com>
The libFuzzer generated binary by default writes failure input
into the current directory. Set cmd.Dir to the temporary directory
so it won't write to GOROOT when running the test.
Change-Id: I3e4ce7e3f845be5c9f09511c36e7a9a396eafad2
Reviewed-on: https://go-review.googlesource.com/c/go/+/459556
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently in libfuzzer mode, we put our counters in section
__sancov_cntrs. When linking with C/C++ code that also has fuzzer
counters, apparently the C linker combines our counters and their
counters and registers them together. But in the Go runtime we
also have code to register our counters. So the Go counters ended
up registered twice, causing problems.
Since we already have code to register our counters, put them in
a Go-specific section so it won't be combined with the C counters.
Fixes#57449.
Change-Id: If3d41735124e7e301572d4b7aecf7d057ac134c0
Reviewed-on: https://go-review.googlesource.com/c/go/+/459055
Reviewed-by: Nicolas Hillegeer <aktau@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Issues #33565 and #56784 were caused by hangs in the child process
after fork, while it ran atfork handlers that ran into slow paths that
didn't work in the child.
CL 451735 worked around those two issues by calling a couple functions
at startup to try to warm up those child paths. That mostly worked,
but it broke programs using cgo with certain macOS frameworks (#57263).
CL 459175 reverted CL 451735.
This CL introduces a different fix: bypass the atfork child handlers
entirely. For a general fork call where the child and parent are both
meant to keep executing the original program, atfork handlers can be
necessary to fix any state that would otherwise be tied to the parent
process. But Go only uses fork as preparation for exec, and it takes
care to limit what it attempts to do in the child between the fork and
exec. In particular it doesn't use any of the things that the macOS
atfork handlers are trying to fix up (malloc, xpc, others). So we can
use the low-level fork system call (__fork) instead of the
atfork-wrapped one.
The full list of functions that can be called in a child after fork in
exec_libc2.go is:
- ptrace
- setsid
- setpgid
- getpid
- ioctl
- chroot
- setgroups
- setgid
- setuid
- chdir
- dup2
- fcntl
- close
- execve
- write
- exit
I disassembled all of these while attached to a hung exec.test binary
and confirmed that nearly all of them are making direct kernel calls,
not using anything that the atfork handler needs to fix up.
The exceptions are ioctl, fcntl, and exit.
The ioctl and fcntl implementations do some extra work around the
kernel call but don't call any other functions, so they should still
be OK. (If not, we could use __ioctl and __fcntl instead, but without
a good reason, we should keep using the standard entry points.)
The exit implementation calls atexit handlers. That is almost
certainly inappropriate in a failed fork child, so this CL changes
that call to __exit on darwin. To avoid making unnecessary changes at
this point in the release cycle, this CL leaves OpenBSD calling plain
exit, even though that is probably a bug in the OpenBSD port
(filed #57446).
Fixes#33565.
Fixes#56784.
Fixes#57263.
Change-Id: I26812c26a72bdd7fcf72ec41899ba11cf6b9c4ab
Reviewed-on: https://go-review.googlesource.com/c/go/+/459176
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Revert CL 451735 (1f4394a0c9), which fixed#33565 and #56784
but also introduced #57263.
I have a different fix to apply instead. Since the first fix was
never backported, it will be easiest to backport the new fix
if the new fix is done in a separate CL from the revert.
Change-Id: I6c8ea3a46e542ee4702675bbc058e29ccd2723e0
Reviewed-on: https://go-review.googlesource.com/c/go/+/459175
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
The runtime/internal/startlinetest package contains a call to a
function defined in runtime_test. Generally this is fine as this
package is only linked in for runtime_test. Except that for "go
install -buildmode=shared std", which include all packages in std,
including this test-only internal package. In this mode, the
caller is included in the linking but the callee is not, causing
linking error. Work around it by calling
runtime_test.callerStartLine via a function pointer. The function
pointer is only set in runtime_test. In the shared std build, the
function pointer will not be set, and this is fine.
Fixes#57334.
Change-Id: I7d871c50ce6599c6ea2802cf6e14bb749deab220
Reviewed-on: https://go-review.googlesource.com/c/go/+/458696
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Only one of the runtime/race/internal/amd64vN packages should be
included in a build. Generally this is true because the
runtime/race package would import only one of them depending on
the build configuration. But for "go install -buildmode=shared std"
it includes all Go packages in std, which includes both, which
then causes link-time failure due to duplicated symbols. To avoid
this, we add build tags to the internal packages, so, depending on
the build configuation, only one package would contain buildable
go files therefore be included in the build.
For #57334.
Change-Id: I52ddc3a40e16c7d04b4dd861e9689918d27e8509
Reviewed-on: https://go-review.googlesource.com/c/go/+/458695
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
The processPod() helper (invoked by processCoverTestDir, which is in
turn called by _testmain.go) was opening and reading counter data
files, but never closing them. Add a call to close the files after
they have been read.
Fixes#57407.
Change-Id: If9a489f92e4bab72c5b2df8697e14420a6f7b8f5
Reviewed-on: https://go-review.googlesource.com/c/go/+/458835
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
CL 452616 disables path security checks by default, enabling them
only when GODEBUG=tarinsecurepath=0 or GODEBUG=zipinsecurepath=0
is set. Remove now-obsolete documenation of the path checks.
For #55356
Change-Id: I4ae57534efe9e27368d5e67773a502dd0e56eff4
Reviewed-on: https://go-review.googlesource.com/c/go/+/458875
Reviewed-by: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
This reverts CL 407414.
When forwarding an inbound request that contains an existing
X-Forwarded-Host or X-Forwarded-Proto header, a proxy might want
to preserve the header from the inbound request, replace it with
its own header, or not include any header at all.
CL 407414 replaces inbound X-Forwarded-{Host,Proto} headers by default,
and allows a Director func to disable sending these headers at all.
However, the Director hook API isn't sufficiently flexible to permit the
previous behavior of preserving inbound values unchanged.
The new Rewrite API does have this flexibility; users of Rewrite can
easily pick the exact behavior they want.
Revert the change to ReverseProxy when using a Director func.
Users who want a convenient way to set X-Forwarded-* headers to
reasonable values can migrate to Rewrite at their convenience,
and users depending on the current behavior will be unaffected.
For #50465.
Fixes#57132.
Change-Id: Ic42449c1bb525d6c9920bf721efbc519697f4f20
Reviewed-on: https://go-review.googlesource.com/c/go/+/457595
Run-TryBot: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
The function LoweredAtomicCas32 is implemented using the LL-SC instruction pair
on loong64, mips64x, riscv64. However,the LL instruction on loong64, mips64x,
riscv64 is sign-extended, so it is necessary to sign-extend the 2nd parameter
"old" of the LoweredAtomicCas32, so that the instruction BNE after LL can get
the desired result.
The function prototype of LoweredAtomicCas32 in golang:
func Cas32(ptr *uint32, old, new uint32) bool
When using an intrinsify implementation:
case 1: (*ptr) <= 0x80000000 && old < 0x80000000
E.g: (*ptr) = 0x7FFFFFFF, old = Rarg1= 0x7FFFFFFF
After run the instruction "LL (Rarg0), Rtmp": Rtmp = 0x7FFFFFFF
Rtmp ! = Rarg1(old) is false, the result we expect
case 2: (*ptr) >= 0x80000000 && old >= 0x80000000
E.g: (*ptr) = 0x80000000, old = Rarg1= 0x80000000
After run the instruction "LL (Rarg0), Rtmp": Rtmp = 0xFFFFFFFF_80000000
Rtmp ! = Rarg1(old) is true, which we do not expect
When using an non-intrinsify implementation:
Because Rarg1 is loaded from the stack using sign-extended instructions
ld.w, the situation described in Case 2 above does not occur
Benchmarks on linux/loong64:
name old time/op new time/op delta
Cas 50.0ns ± 0% 50.1ns ± 0% ~ (p=1.000 n=1+1)
Cas64 50.0ns ± 0% 50.1ns ± 0% ~ (p=1.000 n=1+1)
Cas-4 56.0ns ± 0% 56.0ns ± 0% ~ (p=1.000 n=1+1)
Cas64-4 56.0ns ± 0% 56.0ns ± 0% ~ (p=1.000 n=1+1)
Benchmarks on Loongson 3A4000 (GOARCH=mips64le, 1.8GHz)
name old time/op new time/op delta
Cas 70.4ns ± 0% 70.3ns ± 0% ~ (p=1.000 n=1+1)
Cas64 70.7ns ± 0% 70.6ns ± 0% ~ (p=1.000 n=1+1)
Cas-4 81.1ns ± 0% 80.8ns ± 0% ~ (p=1.000 n=1+1)
Cas64-4 80.9ns ± 0% 80.9ns ± 0% ~ (p=1.000 n=1+1)
Fixes#57282
Change-Id: I190a7fc648023b15fa392f7fdda5ac18c1561bac
Reviewed-on: https://go-review.googlesource.com/c/go/+/457135
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Wayne Zuo <wdvxdr@golangcn.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: David Chase <drchase@google.com>
Some compilers default to having -fstack-protector on, which breaks
when using internal linking because the linker doesn't know how to
find the support functions.
Updates #52919.
Updates #54313.
Fixes#57261.
Change-Id: Iaae731851407af4521fff2dfefc5b7e3e92cf284
Reviewed-on: https://go-review.googlesource.com/c/go/+/456855
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Provide more information about why this test might be hanging waiting
for PutIdleConn to be called (#56587): If the round trip that should
result in PutIdleConn being invoked completes, report that to the
goroutine waiting for PutIdleConn.
For #56587
Change-Id: Ie476ea0ce4a48d2bda6b9b109f89d675a10e7e45
Reviewed-on: https://go-review.googlesource.com/c/go/+/457775
Auto-Submit: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
I made this test parallel in CL 439196, which exposed it to the
fork/exec race condition described in #22315. The ETXTBSY errors from
that race should resolve on their own, so we can simply retry the call
to get past them.
Fixes#56811.
Updates #22315.
Change-Id: I2c6aa405bf3a1769d69cf08bf661a9e7f86440b4
Reviewed-on: https://go-review.googlesource.com/c/go/+/458016
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Devirtualization can turn OCALLINTER into OCALLMETH, but then we want
to actually desugar into OCALLFUNC instead for later phases. Just
needs a missing call to typecheck.FixMethodCall.
Fixes#57309.
Change-Id: I331fbd40804e1a370134ef17fa6dd501c0920ed3
Reviewed-on: https://go-review.googlesource.com/c/go/+/457715
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
- Rephrase the notion of "comparability" from a property
of values (operands) to a property of types and adjust
dependent prose.
- Introduce the notion of "strict comparability".
- Fix the definitions of comparability for type interfaces
and type parameters.
- Define the predeclared identifier "comparable" as stricly
comparable.
These changes address existing problems in the spec as outlined
in the section on "Related spec issues" in issue #56548.
For #56548.
Change-Id: Ibc8c2f36d92857a5134eadc18358624803d3dd21
Reviewed-on: https://go-review.googlesource.com/c/go/+/457095
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Bypass: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Rather than reporting an impossible conversion error when mixing an
untyped value with a pointer type in an operation, report a type
mismatch error. This fixes a regression in error quality compared
to pre-1.18.
For the fix, clean up the implementation of canMix, add documentation,
and give it a better name.
Adjust test case for corresponding error code bacause we now get a
better error message (and error code) for the old error code example.
Fixes#57160.
Change-Id: Ib96ce7cbc44db6905fa2f1c90a3769af609e101b
Reviewed-on: https://go-review.googlesource.com/c/go/+/457055
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
The comment on `slicebytetostringtmp` mention that `==` operator does
not allocate []byte to string conversion, but the test was testing only
`==` and `!=` and the compiler actually optimizes all comparison
operators.
Also added a test for concatenation comparison, which also should not
allocate.
Change-Id: I6f4c5c4f238808138fa901732e1fd5b6ab25f725
Reviewed-on: https://go-review.googlesource.com/c/go/+/456415
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Auto-Submit: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Previous implementation has a check on pointer size but only if ptrSize
is 4. but it does not check on ptrSize 8 causing the panic on read for
some inputs.
The explicit check for pointer size 8 is added and error is returned if
ptrSize is not 4 and not 8.
Fixes#57002
Change-Id: Id51de72bdef4da9955d086bfc2a5d735678ee2ac
Reviewed-on: https://go-review.googlesource.com/c/go/+/454616
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Similar to CL 456556 but for ppc64 instead of arm64.
Change docs about how booleans are stored in registers for ppc64.
We now don't promise to keep the upper bits zeroed; they might be junk.
To test, I changed the boolean generation instructions (MOVBZload* and ISEL*
with boolean type) to OR in 0x100 to the result. all.bash still passed,
so I think nothing else is depending on the upper bits of booleans.
Update #57184
Change-Id: Ie66f8934a0dafa34d0a8c2a37324868d959a852c
Reviewed-on: https://go-review.googlesource.com/c/go/+/456437
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: KAMPANAT THUMWONG (KONG PC) <1992kongpc.kth@gmail.com>
Run-TryBot: Archana Ravindar <aravind5@in.ibm.com>
Tested by temporarily changing sysdir to use a directory where
the expected files were all symlinks. We should consider using
a different approach that doesn't rely on sysdir, but for now
do a minimal fix.
Fixes#57210
Change-Id: Ifb1becef03e014ceb48290ce13527b3e103c0e07
Reviewed-on: https://go-review.googlesource.com/c/go/+/456557
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
ARM64 maintains booleans in the low byte of registers. Upper parts
of that register are junk.
This rule is using all 32 bits of a boolean-containing register, which
is wrong. Change the rule to only look at the low bit.
Fixes#57184
Change-Id: Ibbef86b2be859df3d06d993db00e1231c481c428
Reviewed-on: https://go-review.googlesource.com/c/go/+/456556
Auto-Submit: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
In the wrappers for getgrnam_r and similar, the structs to be returned
are allocated on the C stack and may be uninitialized. If the call to
the wrapped C function returns an error (such as ERANGE), it may leave
the struct uninitialized, expecting that the caller will not read it.
However, when that struct is returned to Go, it may be read by the Go
garbage collector. If the uninitialized struct fields happen to
contain wild pointers, the Go garbage collector will throw an error.
(Prior to CL 449335, the Go runtime would not scan the struct fields
because they did not reside in Go memory.)
Fix this by always zeroing the struct before the C call.
Fixes#57170.
Change-Id: I241ae8e4added6f9a406dac37a7f6452341aa0cf
Reviewed-on: https://go-review.googlesource.com/c/go/+/456121
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Type inference uses a trick of "renaming" type parameters in the type
parameter list to avoid cycles during unification. This separates the
identity of type parameters from type arguments. When this trick was
introduced in CL 385494, we restricted its application to scenarios
where inference is truly self-recursive: the type parameter list being
inferred was the same as the type parameter list of the outer function
declaration. Unfortunately, the heuristic used to determine
self-recursiveness was flawed: type-checking function literals clobbers
the type-checker environment, losing information about the outer
signature.
We could fix this by introducing yet more state into the type-checker
(e.g. a 'declSig' field that would hold the signature of the active
function declaration), but it is simpler to just avoid this optimization
and always perform type parameter renaming. We can always optimize
later.
This CL removes the check for true self-recursion, always performing the
renaming.
Fixesgolang/go#57155
Change-Id: I34c7617005c1f0ccfe2192da0e5ed104be6b92c9
Reviewed-on: https://go-review.googlesource.com/c/go/+/456236
Run-TryBot: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Fix a buglet in cmd/cover in how we handle package name/path for the
"go build -o foo.exe *.go" and "go run *.go" cases.
The go command assigns a dummy import path of "command-line-arguments"
to the main package built in these cases; rather than expose this
dummy to the user in coverage reports, the cover tool had a special
case hack intended to rewrite such package paths to "main". The hack
was too general, however, and was rewriting the import path of all
packages with (p.name == "main") to an import path of "main". The hack
also produced unexpected results for cases such as
go test -cover foo.go foo_test.go
This patch removes the hack entirely, leaving the package path for
such cases as "command-line-arguments".
Fixes#57169.
Change-Id: Ib6071db5e3485da3b8c26e16ef57f6fa1712402c
Reviewed-on: https://go-review.googlesource.com/c/go/+/456237
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
These short timeouts can overrun due to system scheduling delay
(or GC latency) on a slow or heavily-loaded host.
Moreover, if the test deadlocks we will probably want to know what the
GC goroutines were doing at the time. With an arbitrary timeout, we
never get that information; however, if we allow the test to time out
completely we will get a goroutine dump (and, if GOTRACEBACK is
configured in the environment, that may even include GC goroutines).
Fixes#57166.
Change-Id: I136501883373c3ce4e250dc8340c60876b375f44
Reviewed-on: https://go-review.googlesource.com/c/go/+/456118
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
- Start handling signals in 'go test' just before starting the test
subprocess instead of just after. (It is unlikely that starting the
process will cause cmd/go to hang in a way that requires signals to
debug, and it is possible that something the test does — such as
sending os.Interrupt to its parent processes — will immediately
send a signal that needs to be handled.)
- In the test-test, don't try to re-parse the parent PIDs after
sending signals, and sleep for a much shorter time interval.
(Overrunning the sleep caused the next call to strconv.Atoi — which
shouldn't even happen! — to fail with a parse error, leading to the
failure mode observed in
https://build.golang.org/log/f0982dcfc6a362f9c737eec3e7072dcc7ef29e32.)
Fixes#56083.
Updates #53563.
Change-Id: I346a95bdda5619632659ea854f98a9e17a6aede7
Reviewed-on: https://go-review.googlesource.com/c/go/+/456115
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
The getpwuid_r function is expected to return ERANGE on overflow.
Accept -1 on AIX as we see that in practice.
This problem was uncovered by, but not caused by, CL 455815,
which introduced a test that forced a buffer overflow.
Change-Id: I3ae94faf1257d2c73299b1478e49769bb807fc4d
Reviewed-on: https://go-review.googlesource.com/c/go/+/456075
Auto-Submit: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
CL 455716 added TestFromFS. This test was failing on Plan 9
because fromFS didn't return an empty string in case of error.
This change fixes TestFromFS by returning an empty string
in case of error.
Fixes#57142.
Change-Id: Ie50dfba5e70154d641f762fa43f1c26c3d12b6f6
Reviewed-on: https://go-review.googlesource.com/c/go/+/455835
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Damien Neil <dneil@google.com>
Run-TryBot: David du Colombier <0intro@gmail.com>
Empirically, unread PTY output may be discarded on macOS when the
child process exits.
Fixes#57141.
Tested with 'go test cmd/go -run=TestTerminalPassthrough -count=1000'
on a darwin-amd64-12_0 gomote.
Change-Id: I11508e6429c61488f30e10d9ae0cc94fdf059257
Reviewed-on: https://go-review.googlesource.com/c/go/+/455915
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
getpwnam_r and friends return the errno as the result,
not in the global errno. The code changes in CL 449316
inadvertently started using the global errno.
So if a lookup didn't fit in the first buffer size,
it was treated as not found instead of growing the buffer.
Fixes#56942.
Change-Id: Ic5904fbeb31161bccd858e5adb987e919fb3e9d9
Reviewed-on: https://go-review.googlesource.com/c/go/+/455815
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
CL 424854 changed the unified IR writer's handling of type
declarations to write the underlying type rather than the RHS type
expression's type. This in turn allows us to simplify the go/types
importer, because now there's no need to delay caling SetUnderlying.
Fixes#57015.
Change-Id: I80caa61f6cad5b7f9d829939db733a66cfca621c
Reviewed-on: https://go-review.googlesource.com/c/go/+/424876
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
[Re-land of CL 424854, which was reverted as CL 425214.]
When handling a type declaration like:
```
type B A
```
unified IR has been writing out that B's underlying type is A, rather
than the underlying type of A.
This is a bit awkward to implement and adds complexity to importers,
who need to handle resolving the underlying type themselves. But it
was necessary to handle when A was declared like:
```
//go:notinheap
type A int
```
Because we expected A's not-in-heap'ness to be conferred to B, which
required knowing that A was on the path from B to its actual
underlying type int.
However, since #46731 was accepted, we no longer need to support this
case. Instead we can write out B's actual underlying type.
One stumbling point though is the existing code for exporting
interfaces doesn't work for the underlying type of `comparable`, which
is now needed to implement `type C comparable`. As a bit of a hack, we
we instead export its underlying type as `interface{ comparable }`.
Fixes#54512.
Change-Id: I9aa087e0a277527003195ebc7f4fbba6922e788c
Reviewed-on: https://go-review.googlesource.com/c/go/+/455279
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Do not permit access to Windows reserved device names (NUL, COM1, etc.)
via os.DirFS and http.Dir filesystems.
Avoid escapes from os.DirFS(`\`) on Windows. DirFS would join the
the root to the relative path with a path separator, making
os.DirFS(`\`).Open(`/foo/bar`) open the path `\\foo\bar`, which is
a UNC name. Not only does this not open the intended file, but permits
reference to any file on the system rather than only files on the
current drive.
Make os.DirFS("") invalid, with all file access failing. Previously,
a root of "" was interpreted as "/", which is surprising and probably
unintentional.
Fixes CVE-2022-41720
Fixes#56694
Change-Id: I275b5fa391e6ad7404309ea98ccc97405942e0f0
Reviewed-on: https://team-review.git.corp.google.com/c/golang/go-private/+/1663834
Reviewed-by: Tatiana Bradley <tatianabradley@google.com>
Reviewed-by: Julie Qiu <julieqiu@google.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/455362
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Jenny Rakoczy <jenny@golang.org>
Reviewed-on: https://go-review.googlesource.com/c/go/+/455716
Reviewed-by: Jenny Rakoczy <jenny@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
CL 454475 switched from checking CompressedSize to UncompressedSize
when determining if we should consider an archive malformed because
it contains data and added a test for an example of this (a JAR). We
should also remove the hasDataDescriptor check, since that is basically
an alias for CompressedSize > 0. The test didn't catch this because we
didn't actually attempt to read from the returned reader.
Change-Id: Ibc4c1aa9c3a733f3ebf4a956d1e2f8f4900a29cd
Reviewed-on: https://go-review.googlesource.com/c/go/+/455523
Run-TryBot: Roland Shoemaker <roland@golang.org>
Reviewed-by: Julie Qiu <julieqiu@google.com>
Auto-Submit: Roland Shoemaker <roland@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Pass -Wl,--no-insert-timestamp to the external linker on windows, so
as to suppress generation of the PE file header data/time stamp. This
is in order to make it possible to get reproducible CGO builds on
windows (note that we already zero the timestamp field in question for
internal linkage).
Updates #35006.
Change-Id: I3d69cf1fd32e099bd9bb4b0431a4c5f43e4b08f3
Reviewed-on: https://go-review.googlesource.com/c/go/+/455535
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Provide file/line numbers for errors when we have them.
Make the assembler error text closer to the equivalent errors from the compiler.
Abort further processing when we come across errors.
Fixes#53994
Change-Id: I4d6a037d6d713c1329923fce4c1189b5609f3660
Reviewed-on: https://go-review.googlesource.com/c/go/+/455276
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@google.com>
Auto-Submit: Keith Randall <khr@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Put comments about what operations do per block of related opcodes
instead of on each line. This is less repetitive and lets us be a bit
more verbose in our descriptions.
Doesn't change the generated code at all.
Change-Id: I98fbd4029df6537b10aac2113a00df121d0fca1b
Reviewed-on: https://go-review.googlesource.com/c/go/+/433736
Auto-Submit: Keith Randall <khr@google.com>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
In the profiler, when unwinding the stack, we have special
handling for VDSO calls. Currently, the special handling is only
used when the normal unwinding fails. If the signal lands in the
function that makes the VDSO call (e.g. nanotime1) and after the
stack switch, the normal unwinding doesn't fail but gets a stack
trace with exactly one frame (the nanotime1 frame). The stack
trace stops because of the stack switch. This 1-frame stack trace
is not as helpful. Instead, if vdsoSP is set, we know we are in
VDSO call or right before or after it, so use vdsoPC and vdsoSP
for unwinding. Do the same for libcall.
Also remove _TraceTrap for VDSO unwinding, as vdsoPC and vdsoSP
correspond to a call, not an interrupted instruction.
Fixes#56574.
Change-Id: I799aa7644d0c1e2715ab038a9eef49481dd3a7f5
Reviewed-on: https://go-review.googlesource.com/c/go/+/455166
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Without it, at least on ARM64 with older BFD linker, it will
include the file of the object file (which is of a temporary path)
as a debug symbol into the binary, causing the build to be
nondeterministic. Adding a .file directive makes it to create a
STT_FILE symbol with deterministic input, and prevent the linker
creating one using the temporary object file path.
Fixes#57035.
Change-Id: I3ab716b240f60f7a891af2f7e10b467df67d1f31
Reviewed-on: https://go-review.googlesource.com/c/go/+/454838
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
This was a regression test added for a 'git' command line
used for build stamping. Unfortunately, 'gpg' has proved to
be extremely fragile:
* In recent versions, it appears to always require 'gpg-agent' to be
installed for anything involving secret keys, but for some reason is
not normally marked as requiring gpg-agent in Debian's package
manager.
* It tries to create a Unix domain socket in a subdirectory of $TMPDIR
without checking the path length, which fails when $TMPDIR is too
long to fit in the 'sun_path' field of a sockaddr_un struct (which
typically tops out somewhere between 92 and 108 bytes).
We could theoretically address those by artificially reducing the
script's TMPDIR length and checking for gpg-agent in addition to gpg,
but arguably those should both be fixed upstream instead. On balance,
the incremental value that this test provides does not seem worth the
complexity of dealing with such a fragile third-party tool.
Updates #50675.
Updates #48802.
Fixes#57034.
Change-Id: Ia3288c2f84f8db86ddfa139b4d1c0112d67079ef
Reviewed-on: https://go-review.googlesource.com/c/go/+/454502
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
(Until it can be made hermetic.)
The gopkg.in service has had a lot of flakiness lately. Go users in
general are isolated from that flakiness by the Go module mirror
(proxy.golang.org), but this test intentionally bypasses the module
mirror because the mirror itself uses cmd/go to download the module.
In the long term, we can redirect the gopkg.in URL to the local
(in-process) vcweb server added for #27494.
In the meantime, let's skip the test to reduce the impact of upstream
outages.
For #54503.
Change-Id: Icf3de7ca416db548e53864a71776fe22b444fcea
Reviewed-on: https://go-review.googlesource.com/c/go/+/454503
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Historically, on Windows filepath.Join("c:", elt) does not insert
a path separator between "c:" and elt, but preserves leading slashes
in elt. Restore this behavior, which was inadvertently changed by
CL 444280.
Fixes#56988
Change-Id: Id728bf311f4093264f8c067d8b801ea9ebef5b5f
Reviewed-on: https://go-review.googlesource.com/c/go/+/453497
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
gcimporter.TestImportTypeparamTests still needs to create full export
data because it loads lots of source files from GOROOT/test that
expect to be able to import arbitrary subsets of the standard library,
so we now skip it in short mode.
On a clean build cache, this reduces
'go test -short cmd/compile/internal/importer go/internal/gcimporter'
on my machine from 21–28s per test to <6s per test.
Updates #56967.
Updates #47257.
Change-Id: I8fd80293ab135e0d2d213529b74e0ca6429cdfc7
Reviewed-on: https://go-review.googlesource.com/c/go/+/454498
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
We're unlikely to get this package out of the door all that soon. For
now add a note that SetFallbackRoots will be most commonly used with
an TBA package, and link the tracking issue.
We could also just remove the "It will most commonly be used ..."
sentence.
Change-Id: Ie96134d757f5b4c69f1878d53c92b5ed602671e4
Reviewed-on: https://go-review.googlesource.com/c/go/+/454056
Reviewed-by: Julie Qiu <julieqiu@google.com>
Run-TryBot: Roland Shoemaker <roland@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Roland Shoemaker <roland@golang.org>
The darwin arm64 port was added in Go 1.16 and is a first-class port,
so it should be tracked by cmd/api. This CL does that, backfilling
API files as needed.
It also removes a spurious cgo.Incomplete API feature.
Change-Id: Idd995677915e81bf1c2e09be65b31e084b75f668
Reviewed-on: https://go-review.googlesource.com/c/go/+/453260
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Deprecating an API creates notices that go out to potentially
millions of Go developers encouraging them to update their code.
The choice to deprecate an API is as important as the choice to
add a new API. We should track those and make them explicit.
This will also ensure that deprecations go through proposal review.
Change-Id: Ide9f60c32e5a88fb133e0dfedd984b8b0f70f510
Reviewed-on: https://go-review.googlesource.com/c/go/+/453259
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Previously, we were using internal/goroot.PkgfileMap to locate
cached export data. However, PkgfileMap regenerates export data
for the entire standard library, whereas gcimporter may only need
a single package.
Under the new approach, we load the export data (still using
'go list -export') for each GOROOT package individually, avoiding work
to rebuild export data for packages that are not needed.
This is a tradeoff: if most packages in GOROOT are actually needed, we
end up making many more calls to 'go list', and may build packages
sequentially instead of in parallel (but with lower latency to start
using the export data from the earlier packages).
On my workstation, starting from a clean cache for each run,
this reduces the wall time of
'go test go/internal/gcimporter -run=TestImportedTypes'
from 22s real time (2m10s user time) to 6s real (27s user),
and only increases 'go test go/internal/gcimporter' from
28s real (2m16s user) to 30s real (2m19s user).
Updates #56967.
Updates #47257.
Change-Id: I22556acdd9b1acc56533ed4c2728ea29b585c073
Reviewed-on: https://go-review.googlesource.com/c/go/+/454497
Reviewed-by: Michael Matloob <matloob@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
pkg/obj will be empty, so just remove it.
pkg/$GOOS_$GOARCH will be empty unless the user has specified
GODEBUG=installgoroot=all, so check if it's empty, and if so, delete
it.
Also remove xreaddirfiles, which is unused.
Also remove the copy of pkg/$GOOS_$GOARCH in the cmd/go test
TestNewReleaseRebuildsStalePackagesInGOPATH. The directory is empty so
copying it has no effect.
For #47257
Change-Id: Ief90b882d157bd16078cd5d2b83a915bfc831f9a
Reviewed-on: https://go-review.googlesource.com/c/go/+/453496
Run-TryBot: Michael Matloob <matloob@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
The test attempted to find all stdlib packages by scanning
pkg/$GOOS_$GOARCH for .a files and then tried to import all of them.
Now that .a files are no longer being placed there, the test is a
noop. Fix this by using go list std (and filtering out testonly
packages) and trying to import all of those to recreate what the test
intended to do.
This also removes a dependency on the pkg/$GOOS_$GOARCH directory
which will stop being produced by dist in CL 453496.
For #47257
Change-Id: I7c1944a89db9da9269def3d64a11408a60d73d46
Reviewed-on: https://go-review.googlesource.com/c/go/+/453858
Reviewed-by: Michael Matloob <matloob@golang.org>
Run-TryBot: Michael Matloob <matloob@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Ordinary interface types now satisfy comparable constraints. This
is a fully backward-compatible change: it simply permits additional
code to be valid that wasn't valid before.
This change makes the new comparable semantics the default behavior,
depending on the Go -lang version.
It also renames the flag types2.Config.AltComparableSemantics to
types2.Config.OldComparableSemantics and inverts its meaning
(or types.Config.oldComparableSemantics respectively).
Add new predicate Satisfies (matching the predicate Implements but
for constraint satisfaction), per the proposal description.
Adjust some existing tests by setting -oldComparableSemantics
and add some new tests that verify version-dependent behavior.
The compiler flag -oldcomparable may be used to temporarily
switch back to the Go 1.18/1.19 behavior should this change
cause problems, or to identify that a problem is unrelated
to this change. The flag will be removed for Go 1.21.
For #52509.
For #56548.
For #57011.
Change-Id: I8b3b3d9d492fc24b0693567055f0053ccb5aeb42
Reviewed-on: https://go-review.googlesource.com/c/go/+/454575
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
This refines the error output that was previously adjusted in CL 437298.
Longer term, we should consider unraveling the call chains involving
formatOutput to avoid passing so many parameters through so many
different formatting functions.
Updates #25842.
Change-Id: I3b9d03bf5968902d8ccc4841ab4dbe114a2239e4
Reviewed-on: https://go-review.googlesource.com/c/go/+/451218
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
In CL 449955 we made reading of directories with associated file data
an error, since it is a "must not" in the zip specification. It turns
out that a number of implementations make the mistake of not setting
the correct compression method on directories (in particular the Java
jar tool does this when storing the META-INF directory). If the
compression method used is not 0 (stored) then the compressed size of
the directory can be > 0, despite the uncompressed size still being 0.
Since this mistake is not uncommon, we are forced to tolerate it. We
still fail if the recorded uncompressed size is > 0, which should be
a significantly harder mistake to make.
Change-Id: Ia732b10787f26ab937ac9cf9869ac3042efb8118
Reviewed-on: https://go-review.googlesource.com/c/go/+/454475
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Roland Shoemaker <roland@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Roland Shoemaker <roland@golang.org>
CL 353849 removed validation of signature parameters being passed to
Verify which led to two distinct problems. If passed a R or S == 0,
encodeSignature would panic since it expects them to be non-zero.
encodeSignature would also normalize (i.e. make non-negative) parameters
by zero padding them, which would result in a signature being passed to
VerifyASN1 which did not match the input signature, resulting in success
in cases where it should've failed. This change re-adds the verification
that 0 < r,s < N before calling ecnodeSignature.
This was caught because tink runs the wycheproof ECDSA vectors against
Verify, where we only run the vectors against VerifyASN1. We should be
doing both.
Change-Id: I1dcf41626b4df2b43296e8b878dc607ff316a892
Reviewed-on: https://go-review.googlesource.com/c/go/+/453675
Auto-Submit: Roland Shoemaker <roland@golang.org>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Roland Shoemaker <roland@golang.org>
Ordinary interface types now satisfy comparable constraints.
This change makes the new comparable semantics the default behavior,
depending on the Go -lang version.
It also renames the flag types2.Config.AltComparableSemantics to
types2.Config.OldComparableSemantics and inverts its meaning
(or types.Config.oldComparableSemantics respectively).
Adjust some existing tests by setting -oldComparableSemantics
and add some new tests that verify version-dependent behavior.
The compiler flag -oldcomparable may be used to temporarily
switch back to the Go 1.18/1.19 behavior should this change
cause problems, or to identify that a problem is unrelated
to this change. The flag will be removed for Go 1.21.
For #52509.
For #56548.
Change-Id: Ic2b22db9433a8dd81dc1ed9d30835f0395fb7205
Reviewed-on: https://go-review.googlesource.com/c/go/+/453978
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Robert Griesemer <gri@google.com>
TestLookupDotsWithRemoteSource and TestLookupGoogleSRV
were disabled because they look up the no-longer-present
SRV record for _xmpp-server._tcp.google.com.
Change the tests to look for _ldap._tcp.google.com and
reenable them.
Fixes#56708.
Change-Id: I26475fa3ff6fc008048a4e5f24f0e96ee12f655c
Reviewed-on: https://go-review.googlesource.com/c/go/+/453861
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Read the full (empty) response body before closing it,
to avoid cancelling the request while the server handler
is still running.
Wrap the ResponseWriter before calling NewResponseController:
This test is intended to verify that wrapping the controller
works properly, but neglected to actually wrap the controller.
Fixes#56961.
Change-Id: I00269f897448ab34676338707b7a04d19ff17963
Reviewed-on: https://go-review.googlesource.com/c/go/+/453860
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
if q != nil {
p = &q.f
}
Which gets rewritten to a conditional move:
tmp := &q.f
p = Select q!=nil, tmp, p
Unfortunately, we can't compute &q.f before we've checked if q is nil,
because if it is nil, &q.f is an invalid pointer (if f's offset is
nonzero but small).
Normally this is not a problem because the tmp variable above
immediately dies, and is thus not live across any safepoint. However,
if later there is another &q.f computation, those two computations are
CSEd, causing tmp to be used at both use points. That will extend
tmp's lifetime, possibly across a call.
Fixes#56990
Change-Id: I3ea31be93feae04fbe3304cb11323194c5df3879
Reviewed-on: https://go-review.googlesource.com/c/go/+/454155
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Document that a slice can be converted to either an array or a pointer
to an array of a matching underlying array type. This was documented in
the "Conversions from slice to array or array pointer" subsection, but
not in the list of conversion rules.
Updates #46505.
Change-Id: I16a89a63ef23c33580129952415e977a8f334009
Reviewed-on: https://go-review.googlesource.com/c/go/+/452936
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Tim King <taking@google.com>
I misunderstood CL 420774. We didn't remove GO_LDSO, just that
make.bash no longer tries to set it automatically. If GO_LDSO is
explicitly set at make.bash, it is still used as the default
dynamic interpreter.
For #54202.
Change-Id: Ided775438b8e4b87a6acd9bc87657657dbd3d91c
Reviewed-on: https://go-review.googlesource.com/c/go/+/453601
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
A couple of the windows runtime tests were being gated by "if
testenv.Builder() == ..." guards that referred to builders that have
long since been obsoleted (e.g. "windows-amd64-gce"). Use a more
generic guard instead, checking for windows-<goarch> prefix.
Change-Id: Ibdb9ce2b0cfe10bba986bd210a5b8ce5c1b1d675
Reviewed-on: https://go-review.googlesource.com/c/go/+/453035
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
No one ever runs 'go tool api', because the invocation
has gotten unwieldy enough that it's not practical.
And we don't support it as a standalone tool for other
packages - it's not even in the distribution.
Making it an ordinary package test lets us invoke it
more easily from cmd/dist (as go test cmd/api -check)
and avoids the increasingly baroque code in run.go to
build a command line.
Left in cmd/api even though it's no longer a command
because (1) it uses a package from cmd/vendor and
(2) it uses internal/testenv. Otherwise it could be misc/api.
Fixes#56845.
Change-Id: I00a13d9c19b1e259fa0e6bb93d1a4dca25f0e8c1
Reviewed-on: https://go-review.googlesource.com/c/go/+/453258
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
All mips variant perform syscalls similarly. R2 (v0) holds r1 and R3
(v1) holds r2 of a syscall. The latter is only used by 2-ret syscalls.
A 1-ret syscall would not touch R3 but keeps it as is, making r2 be a
random value. Always reset it to 0 before SYSCALL to fix the issue.
Fixes#56426
Change-Id: Ie49965c0c3c224c4a895703ac659205cd040ff56
Reviewed-on: https://go-review.googlesource.com/c/go/+/452975
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Meng Zhuo <mzh@golangcn.org>
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Meng Zhuo <mzh@golangcn.org>
These should be deprecated, but per go.dev/wiki/Deprecated,
that should only happen two releases after the replacement is
available (so Go 1.22).
The deprecation of this package was part of the discussion
of proposal #52221. All that remains is waiting for the new
package to be widely available.
Change-Id: I580a4af6514eb77d7ec31b443d07259a4a2cf030
Reviewed-on: https://go-review.googlesource.com/c/go/+/453256
Reviewed-by: Roland Shoemaker <roland@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
This was disabled in CL 452676 out of an abundance of caution,
but further analysis has shown that the failures were not being
caused by this optimization. Instead the sequence of commits was:
CL 450136 cmd/compile: handle simple inlined calls in staticinit
...
CL 449937 archive/tar, archive/zip: return ErrInsecurePath for unsafe paths
...
CL 451555 cmd/compile: fix static init for inlined calls
The failures in question became compile failures in the first CL
and started building again after the last CL.
But in the interim the code had been broken by the middle CL.
CL 451555 was just the first time that the tests could run and fail.
For #30820.
Change-Id: I65064032355b56fdb43d9731be2f9f32ef6ee600
Reviewed-on: https://go-review.googlesource.com/c/go/+/452817
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
There has been no proposal discussion about adding these notices.
Also, even if we did decide to add them, then since their replacements
are only appearing in Go 1.20, go.dev/wiki/Deprecation says that we
should wait until Go 1.22 to add the deprecation notice.
Filed #56906 for the proposal discussion.
Fixes#56905.
Change-Id: If86cce65aa00b4b62b2b18e82503431dcbdbcfed
Reviewed-on: https://go-review.googlesource.com/c/go/+/452761
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
CL 450136 fixed a different copy of this error but missed this one.
With the compiler fix from CL 451555 rolled back to produce the error,
this is the text before this CL:
b.go:9:15: internal compiler error: 'init': Value live at entry. It shouldn't be. func init, node a.i, value nil
And this CL changes it to:
b.go:9:15: internal compiler error: 'init': value a.i (nil) incorrectly live at entry
matching the same change in the earlier CL.
Change-Id: I33e6b91477e1a213a6918c3ebdea81273be7d235
Reviewed-on: https://go-review.googlesource.com/c/go/+/452816
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
If needed, the built-in function append allocates a new underlying
array. While we (probably) don't want to specify exactly how much
is allocated (the prose is deliberately vague), if there's more
space allocated than needed (cap > len after allocation), that
extra space is zeroed. Use an explicit link to the section on
Allocation which explicitly states that newly allocated memory
is zeroed.
Fixes#56684.
Change-Id: I9805d37c263b87860ea703ad143f738a0846247e
Reviewed-on: https://go-review.googlesource.com/c/go/+/452619
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
At parse time we don't know if a[i] is an index expression or a
type (or function) instantiation. Because instantiations accept
a list of type arguments, and argument lists permit a trailing
comma, a[i,] is either an instantiation or index expression.
Document that a trailing comma is permitted in the syntax for
index expressions.
For comparison, the same problem arises with conversions which
cannot be distinguished from function calls at parse time. The
spec also permits a trailing comma for conversions T(x,). The
grammar adjustment is the same (see line 5239).
Fixes#55007.
Change-Id: Ib9101efe52031589eb95a428cc6dff940d939f9e
Reviewed-on: https://go-review.googlesource.com/c/go/+/452618
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Update net/http to enable tests that pass with the latest update
to the vendored x/net.
Update a few tests:
Windows apparently doesn't guarantee that time.Since(time.Now())
is >=0, so to set a definitely-expired write deadline, use a time
firmly in the past rather than now.
Put a backoff loop on TestServerReadTimeout to avoid failures
when the timeout expires mid-TLS-handshake. (The TLS handshake
timeout is set to min(ReadTimeout, WriteTimeout, ReadHeaderTimeout);
there's no way to set a long TLS handshake timeout and a short
read timeout.)
Don't close the http.Server in TestServerWriteTimeout while the
handler may still be executing, since this can result in us
getting the wrong error.
Change the GOOS=js fake net implementation to properly return
ErrDeadlineExceeded when a read/write deadline is exceeded,
rather than EAGAIN.
For #49837
For #54136
Change-Id: Id8a4ff6ac58336ff212dda3c8799b320cd6b9c19
Reviewed-on: https://go-review.googlesource.com/c/go/+/449935
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This reverts CL 450375.
Reason for revert: This change causes test failures (and possibly other
problems) for users depending on the existing validation behavior.
Rolling back the change for now to give us more time to consider its
impact. This landed late in the cycle and isn't urgent; it can wait
for 1.21 if we do want to make the change.
Fixes#56884
For #56732
Change-Id: I082023c67f1bbb933a617453ab92b67abba876ef
Reviewed-on: https://go-review.googlesource.com/c/go/+/452795
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Revise the code added in CL 452763 that skips some of the -race tests
on older windows builders. The old-style skip was doing a log.Printf,
which wound up being interpreted in "-list" mode. Fix is to pass in a
special rtPreFunc when registering the test (thanks Austin for the
fix suggestion).
Updates #56904.
Change-Id: Ia6ea31fb7f011b539173f47357ab3bf7389f256d
Reviewed-on: https://go-review.googlesource.com/c/go/+/452769
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Than McIntosh <thanm@google.com>
This change is being made late in the release cycle.
Disable it by default. Insecure path checks may be enabled by setting
GODEBUG=tarinsecurepath=0 or GODEBUG=zipinsecurepath=0.
We can enable this by default in Go 1.21 after publicizing the change
more broadly and giving users a chance to adapt to the change.
For #55356.
Change-Id: I549298b3c85d6c8c7fd607c41de1073083f79b1d
Reviewed-on: https://go-review.googlesource.com/c/go/+/452616
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Damien Neil <dneil@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
The Go 1.20 code freeze has recently started. This is a time to update
all golang.org/x/... module versions that contribute packages to the
std and cmd modules in the standard library to latest master versions.
This CL updates them with x/build/cmd/updatestd.
For #36905.
Change-Id: Ie0ec91daeb848f00f64686003012297161ad02fa
Reviewed-on: https://go-review.googlesource.com/c/go/+/452766
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
This was an oversight from CL 452457 that I noticed while
investigating #56889.
This change essentially undoes CL 335409, which is no longer needed
after CL 450739 because we no longer attempt to use cgo by default
when no C compiler is present.
Updates #47257.
Updates #40042.
Updates #47215.
Change-Id: I29c7ce777a9ec7ba5820dc1d836b12a61b86bc37
Reviewed-on: https://go-review.googlesource.com/c/go/+/452677
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
The gcc toolchain on the aix-ppc64 builder apparently does not achieve
reproducible builds for packages that use cgo, which causes the
binaries in cmd that use package "net" (cmd/go, cmd/pprof, and
cmd/trace) to appear stale whenever the Go build cache is cleared.
For now, we work around the staleness by rebuilding std and simply not
checking whether cmd is stale.
For #56896.
Updates #47257.
Change-Id: I15f86e72dee53904b881710d5d5d613872361510
Reviewed-on: https://go-review.googlesource.com/c/go/+/452680
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Ayappan Perumal <ayappanec@gmail.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
If the build cache is cleaned (using 'go clean -cache' or similar), or
if a toolchain is freshly installed without warming the cache, the
build cache might not contain the dependencies needed to verify that
cmd/addr2line is not stale. In that case, the test should refill the
cache instead of failing.
Fixes#56889.
Change-Id: Ic6cf13b92bafa9c795e50eb0e4e1a9ae00ee8538
Reviewed-on: https://go-review.googlesource.com/c/go/+/452458
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
Address the release notes TODO regarding the loopclosure analyzer,
documenting the new warning for parallel subtests.
In doing so, choose a structure for the vet section, opting for h4
headings. In recent years, we have used either h4 headings or simple
paragraphs to document vet changes. This year, I thought it worthwhile
to put the timeformat and loopclosure changes into separate sections.
Also document the improvements to reference capture detection introduced
in CL 452615.
Change-Id: I05886f7025d66bb7f2609f787f69d1a769ca6f5e
Reviewed-on: https://go-review.googlesource.com/c/go/+/450735
Run-TryBot: Robert Findley <rfindley@google.com>
Reviewed-by: Tim King <taking@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This CL adds -d=inlstaticinit to control whether static initialization
of inlined function calls (added in CL 450136) is allowed.
We've needed to fix it once already (CL 451555) and Google-internal
testing is hitting additional failure cases, so putting this
optimization behind a feature flag seems appropriate regardless.
Also, while we diagnose and fix the remaining cases, this CL also
disables the optimization to avoid miscompilations.
Updates #56894.
Change-Id: If52a358ad1e9d6aad1c74fac5a81ff9cfa5a3793
Reviewed-on: https://go-review.googlesource.com/c/go/+/452676
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
If GOMAXPROCS=1, force serial execution, which is better for
debugging build problems and also minimizes footprint, if that
happens to matter.
This wasn't good when the bootstrap was 1.4 because there
default GOMAXPROCS=1, but that is no longer the bootstrap
version.
Change-Id: I637e25c8acb4758795fceef63921eda359a7be29
Reviewed-on: https://go-review.googlesource.com/c/go/+/452556
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: David Chase <drchase@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
This CL changes cmd/compile to reject anonymous interface cycles like:
type I interface { m() interface { I } }
We don't anticipate any users to be affected by this change in
practice. Nonetheless, this CL also adds a `-d=interfacecycles`
compiler flag to suppress the error. And assuming no issue reports
from users, we'll move the check into go/types and types2 instead.
Updates #56103.
Change-Id: I1f1dce2d7aa19fb388312cc020e99cc354afddcb
Reviewed-on: https://go-review.googlesource.com/c/go/+/445598
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
Modify skip rule for TestVectoredHandlerExceptionInNonGoThread to
trigger on both the base builder (windows-amd64-2012) and the newcc
canary builder (windows-amd64-2012-newcc).
Updates #49681.
Change-Id: I58109fc2e861b943cb66be0feec348671be84ab3
Reviewed-on: https://go-review.googlesource.com/c/go/+/452436
Run-TryBot: Than McIntosh <thanm@google.com>
Auto-Submit: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
This CL adds windows/arm64 to the list of ports that supports PIE
build mode. It is probably an oversight that this port is not marked
as pie-capable because windows/arm64 only supports PIE build mode.
Fixes#56872
Change-Id: I2bdd3ac207280f47ddcf8c2582f13025dafb9278
Reviewed-on: https://go-review.googlesource.com/c/go/+/452415
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This will let us reuse it in crypto/ecdsa for the NIST scalar fields.
The main change in API is around encoding and decoding. The SetBytes +
ExpandFor sequence was hacky: SetBytes could produce a bigger size than
the modulus if leading zeroes in the top byte overflowed the limb
boundary, so ExpandFor had to check for and tolerate that. Also, the
caller was responsible for checking that the overflow was actually all
zeroes (which we weren't doing, exposing a crasher in decryption and
signature verification) and then for checking that the result was less
than the modulus. Instead, make SetBytes take a modulus and return an
error if the value overflows. Same with Bytes: we were always allocating
based on Size before FillBytes anyway, so now Bytes takes a modulus.
Finally, SetBig was almost only used for moduli, so replaced
NewModulusFromNat and SetBig with NewModulusFromBig.
Moved the constant-time bitLen to math/big.Int.BitLen. It's slower, but
BitLen is primarily used in cryptographic code, so it's safer this way.
Change-Id: Ibaf7f36d80695578cb80484167d82ce1aa83832f
Reviewed-on: https://go-review.googlesource.com/c/go/+/450055
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Several operations emulate instructions available on power9. Use
the GOPPC64_power9 macro provided by the compiler to select the
native instructions if the minimum cpu requirements are met.
Likewise rework the LXSDX_BE to simplify usage when overriding
it. It is only used in one place.
All three configurations are tested via CI.
On POWER9:
pkg:crypto/cipher goos:linux goarch:ppc64le
AESCBCEncrypt1K 949MB/s ± 0% 957MB/s ± 0% +0.83%
AESCBCDecrypt1K 1.82GB/s ± 0% 1.99GB/s ± 0% +8.93%
pkg:crypto/aes goos:linux goarch:ppc64le
Encrypt 1.01GB/s ± 0% 1.05GB/s ± 0% +4.36%
Decrypt 987MB/s ± 0% 1024MB/s ± 0% +3.77%
Change-Id: I56d0eb845647dd3c43bcad71eb281b499e1d1789
Reviewed-on: https://go-review.googlesource.com/c/go/+/449116
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Auto-Submit: Paul Murphy <murp@ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Joedian Reid <joedian@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
This patch reworks the handling of DLL import symbols in the PE host
object loader to ensure that the Go linker can deal with them properly
during internal linking.
Prior to this point the strategy was to immediately treat an import
symbol reference of the form "__imp__XXX" as if it were a reference to
the corresponding DYNIMPORT symbol XXX, except for certain special
cases. This worked for the most part, but ran into problems in
situations where the target ("XXX") wasn't a previously created
DYNIMPORT symbol (and when these problems happened, the root cause was
not always easy to see).
The new strategy is to not do any renaming or forwarding immediately,
but to delay handling until host object loading is complete. At that
point we make a scan through the newly introduced text+data sections
looking at the relocations that target import symbols, forwarding
the references to the corresponding DYNIMPORT sym where appropriate
and where there are direct refs to the DYNIMPORT syms, tagging them
for stub generation later on.
Updates #35006.
Updates #53540.
Change-Id: I2d42b39141ae150a9f82ecc334001749ae8a3b4a
Reviewed-on: https://go-review.googlesource.com/c/go/+/451738
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
Adapt the testcshared tests to handle the case where the path output
by invoking
gcc -print-prog-name=dlltool
is a path lacking the final ".exe" suffix (this seems to be what clang
is doing); tack it on before using if this is the case.
Updates #35006.
Updates #53540.
Change-Id: I04fb7b9fc90677880b1ced4a4ad2a8867a3f5f86
Reviewed-on: https://go-review.googlesource.com/c/go/+/451816
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Add a new debugging flag "-capturehostobjs" that instructs the linker
to capture copies of all object files loaded in during the host object
loading portion of CGO internal linking. The intent is to make it
easier to analyze the objects after the fact (as opposed to having to
dig around inside archives, which can be a "find needle in haystack"
exercise).
Change-Id: I7023a5b72b1b899ea9b3bd6501f069d1f21bbaf0
Reviewed-on: https://go-review.googlesource.com/c/go/+/451737
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
When ctxt.Debugvlog > 1, produce additional trace output to describe
which object files are being pulled out of host archive libraries and
why they were pulled (e.g. which symbol had a reference to something
in a library). Intended to make it easier to debug problems with cgo
internal linking.
Change-Id: Icd64aff244b9145162a00cb51642ef32f26adfba
Reviewed-on: https://go-review.googlesource.com/c/go/+/451736
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This is faster than the current code because computing RR involves
one more shiftIn and using it involves an extra multiplication, but each
exponentiation was doing montgomeryRepresentation twice, once for x and
once for 1, and now they share the RR precomputation.
More importantly, it allows precomputing the value and attaching it to
the private key in a future CL.
name old time/op new time/op delta
DecryptPKCS1v15/2048-8 1.46ms ± 0% 1.40ms ± 7% -3.69% (p=0.003 n=10+9)
DecryptPKCS1v15/3072-8 4.23ms ± 0% 4.13ms ± 4% -2.36% (p=0.004 n=9+9)
DecryptPKCS1v15/4096-8 9.42ms ± 0% 9.08ms ± 3% -3.69% (p=0.000 n=9+10)
EncryptPKCS1v15/2048-8 221µs ± 0% 137µs ± 1% -37.91% (p=0.000 n=9+10)
DecryptOAEP/2048-8 1.46ms ± 0% 1.39ms ± 1% -4.97% (p=0.000 n=9+10)
EncryptOAEP/2048-8 221µs ± 0% 138µs ± 0% -37.71% (p=0.000 n=8+10)
SignPKCS1v15/2048-8 1.68ms ± 0% 1.53ms ± 1% -8.85% (p=0.000 n=9+10)
VerifyPKCS1v15/2048-8 220µs ± 0% 137µs ± 1% -37.84% (p=0.000 n=9+10)
SignPSS/2048-8 1.68ms ± 0% 1.52ms ± 1% -9.16% (p=0.000 n=8+8)
VerifyPSS/2048-8 234µs ±12% 138µs ± 1% -40.87% (p=0.000 n=10+9)
Change-Id: I6c650bad9019765d793fd37a529ca186cf1eeef7
Reviewed-on: https://go-review.googlesource.com/c/go/+/445019
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
I have never encountered multi-prime RSA in the wild. A GitHub-wide
search reveals exactly two explicit uses of it (and a couple of tools
that leave the number configurable but defaulting to two).
https://github.com/decred/tumblebit/blob/31898baea/puzzle/puzzlekey.go#L38https://github.com/carl-mastrangelo/pixur/blob/95d4a4208/tools/genkeys/genkeys.go#L13
Multi-prime RSA has a slight performance advantage, but has limited
compatibility and the number of primes must be chosen carefully based on
the key size to avoid security issues. It also requires a completely
separate and rarely used private key operation code path, which if buggy
or incorrect would leak the private key.
Mark it as deprecated, and remove the dedicated CRT optimization,
falling back instead to the slower but safer non-CRT fallback.
Change-Id: Iba95edc044fcf9b37bc1f4bb59c6ea273975837f
Reviewed-on: https://go-review.googlesource.com/c/go/+/445017
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
Infamously, big.Int does not provide constant-time arithmetic, making
its use in cryptographic code quite tricky. RSA uses big.Int
pervasively, in its public API, for key generation, precomputation, and
for encryption and decryption. This is a known problem. One mitigation,
blinding, is already in place during decryption. This helps mitigate the
very leaky exponentiation operation. Because big.Int is fundamentally
not constant-time, it's unfortunately difficult to guarantee that
mitigations like these are completely effective.
This patch removes the use of big.Int for encryption and decryption,
replacing it with an internal nat type instead. Signing and verification
are also affected, because they depend on encryption and decryption.
Overall, this patch degrades performance by 55% for private key
operations, and 4-5x for (much faster) public key operations.
(Signatures do both, so the slowdown is worse than decryption.)
name old time/op new time/op delta
DecryptPKCS1v15/2048-8 1.50ms ± 0% 2.34ms ± 0% +56.44% (p=0.000 n=8+10)
DecryptPKCS1v15/3072-8 4.40ms ± 0% 6.79ms ± 0% +54.33% (p=0.000 n=10+9)
DecryptPKCS1v15/4096-8 9.31ms ± 0% 15.14ms ± 0% +62.60% (p=0.000 n=10+10)
EncryptPKCS1v15/2048-8 8.16µs ± 0% 355.58µs ± 0% +4258.90% (p=0.000 n=10+9)
DecryptOAEP/2048-8 1.50ms ± 0% 2.34ms ± 0% +55.68% (p=0.000 n=10+9)
EncryptOAEP/2048-8 8.51µs ± 0% 355.95µs ± 0% +4082.75% (p=0.000 n=10+9)
SignPKCS1v15/2048-8 1.51ms ± 0% 2.69ms ± 0% +77.94% (p=0.000 n=10+10)
VerifyPKCS1v15/2048-8 7.25µs ± 0% 354.34µs ± 0% +4789.52% (p=0.000 n=9+9)
SignPSS/2048-8 1.51ms ± 0% 2.70ms ± 0% +78.80% (p=0.000 n=9+10)
VerifyPSS/2048-8 8.27µs ± 1% 355.65µs ± 0% +4199.39% (p=0.000 n=10+10)
Keep in mind that this is without any assembly at all, and that further
improvements are likely possible. I think having a review of the logic
and the cryptography would be a good idea at this stage, before we
complicate the code too much through optimization.
The bulk of the work is in nat.go. This introduces two new types: nat,
representing natural numbers, and modulus, representing moduli used in
modular arithmetic.
A nat has an "announced size", which may be larger than its "true size",
the number of bits needed to represent this number. Operations on a nat
will only ever leak its announced size, never its true size, or other
information about its value. The size of a nat is always clear based on
how its value is set. For example, x.mod(y, m) will make the announced
size of x match that of m, since x is reduced modulo m.
Operations assume that the announced size of the operands match what's
expected (with a few exceptions). For example, x.modAdd(y, m) assumes
that x and y have the same announced size as m, and that they're reduced
modulo m.
Nats are represented over unsatured bits.UintSize - 1 bit limbs. This
means that we can't reuse the assembly routines for big.Int, which use
saturated bits.UintSize limbs. The advantage of unsaturated limbs is
that it makes Montgomery multiplication faster, by needing fewer
registers in a hot loop. This makes exponentiation faster, which
consists of many Montgomery multiplications.
Moduli use nat internally. Unlike nat, the true size of a modulus always
matches its announced size. When creating a modulus, any zero padding is
removed. Moduli will also precompute constants when created, which is
another reason why having a separate type is desirable.
Updates #20654
Co-authored-by: Filippo Valsorda <filippo@golang.org>
Change-Id: I73b61f87d58ab912e80a9644e255d552cbadcced
Reviewed-on: https://go-review.googlesource.com/c/go/+/326012
Run-TryBot: Filippo Valsorda <filippo@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Joedian Reid <joedian@golang.org>
Currently, tfork on openbsd/mips64 returns the thread ID on success and
a negative error number on error. In CL#447175, newosproc was changed
to assume that a non-zero value is an error - return zero on success to
match this expectation.
Change-Id: I955efad49b149146165eba3d05fe40ba75caa098
Reviewed-on: https://go-review.googlesource.com/c/go/+/451257
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Joedian Reid <joedian@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Joel Sing <joel@sing.id.au>
This test exercises the case where a net.Conn error occurs while
writing a response body. It injects an error by setting a timeout
on the Conn. If this timeout expires before response headers are
written, the test fails. The test attempts to recover from this
failure by extending the timeout and retrying.
Set the timeout after the response headers are removed, and
remove the retry loop.
Fixes#56274.
Change-Id: I293f8bedb7b20a21d14f43ea9bb48fc56b59441c
Reviewed-on: https://go-review.googlesource.com/c/go/+/452175
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
Set a logger in newClientServerTest that directs the server
log output to the testing.T's log, so log output gets properly
associated with the test that caused it.
Change-Id: I13686ca35c3e21adae16b2fc37ce36daea3df9d5
Reviewed-on: https://go-review.googlesource.com/c/go/+/452075
Run-TryBot: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
To pick up CL 451515.
This CL also updates x/net because x/crypto's dependency was bumped
while tagging v0.3.0.
Done by
go get -d golang.org/x/crypto@2c476679df9a
go mod tidy
go mod vendor
Change-Id: I432a04586be3784b1027aa9b62d86c0df6d4a97e
Reviewed-on: https://go-review.googlesource.com/c/go/+/452096
Reviewed-by: Roland Shoemaker <roland@golang.org>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Before emitting a "wanted Foo but got Bar" message for an interface
type match failure, check that Foo and Bar are different. If they
are not, add package paths to first unexported struct field seen,
because that is the cause (a cause, there could be more than one).
Replicated in go/types.
Added tests to go/types and cmd/compile/internal/types2
Fixes#54258.
Change-Id: Ifc2b2067d62fe2138996972cdf3b6cb7ca0ed456
Reviewed-on: https://go-review.googlesource.com/c/go/+/422914
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: David Chase <drchase@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
We save non-go files in the cached srcfiles file because we want the
non-go files for vet, but we shouldn't report them in CompiledGoFiles.
Filter them out before adding them to CompiledGoFiles.
Fixes#28749
Change-Id: I889d4bbf8c4ec1348584a62ef5e4f8b3f05e97da
Reviewed-on: https://go-review.googlesource.com/c/go/+/451285
Run-TryBot: Michael Matloob <matloob@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
In the script engine, if a command does not return a Wait function and
it succeeds, we won't call checkStatus. That means that commands that
don't have a wait function, have a "!" indicating that they are
supposed to fail, and then succeed will spuriously not fail the script
engine test even they were supposed to fail but didn't.
Change-Id: Ic88c3cdd628064d48f14a8a4a2e97cded48890fb
Reviewed-on: https://go-review.googlesource.com/c/go/+/451284
Reviewed-by: Michael Matloob <matloob@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Matloob <matloob@golang.org>
Inline the only use of checkMarks which also allows to drop the
always-true report argument. This also ensures the correct line gets
reported in case of an error.
Also remove the unused markTree function and drop the unused testing.T
argument from makeTree.
Change-Id: I4033d3e5ecd929d08ce03c563aa99444e102d931
Reviewed-on: https://go-review.googlesource.com/c/go/+/451615
Reviewed-by: Joedian Reid <joedian@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
If a compatible trampoline has been inserted by a previously laid
function in the same section, and is known to be sufficiently close,
it can be reused.
When testing if the trampoline can be reused, the addend of the direct
call should be ignored. It is already encoded in the trampoline. If the
addend is non-zero, and the target sufficiently far away, and just
beyond direct call reach, this may cause the trampoline to be
incorrectly reused.
This was observed on go1.17.13 and openshift-installer commit f3c53b382
building in release mode with the following error:
github.com/aliyun/alibaba-cloud-sdk-go/services/cms.(*Client).DescribeMonitoringAgentAccessKeyWithChan.func1: direct call too far: runtime.duffzero+1f0-tramp0-1 -2000078
Fixes#56775
Change-Id: I54af957302506d4e3cd5d3121542c83fe980e912
Reviewed-on: https://go-review.googlesource.com/c/go/+/451415
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Than McIntosh <thanm@google.com>
This CL cleans up the literal pool implementation and inserts an UNDEF
instruction before the literal pool if the last instruction of the
function is not an unconditional jump instruction, RET or ERET
instruction.
Change-Id: Ifecb9e3372478362dde246c1bc9bc8d527a469d5
Reviewed-on: https://go-review.googlesource.com/c/go/+/424134
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Joedian Reid <joedian@golang.org>
Run-TryBot: Eric Fang <eric.fang@arm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This change adds a new GODEBUG flag called pagetrace that writes a
low-overhead trace of how pages of memory are managed by the Go runtime.
The page tracer is kept behind a GOEXPERIMENT flag due to a potential
security risk for setuid binaries.
Change-Id: I6f4a2447d02693c25214400846a5d2832ad6e5c0
Reviewed-on: https://go-review.googlesource.com/c/go/+/444157
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently, we judge whether we need to fix up the branch instruction
based on Optab.type_ field, but the type_ field in optab may change.
This CL marks the branch instruction in optab, and checks whether to
do fixing up according to the mark. Depending on the constant parameter
range of the branch instruction, there are two labels, BRANCH14BITS,
BRANCH19BITS. For the 26-bit branch, linker will handle it.
Besides this CL removes the unnecessary alignment of the DWORD
instruction. Because the ISA doesn't require it and no 64-bit load
assume it. The only effect is that there is some performance penalty
for loading from DWORDs if the 8-byte DWORD instruction crosses the
cache line, but this is very rare.
Change-Id: I993902b3fb5ad8e081dd6c441e86bcf581031835
Reviewed-on: https://go-review.googlesource.com/c/go/+/424135
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Eric Fang <eric.fang@arm.com>
Reviewed-by: Joedian Reid <joedian@golang.org>
are provided
Improve the accuracy of recorded types and instances for function calls,
by instantiating their signature before checking arguments if all type
arguments are provided. This avoids a problem where fully instantiated
function signatures are are not recorded as such following an error
checking their arguments.
Fixesgolang/go#51803
Change-Id: Iec4cbd219a2cd19bb1bcf2a5c4019f556e4304b1
Reviewed-on: https://go-review.googlesource.com/c/go/+/451436
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
lfstack does very unsafe things. In particular, it will not
work with nodes that live on the heap. In normal use by the runtime,
that is the case (it is only used for gc work bufs). But the lfstack
test does use heap objects. It goes through some hoops to prevent
premature deallocation, but those hoops are not enough to convince
-d=checkptr that everything is ok.
Instead, allocate the test objects outside the heap, like the runtime
does for all of its lfstack usage. Remove the lifetime workaround
from the test.
Reported in https://groups.google.com/g/golang-nuts/c/psjrUV2ZKyI
Change-Id: If611105eab6c823a4d6c105938ce145ed731781d
Reviewed-on: https://go-review.googlesource.com/c/go/+/448899
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
For a while now, we've had intermittent reports about problems with
os/exec on macOS, but no clear way to reproduce them. Recent changes
in the os/exec package test seem to have aligned the stars just right,
at least on my two x86 and ARM MacBook Pro laptops, to make the
package test hang with roughly 50% probability. When it does hang, the
stacks I see in the hung process match the ones reported for the
Go-based hangs in #33565. (They do not match the ones reported in the
so-called C reproducer in that issue, but I think that reproducer is
actually reproducing a different race, between fork and exit.)
The stacks obtained from the hung child processes are in
libSystem_atfork_child, which is supposed to reinitialize various
parts of the C library in the new process.
One common stack dies in _notify_fork_child calling _notify_globals
(inlined) calling _os_alloc_once, because _os_alloc_once detects that
the once lock is held by the parent process and then calls
_os_once_gate_corruption_abort. The allocation is setting up the
globals for the notification subsystem. See the source code at [1].
To work around this, we can allocate the globals earlier in the Go
program's lifetime, before any execs are involved, by calling any
notify routine that is exported, calls _notify_globals, and doesn't do
anything too expensive otherwise. notify_is_valid_token(0) fits the bill.
The other common stack dies in xpc_atfork_child calling
_objc_msgSend_uncached which ends up in
WAITING_FOR_ANOTHER_THREAD_TO_FINISH_CALLING_+initialize. Of course,
whatever thread the child is waiting for is in the parent process and
is not going to finish anything in the child process. There is no
public source code for these routines, so it is unclear exactly what
the problem is. However, xpc_atfork_child turns out to be exported
(for use by libSystem_atfork_child, which is in a different library,
so xpc_atfork_child is unlikely to be unexported any time soon).
It also stands to reason that since xpc_atfork_child is called at the
start of any forked child process, it can't be too harmful to call at
the start of an ordinary Go process. And whatever caches it needs for
a non-deadlocking fast path during exec empirically do get initialized
by calling it at startup.
This CL introduces a function osinit_hack, called at osinit time,
which calls notify_is_valid_token(0) and xpc_atfork_child().
Doing so makes the os/exec test pass reliably on both my laptops -
I can run it successfully hundreds of times in a row when my previous
record was twice in a row.
Fixes#33565.
Fixes#56784.
[1] https://opensource.apple.com/source/Libnotify/Libnotify-241/notify_client.c.auto.html
Change-Id: I16a14a800893c40244678203532a3e8d6214b6bd
Reviewed-on: https://go-review.googlesource.com/c/go/+/451735
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently for every CPU profile sample, we apply its weight to all
call edges of the entire call stack. Frames higher up the stack
are unlikely to be repeated calls (e.g. runtime.main calling
main.main). So adding weights to call edges higher up the stack
may be not reflecting the actual call edge weights in the program.
This CL changes it to add weights to only the edge between the
last two frames.
Without a branch profile (e.g. LBR records) this is not perfect,
but seems more reasonable.
Change-Id: I0aee75cc608a152adad41c51120b661a6c542283
Reviewed-on: https://go-review.googlesource.com/c/go/+/450915
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Currently, we use CDF to compute a weight threshold and then use
the weight threshold to determine whether a call site is hot. As
when we compute the CDF we already have a list of hot call sites
that make up the given percentage of the CDF, just use that list.
Also, when computing the CDF threshold, include the very last node
that makes it to go over the threshold. (I.e. if the CDF threshold
is 50% and one hot node takes 60% of weight, we should include that
node instead of excluding it. In practice it rarely matters,
probably only for testing and micro-benchmarks.)
Change-Id: I535ae9cd6b679609e247c3d0d9ee572c1a1187cc
Reviewed-on: https://go-review.googlesource.com/c/go/+/450737
Reviewed-by: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
For implementing interface to empty interface conversion, the compiler
generate code like:
var res *uint8
res = itab
if res != nil {
res = res.type
}
However, itab has type *uintptr, so the assignment is broken. The
problem is not shown up, until CL 450215, which call typecheck on this
broken assignment.
To fix this, just cast itab to *uint8 when doing the conversion.
Fixes#56768
Change-Id: Id42792d18e7f382578b40854d46eecd49673792c
Reviewed-on: https://go-review.googlesource.com/c/go/+/451256
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Temporary registers are sometimes needed for an architecture backend
which needs to use several machine instructions to implement a single
SSA instruction.
Mark such instructions so that regalloc can reserve the temporary register
for it. That way we don't have to reserve a fixed register like we do now.
Convert the temp-register-using instructions on amd64 to use this
new mechanism. Other archs can follow as needed.
Change-Id: I1d0c8588afdad5cd18b4398eb5a0f755be5dead7
Reviewed-on: https://go-review.googlesource.com/c/go/+/398556
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: David Chase <drchase@google.com>
CL 450136 made the compiler to be able to handle simple inlined calls in
staticinit. However, it's missed a condition when checking substituting
arg for param. If there's any non-trivial closures, it has captured one
of the param, so the substitution could not happen.
Fixes#56778
Change-Id: I427c9134e333e2f9af136c1a124da4d37d326f10
Reviewed-on: https://go-review.googlesource.com/c/go/+/451555
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: David Chase <drchase@google.com>
Generating 8192 bit keys times out on builders relatively frequently. We
just need something that isn't a boringAllowCert allowed key size so we
can test that a non-boringAllowCert signed intermediate works, so just
use 512 instead since it'll be significantly faster.
Fixes#56798
Change-Id: I416e0d8c3aa11ff44e9870755efa95c74d1013f1
Reviewed-on: https://go-review.googlesource.com/c/go/+/451656
Auto-Submit: Roland Shoemaker <roland@golang.org>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Run-TryBot: Roland Shoemaker <roland@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Will become available with Go 1.21.
Recognizing the `clear` built-in early is not causing any problems:
if existing code defines a `clear`, that will be used as before. If
code doesn't define `clear` the error message will make it clear
that with 1.21 the function will be available. It's still possible
to define a local `clear` and get rid of the error; but more likely
the name choice should be avoided going forward, so this provides a
useful early "heads-up".
For #56351.
Change-Id: I3d0fb1eb3508fbc78d7514b6238eac89610158c9
Reviewed-on: https://go-review.googlesource.com/c/go/+/448076
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
go/parser can correctly parse interfaces that instantiate and embed
generic interfaces, but not structs. This is because in the case of
structs, it does not expect RBRACK as a token trailing COMMA in the type
argument, even though it is allowed by the spec.
For example, go/parser produces an error for the type declaration below:
type A struct {
B[byte, []byte,]
}
Fixes#56748
Change-Id: Ibb2addd6cf9b381d8470a6d20eedb93f13f93cd6
Reviewed-on: https://go-review.googlesource.com/c/go/+/450175
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
When it reports a parse error, it uses the "value" variable as the
value element of the parse error. Previously, in some of the cases,
the "value" variable is always updated to the next chunk of the value
to be parsed (even if an earlier chunk is invalid). The reported
parse error is confusing in this case.
This CL addresses this issue by holding the original value, and when
it fails to parse the time, use it to create the parse error.
Fixes#56730.
Change-Id: I445b1d8a1b910208d0608b2186881746adb550e0
GitHub-Last-Rev: 67b1102b5e
GitHub-Pull-Request: golang/go#56754
Reviewed-on: https://go-review.googlesource.com/c/go/+/450936
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Joedian Reid <joedian@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Given code like
func itou(i int) uint { return uint(i) }
var x = itou(-1)
the static inliner from CL 450136 was rewriting the code to
var x = uint(-1)
which is not valid Go code. Fix this by converting the
constants appropriately during inlining.
Fixes golang.org/x/image/vector test.
Change-Id: I13448df8504c6a70525b1cdc36e2c947e22cdd33
Reviewed-on: https://go-review.googlesource.com/c/go/+/451376
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
In -race mode, the dist test command only registers the std, race,
osusergo, and amd64ios tests before returning early from
(*tester).registerTests. Prior to CL 450018, the osusergo and amd64ios
tests weren't even affected by -race mode, so it seems their inclusion
was unintentional. CL 450018 lifted the logic to run tests in race
mode, which means these tests went from running without -race to
running with -race. Unfortunately, amd64ios is not compatible with
-race, so it is now failing on the darwin-amd64-race builder.
Fix this by omitting the osusergo and amd64ios tests from -race mode,
since it seems like they were really intended to be included anyway.
This should fix the darwin-amd64-race builder.
Updates #37486.
Change-Id: I554bb60bc729dbb6f1bc926f1ea329768b0d6d81
Reviewed-on: https://go-review.googlesource.com/c/go/+/451437
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Austin Clements <austin@google.com>
testenv.HasCgo reports whether the test binary may have been built
with cgo enabled, but having been built with cgo does not necessarily
imply that the test can invoke the cgo tool itself.
This should fix a test failure on the android builders introduced in
CL 450714.
Change-Id: I2eed4098736e1cb285ca20bc248b0ab3515f0dea
Reviewed-on: https://go-review.googlesource.com/c/go/+/451221
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
We were using a RWMutex RLock around a single memory load,
which is not a good use of a RWMutex--it introduces extra work
for the RLock but contention around a single memory load is unlikely.
And, the tryUpdate method was not acquiring the mutex anyhow.
The new atomic.Pointer type is type-safe and easy to use correctly
for a simple use-case like this.
Change-Id: Ib3859c03414c44d2e897f6d15c92c8e4b5c81a11
Reviewed-on: https://go-review.googlesource.com/c/go/+/451416
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Return a distinguishable error when reading an archive file
with a path that is:
- absolute
- escapes the current directory (../a)
- on Windows, a reserved name such as NUL
Users may ignore this error and proceed if they do not need name
sanitization or intend to perform it themselves.
Fixes#25849Fixes#55356
Change-Id: Ieefa163f00384bc285ab329ea21a6561d39d8096
Reviewed-on: https://go-review.googlesource.com/c/go/+/449937
Reviewed-by: Joseph Tsai <joetsai@digital-static.net>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
Auto-Submit: Damien Neil <dneil@google.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
IsLocal reports whether a path lexically refers to a location
contained within the directory in which it is evaluated.
It identifies paths that are absolute, escape a directory
with ".." elements, and (on Windows) paths that reference
reserved device names.
For #56219.
Change-Id: I35edfa3ce77b40b8e66f1fc8e0ff73cfd06f2313
Reviewed-on: https://go-review.googlesource.com/c/go/+/449239
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Joseph Tsai <joetsai@digital-static.net>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Joedian Reid <joedian@golang.org>
Add "auto" mode for the -pgo build flag. When -pgo=auto is
specified, if there is a default.pgo file in the directory of the
main package, it will be selected and used for the build.
Currently it requires exactly one main package when -pgo=auto is
specified. (We'll support multiple main packages in the future.)
Also apply to other build-related subcommands, "go install", "go
run", "go test", and "go list".
For #55022.
Change-Id: Iab7974ab8932daf0e83506de505e044a8e412466
Reviewed-on: https://go-review.googlesource.com/c/go/+/438737
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
The 'cgo' command invoked by 'go fix' was not valid when built with
-trimpath, but the test was not failing because errors from the
command were being logged and ignored instead of causing tests to
fail. Changing the code and test not to ignore the errors revealed
that a number of existing tests were always, unconditionally
triggering cgo errors which were then ignored.
This change updates those tests to no longer produce cgo errors,
and to check their results when cgo is enabled.
For #51473.
Updates #51461.
Change-Id: Ib9d1ea93f26d30daa824d75ed634eaf530af086d
Reviewed-on: https://go-review.googlesource.com/c/go/+/450714
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
The documentation for cgo has always said:
> The cgo tool is enabled by default for native builds
> on systems where it is expected to work.
Following the spirit of that rule, this CL disables cgo by default
on systems where $CC is unset and the default C compiler
(clang or gcc) is not found in $PATH.
This CL makes builds of Go code on systems with no C compiler
installed automatically fall back to non-cgo mode.
For example, if building a Go program using package net
in a stripped down Linux container, that build will now run
with cgo disabled, instead of attempting the build with cgo enabled
and only succeeding if the right pre-compiled .a files happen to
be loaded into the container.
This CL makes it safe to drop the pre-compiled .a files
from the Go distribution. Systems that don't have a C compiler
will simply disable cgo when building new .a files for that system.
In general keeping the pre-compiled .a files working in cgo mode
on systems without C compilers has had only mixed success due
to the precise build cache. Today we have had to disable various
checks in the precise build cache so that distributed .a files look
up-to-date even if the current machine's C compiler is a different
version than the one used when packaging the distribution.
Each time we improve precision we have a decent chance of
re-invalidating the files. This CL, combined with dropping the .a files
entirely, will let us re-enable those checks and ensure that the
.a files used in a build actually match the C compiler being used.
On macOS, the distributed .a files for cgo-dependent packages
have been stale (not actually used by the go command) since the
release of Go 1.14 in February 2020, due to CL 216304 setting
a CGO_CFLAGS environment variable that won't match the default
setting on users machines. (To keep the distributed .a files working,
that CL should have instead changed the default in the go command.)
The effect is that for the past six Go releases (!!!), the go command
has been unable to build basic programs like src/net/http/triv.go
on macOS without either disabling cgo or installing Xcode's C compiler.
This CL fixes that problem by disabling cgo when there's no C compiler.
Now it will once again be possible to build basic programs with just
a Go toolchain installed.
In the past, disabling cgo on macOS would have resulted in subpar
implementations of crypto/x509, net, and os/user, but as of CL 449316
those packages have all been updated to use libc calls directly,
so they now provide the same implementation whether or not cgo is enabled.
In the past, disabling cgo on macOS would also have made the
race detector unusable, but CL 451055 makes the race detector
work on macOS even when cgo is disabled.
On Windows, none of the standard library uses cgo today, so all
the necessary .a files can be rebuilt without a C toolchain,
and there is no loss of functionality in the standard library when
cgo is disabled. After this CL, the race detector won't work on
Windows without a C toolchain installed, but that turns out to be
true already: when linking race-enabled programs, even if the Go linker
does not invoke the host linker, it still attempts to read some of the
host C toolchain's .a files to resolve undefined references.
On Unix systems, disabling cgo when a C compiler is not present
will mean that builds get the pure Go net resolver, which is used
by default even in cgo builds when /etc/resolv.conf is simple enough.
It will also mean they get the pure os/user code, which reads
/etc/passwd and /etc/group instead of using shared libraries,
and therefore it may miss out on other sources of user information
such as LDAP. The race detector also will not work without a C compiler.
This would be dire except that nearly all Unix systems have a C compiler
installed by default, and on those that don't it is trivial to add one.
In particular, the vast majority of Go developers running on Linux
and other Unix systems will already have a C compiler and will be
unaffected.
Change-Id: I491e8a022fe3a64022e9dc593850d483a0d353fa
Reviewed-on: https://go-review.googlesource.com/c/go/+/450739
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
The use of an empty import "C" to trigger cgo in runtime/race
serves two purposes:
1. Cause the runtime to use the C library to create system threads,
because the race syso implementation expects things like
thread-local storage to work correctly.
2. Derive the right set of //go:cgo_import_dynamic comments
to pass to the Go linker, so that it doesn't diagnose them as
undefined references.
On macOS, (1) is unnecessary because using the C library
(via DLL calls) is the only way the runtime ever creates threads.
We can accomplish (2) by writing those comments ourselves.
Having done that in this CL, cgo is no longer needed to run
the race detector on macOS, which means that having a
pre-compiled set of .a files is no longer necessary,
nor is having Xcode for use with cgo when rebuilding those .a files.
Change-Id: Iee24cc67900eb542141b32beaadafb2c94f5fe26
Reviewed-on: https://go-review.googlesource.com/c/go/+/451055
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Now that all uses of "go test" have been converted over to the new
abstraction, we can delete the old helpers for building "go test"
commands and simplify some code that's only used by the new
abstraction now.
For #37486.
Change-Id: I770cd457e018160d694abcc0b6ac80f7dc2e8425
Reviewed-on: https://go-review.googlesource.com/c/go/+/450020
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Austin Clements <austin@google.com>
This adds support for host tests to goTest and registerTest and
modifies all uses of registerHostTest to use goTest and registerTest.
This eliminates the last case where go test command lines are
constructed by hand. Next we'll clean up all of the infrastructure
support for that.
I traced all exec calls from cmd/dist on linux/amd64 and this makes
only no-op changes (such as re-arranging the order of flags).
Preparation for #37486.
Change-Id: Icb7ec8efdac72bdb819ae24b2f585375d9d9d5b9
Reviewed-on: https://go-review.googlesource.com/c/go/+/450019
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
This converts most of the remaining manual "go test" command line
construction in cmd/dist to use the goTest abstraction and
registerTest.
At this point, the only remaining place that directly constructs go
test command lines is runHostTest.
This fixes a bug in the "nolibgcc:os/user" test. It was clearly
supposed to pass "-run=^Test[^CS]", but the logic to override the
"-run" flag for "nolibgcc:net" caused "nolibgcc:os/user" to pass
*both* "-run=^Test[^CS]" and "-run=". This was then rewritten into
just "-run=" by flattenCmdline, which caused all os/user tests to run,
and not actually skip the expensive tests as intended. (This is a
great example of why the new abstraction is much more robust than
command line construction.)
I traced all exec calls from cmd/dist on linux/amd64 and, other than
the fix to nolibgcc:os/user, this makes only no-op changes (such as
re-arranging the order of flags).
For #37486.
Change-Id: Ie8546bacc56640ea39f2804a87795c14a3fe4c7d
Reviewed-on: https://go-review.googlesource.com/c/go/+/450018
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Currently, dist test has a single test called "cgo_test" that runs a
large number of different "go test"s.
This commit restructures cgo_test into several individual tests, each
of which runs a single "go test" that can be described by a goTest
object and registered with registerTest. Since this lets us raise the
abstraction level of constructing these tests and these tests are
mostly covering the Cartesian product of a small number of orthogonal
dimensions, we pull the common logic for constructing these tests into
a helper function.
For consistency, we now pass -tags=static to the static testtls and
nocgo tests, but this tag doesn't affect the build of these tests at
all. I traced all exec calls from cmd/dist on linux/amd64 and this is
the only non-trivial change.
For #37486.
Change-Id: I53c1efa1c38d785dc71968f05e8d7d636b553e96
Reviewed-on: https://go-review.googlesource.com/c/go/+/450017
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
The overall goal is to make registerTest the primary entry point for
adding dist tests and to convert nearly all dist tests to be
represented by a goTest, registered via registerTest. This will
centralize the logic for creating dist tests corresponding to go tool
tests.
I traced all exec calls from cmd/dist on linux/amd64 and this makes
only no-op changes (such as re-arranging the order of flags).
For #37486.
Change-Id: I4749e6f3666134d3259b54ee6055d76a4235c60c
Reviewed-on: https://go-review.googlesource.com/c/go/+/450016
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This CL rewrites everywhere in dist that manually constructs an
exec.Cmd to run "go test" to use the goTest abstraction. All remaining
invocations of "go test" after this CL construct the command line
manually, but ultimately use addCmd to execute it.
I traced all exec calls from cmd/dist on linux/amd64 and this makes
only no-op changes (such as re-arranging the order of flags).
For #37486.
Change-Id: Idc7497e39bac04def7ddaf2010881c9623e76fd4
Reviewed-on: https://go-review.googlesource.com/c/go/+/450015
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Austin Clements <austin@google.com>
This introduces an abstraction for constructing and running "go test"
commands. Currently, dist test is basically a shell script written in
Go syntax: it mostly just invokes lots of subprocesses, almost all of
which are "go test" invocations, and it constructs those command lines
directly from strings all over the place.
This CL raises the level of abstraction of invoking go test. The
current level of abstraction is not serving us very well: it's
conveniently terse, but the actual logic for constructing a command
line is typically so spread out that it's difficult to predict what
command will actually run. For example, the `gotest` function
constructs the basic command, but many tests want to override at least
some of these flags, so flattenCmdLine has logic specific to `go test`
for eliminating duplicate flags that `go test` itself would reject. At
the same time, the logic for constructing many common flags is
conditional, leading to a bevy of helpers for constructing flags like
`-short` and `-timeout` and `-run` that are scattered throughout
test.go and very easy to forget to call.
This CL centralizes and flattens all of this knowledge into a new
`goTest` type. This type gives dist a single, unified point where we
can change anything about how it invokes "go test".
There's currently some "unnecessary" abstraction in the implementation
of the goTest type to separate "build" and "run" flags. This will
become important later when we convert host tests and to do separate
build and run steps.
The following CLs will convert dist test to use this type rather than
directly constructing "go test" command lines. Finally, we'll strip
out the scattered helper logic for building command lines.
For #37486.
Change-Id: I9f1633fe6c0921696419ce8127ed2ca7b7a4e01b
Reviewed-on: https://go-review.googlesource.com/c/go/+/448802
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Programs that call Seed and then expect a specific sequence
of results from the global random source (using functions such as Int)
can be broken when a dependency changes how much it consumes
from the global random source. To avoid such breakages, programs
that need a specific result sequence should use NewRand(NewSource(seed))
to obtain a random generator that other packages cannot access.
Fixes#56319.
Change-Id: Idac33991b719d2c71f109f51dacb3467a649e01e
Reviewed-on: https://go-review.googlesource.com/c/go/+/451375
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Currently, cmd/go's testterminal18153 is implemented as a special test
that doesn't run as part of cmd/go's regular tests. Because the test
requires stdout and stderr to be a terminal, it is currently run
directly by "dist test" so it can inherit the terminal of all.bash.
This has a few problems. It's yet another special case in dist test.
dist test also has to be careful to not apply its own buffering to
this test, so it can't run in parallel and it limits dist test's own
scheduler. It doesn't run as part of regular "go test", which means it
usually only gets coverage from running all.bash. And since we have to
skip it if all.bash wasn't run at a terminal, I'm sure it often gets
skipped even when running all.bash.
Fix all of this by rewriting this test to create its own PTY and
re-exec "go test" to check that PTY passes through go test. This makes
the test self-contained, so it can be a regular cmd/go test, we can
drop it from dist test, and it's not sensitive to the environment of
all.bash.
Preparation for #37486.
Updates #18153.
Change-Id: I6493dbb0143348e299718f6e311ac8a63f5d69c5
Reviewed-on: https://go-review.googlesource.com/c/go/+/449503
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The registerTest function has a special case for commands that start
with "time", but we don't use this case anywhere. Delete this special
case and its support code.
Preparation for #37486.
Change-Id: Ica180417e7aa4e4fc260cb97467942bae972fdb6
Reviewed-on: https://go-review.googlesource.com/c/go/+/448801
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The cmd/dist cgo_test enumerates a large number of platforms in
various special cases. Some combinations are suspiciously absent. This
CL completes the combinations.
I've confirmed using trybots that the newly-enabled tests pass on
android/* (this is not surprising because the gohostos is never
"android" anyway), windows/arm64, linux/ppc64 (no cgo), linux/loong64
(except for one test, filed #56623), linux/mips*, netbsd/arm (except
for one test, filed #56629), and netbsd/arm64. The windows/arm builder
is out to lunch, so I'm assuming that works. Since netbsd/arm and
arm64 mostly passed these tests, I've also enabled them on netbsd/386
and netbsd/amd64, where they seem to work fine as well.
Preparation for #37486.
Change-Id: I04c3348e4f422d74d51e714647ca3db379e6e919
Reviewed-on: https://go-review.googlesource.com/c/go/+/448016
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Austin Clements <austin@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
This specifically doesn't add support for X25519 certificates.
Refactored parsePublicKey not to depend on the public PublicKeyAlgorithm
values, and ParseCertificate/ParseCertificateRequest to ignore keys that
don't have a PublicKeyAlgorithm even if parsePublicKey supports them.
Updates #56088
Change-Id: I2274deadfe9bb592e3547c0d4d48166de1006df0
Reviewed-on: https://go-review.googlesource.com/c/go/+/450815
Reviewed-by: Roland Shoemaker <roland@golang.org>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Joedian Reid <joedian@golang.org>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Fix noopt build break from CL 450136 by not running test.
I can't reproduce the failure locally, but it's entirely reasonable
for this test to fail when optimizations are disabled, so just don't
run it when optimizations are disabled.
Change-Id: I882760fc7373ba0449379f81d295312a6be49be1
Reviewed-on: https://go-review.googlesource.com/c/go/+/450740
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Michael Stapelberg <stapelberg@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Global variable initializers like
var myErr error = &myError{"msg"}
have been converted to statically initialized data
from the earliest days of Go: there is no init-time
execution or allocation for that line of code.
But if the expression is moved into an inlinable function,
the static initialization no longer happens.
That is, this code has always executed and allocated
at init time, even after we added inlining to the compiler,
which should in theory make this code equivalent to
the original:
func NewError(s string) error { return &myError{s} }
var myErr2 = NewError("msg")
This CL makes the static initialization rewriter understand
inlined functions consisting of a single return statement,
like in this example, so that myErr2 can be implemented as
statically initialized data too, just like myErr, with no init-time
execution or allocation.
A real example of code that benefits from this rewrite is
all globally declared errors created with errors.New, like
package io
var EOF = errors.New("EOF")
Package io no longer has to allocate and initialize EOF each
time a program starts.
Another example of code that benefits is any globally declared
godebug setting (using the API from CL 449504), like
package http
var http2server = godebug.New("http2server")
These are no longer allocated and initialized at program startup either.
The list of functions that are inlined into static initializers when
compiling std and cmd (along with how many times each occurs) is:
cmd/compile/internal/ssa.StringToAux (3)
cmd/compile/internal/walk.mkmapnames (4)
errors.New (360)
go/ast.NewIdent (1)
go/constant.MakeBool (4)
go/constant.MakeInt64 (3)
image.NewUniform (4)
image/color.ModelFunc (11)
internal/godebug.New (12)
vendor/golang.org/x/text/unicode/bidi.newBidiTrie (1)
vendor/golang.org/x/text/unicode/norm.newNfcTrie (1)
vendor/golang.org/x/text/unicode/norm.newNfkcTrie (1)
For the cmd/go binary, this CL cuts the number of init-time
allocations from about 1920 to about 1620 (a 15% reduction).
The total executable code footprint of init functions is reduced
by 24kB, from 137kB to 113kB (an 18% reduction).
The overall binary size is reduced by 45kB,
from 15.335MB to 15.290MB (a 0.3% reduction).
(The binary size savings is larger than the executable code savings
because every byte of executable code also requires corresponding
runtime tables for unwinding, source-line mapping, and so on.)
Also merge test/sinit_run.go, which had stopped testing anything
at all as of CL 161337 (Feb 2019) and initempty.go into a new test
noinit.go.
Fixes#30820.
Change-Id: I52f7275b1ac2a0a32e22c29f9095071c7b1fac20
Reviewed-on: https://go-review.googlesource.com/c/go/+/450136
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Joedian Reid <joedian@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
We already emit types for any and func(error) string in runtime.a
but unlike the other pre-emitted types, we don't then exclude them
from being emitted in other packages. Fix that.
Also add slices of non-func types that we already emit.
Saves 0.3% of .a files in std cmd deps, computed by adding sizes from:
ls -l $(go list -export -f '{{.Export}}' -deps std cmd
The effect is small and not worth doing on its own.
The real improvement is making “what to write always in runtime”
and “what not to write in other packages” more obviously aligned.
Change-Id: Ie5cb5fd7e5a3025d2776d9b4cece775fdf92d3b6
Reviewed-on: https://go-review.googlesource.com/c/go/+/450135
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
For range loops, we use a pointer to the backing store that gets
incremented on each iteration of the loop.
The problem with this scheme is that at the end of the last iteration,
we may briefly have a pointer that points past the end of the backing store
of the slice that is being iterated over. We cannot let the garbage collector
see that pointer.
To fix this problem, have the incremented pointer live briefly as
a uintptr instead of a normal pointer, so it doesn't keep anything
alive. Convert back to a normal pointer just after the loop condition
is checked, but before anything that requires a real pointer representation
(in practice, any call, which is what could cause a GC scan or stack copy).
Fixes#56699
Change-Id: Ia928d23f85a211565357603668bea4e5c534f989
Reviewed-on: https://go-review.googlesource.com/c/go/+/449995
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
On advice of the department of garbage collection, forcing a garbage
collection generally does not improve performance. However,
this-data-is-now-unreachable is a good property to be able to test,
and that requires finalizers and a forced GC. So, to save build time,
this test was removed from the compiler itself, but to verify the
property, it was added to the fma_test (and the end-to-end dependence
on the flag was tested with an inserted failure in testing the
test).
TODO: also turn on the new -d=gccheck=1 debug flag on the ssacheck
builder.
Benchmarking reveals that it is profitable to avoid this GC,
with about 1.5% reduction in both user and wall time.
(48 p) https://perf.golang.org/search?q=upload:20221103.3
(12 p) https://perf.golang.org/search?q=upload:20221103.5
Change-Id: I4c4816d619735838a32388acf0cc5eb1cd5f0db5
Reviewed-on: https://go-review.googlesource.com/c/go/+/447359
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
For #56603, CL 448275 added a _ [0]T field to atomic.Pointer,
so that different kinds of atomic.Pointer are not convertible.
Unfortunately, that breaks code like:
type List struct {
Next atomic.Pointer[List]
}
which should be valid, just as using Next *List is valid.
Instead, we get:
./atomic_test.go:2533:6: invalid recursive type List
./atomic_test.go:2533:6: List refers to
./atomic_test.go:2534:13: "sync/atomic".Pointer refers to
./atomic_test.go:2533:6: List
Fix by using _[0]*T instead.
Change-Id: Icc4c83c691d35961d20cb14b824223d6c779ac5e
Reviewed-on: https://go-review.googlesource.com/c/go/+/450655
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
We already had some tests for special cases such as PSS with 513 bit
keys. The upcoming backend rewrite also happened to crash at 63 and 504
bits for different reasons. Might as well be systematic about it.
Also, make sure SignPSS returns ErrMessageTooLong like SignPKCS1v15 when
the key is too small, instead of panicking or returning an unnamed error.
-all takes a couple minutes on my M1.
Change-Id: I656239a00d0831fa7d187a6d3bb30341d41602f7
Reviewed-on: https://go-review.googlesource.com/c/go/+/443195
Reviewed-by: Roland Shoemaker <roland@golang.org>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Joedian Reid <joedian@golang.org>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Samples in the mutex profile have their count and duration scaled
according to the probability they were sampled. This is done when the
profile is actually requested. The adjustment is done using to the
current configured sampling rate. However, if the sample rate is changed
after a specific sample is recorded, then the sample will be scaled
incorrectly. In particular, if the sampling rate is changed to 0, all of
the samples in the encoded profile will have 0 count and duration. This
means the profile will be "empty", even if it should have had samples.
This CL scales the samples in the profile when they are recorded, rather
than when the profile is requested. This matches what is currently done
for the block profile.
With this change, neither the block profile nor mutex profile are scaled
when they are encoded, so the logic for scaling the samples can be
removed.
Change-Id: If228cf39284385aa8fb9a2d62492d839e02f027f
Reviewed-on: https://go-review.googlesource.com/c/go/+/443056
Auto-Submit: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Joedian Reid <joedian@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
When the GC is scanning some memory (possibly conservatively),
finding a pointer, while concurrently another goroutine is
allocating an object at the same address as the found pointer, the
GC may see the pointer before the object and/or the heap bits are
initialized. This may cause the GC to see bad pointers and
possibly crash.
To prevent this, we make it that the scanner can only see the
object as allocated after the object and the heap bits are
initialized. Currently the allocator uses freeindex to find the
next available slot, and that code is coupled with updating the
free index to a new slot past it. The scanner also uses the
freeindex to determine if an object is allocated. This is somewhat
racy. This CL makes the scanner use a different field, which is
only updated after the object initialization (and a memory
barrier).
Fixes#54596.
Change-Id: I2a57a226369926e7192c253dd0d21d3faf22297c
Reviewed-on: https://go-review.googlesource.com/c/go/+/449017
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Use the same dnsConfig throughout a DNS lookup operation.
Before this CL it was possible to decide to re-read a
modified resolv.conf file during the DNS lookup,
which could lead to inconsistencies between the lookup order
and the name server list.
Change-Id: I0689749272b8263268d00b9a9cb4458cd68b23eb
GitHub-Last-Rev: 64810a22bc
GitHub-Pull-Request: golang/go#56690
Reviewed-on: https://go-review.googlesource.com/c/go/+/449337
Reviewed-by: Damien Neil <dneil@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
As noticed in the review of the CRL RawIssuer updates (https://go-review.googlesource.com/c/go/+/418834), the Raw field on the internal type certificate of crypto/x509 is unused and could be removed.
From looking at encoding/asn1's implementation, it appears this field would be set on unmarshal but not during marshaling. However, we unmarshal into the x509.Certificate class directly, avoiding this internal class entirely.
Change-Id: I1ab592eb939b6fe701206ba77b6727763deaeaf0
GitHub-Last-Rev: 5272e0d369
GitHub-Pull-Request: golang/go#56524
Reviewed-on: https://go-review.googlesource.com/c/go/+/447215
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Joedian Reid <joedian@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Roland Shoemaker <roland@golang.org>
Auto-Submit: Roland Shoemaker <roland@golang.org>
This CL redesign how we get the TLS pointer on windows/amd64.
We were previously reading it from the [TEB] arbitrary data slot,
located at 0x28(GS), which can only hold 1 TLS pointer.
With this CL, we will read the TLS pointer from the TEB TLS slot array,
located at 0x1480(GS). The TLS slot array can hold multiple
TLS pointers, up to 64, so multiple Go runtimes running on the
same thread can coexists with different TLS.
Each new TLS slot has to be allocated via [TlsAlloc],
which returns the slot index. This index can then be used to get the
slot offset from GS with the following formula: 0x1480 + index*8
The slot index is fixed per Go runtime, so we can store it
in runtime.tls_g and use it latter on to read/update the TLS pointer.
Loading the TLS pointer requires the following asm instructions:
MOVQ runtime.tls_g, AX
MOVQ AX(GS), AX
Notice that this approach is also implemented on windows/arm64.
[TEB]: https://en.wikipedia.org/wiki/Win32_Thread_Information_Block
[TlsAlloc]: https://learn.microsoft.com/en-us/windows/win32/api/processthreadsapi/nf-processthreadsapi-tlsalloc
Updates #22192
Change-Id: Idea7119fd76a3cd083979a4d57ed64b552fa101b
Reviewed-on: https://go-review.googlesource.com/c/go/+/431775
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Quim Muntal <quimmuntal@gmail.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Alex Brainman <alex.brainman@gmail.com>
Updates syscall.Open to support opening directories via CreateFileW.
CreateFileW handles are more versatile than FindFirstFile handles.
They can be used in Win32 APIs like GetFileInformationByHandle and
SetFilePointerEx, which are needed by some Go APIs.
Fixes#52747Fixes#36019
Change-Id: I26a00cef9844fb4abeeb18d2f9d854162a146651
Reviewed-on: https://go-review.googlesource.com/c/go/+/405275
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Patrik Nyblom <pnyb@google.com>
Reviewed-by: Alex Brainman <alex.brainman@gmail.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Quim Muntal <quimmuntal@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
CL 448897 changed os.Stat to request GENERIC_READ access when using
CreateFile to examine a file. This is unnecessary; access flags of 0
will permit examining file metadata even if the file isn't readable.
Revert to the old behavior here.
For #56217
Change-Id: I09220b3bbee304bd89f4a94ec9b0af42042b7773
Reviewed-on: https://go-review.googlesource.com/c/go/+/450296
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Run-TryBot: Quim Muntal <quimmuntal@gmail.com>
This loads the keys once per call, not once per block. This
has the effect of unrolling the inner loop too. This allows
decryption to scale better with available hardware.
Noteably, encryption serializes crypto ops, thus no
performance improvement is seen, but neither is it reduced.
Care is also taken to explicitly clear keys from registers
as was done implicitly in the prior version.
Also, fix a couple of typos from copying the asm used to
load ESPERM.
Performance delta on POWER9:
name old time/op new time/op delta
AESCBCEncrypt1K 1.10µs ± 0% 1.10µs ± 0% +0.55%
AESCBCDecrypt1K 793ns ± 0% 415ns ± 0% -47.70%
Change-Id: I52ca939fefa1d776a390a0869e7f4564058942b3
Reviewed-on: https://go-review.googlesource.com/c/go/+/441816
Run-TryBot: Paul Murphy <murp@ibm.com>
Reviewed-by: Joedian Reid <joedian@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
We have been expanding our use of GODEBUG for compatibility,
and the current implementation forces a tradeoff between
freshness and efficiency. It parses the environment variable
in full each time it is called, which is expensive. But if clients
cache the result, they won't respond to run-time GODEBUG
changes, as happened with x509sha1 (#56436).
This CL changes the GODEBUG API to provide efficient,
up-to-date results. Instead of a single Get function,
New returns a *godebug.Setting that itself has a Get method.
Clients can save the result of New, which is no more expensive
than errors.New, in a global variable, and then call that
variable's Get method to get the value. Get costs only two
atomic loads in the case where the variable hasn't changed
since the last call.
Unfortunately, these changes do require importing sync
from godebug, which will mean that sync itself will never
be able to use a GODEBUG setting. That doesn't seem like
such a hardship. If it was really necessary, the runtime could
pass a setting to package sync itself at startup, with the
caveat that that setting, like the ones used by runtime itself,
would not respond to run-time GODEBUG changes.
Change-Id: I99a3acfa24fb2a692610af26a5d14bbc62c966ac
Reviewed-on: https://go-review.googlesource.com/c/go/+/449504
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Under the race detector, checkptr flags uses of unsafe.Slice that
result in slices that straddle multiple Go allocations.
Avoid that scenario by calling existing runtime code.
This fixes a failure on the darwin-.*-race builders introduced in
CL 446178.
Change-Id: I6e0fdb37e3c3f38d97939a8799bb4d10f519c5b9
Reviewed-on: https://go-review.googlesource.com/c/go/+/449936
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@golang.org>
When the GC is scanning some memory (possibly conservatively),
finding a pointer, while concurrently another goroutine is
allocating an object at the same address as the found pointer, the
GC may see the pointer before the object and/or the heap bits are
initialized. This may cause the GC to see bad pointers and
possibly crash.
To prevent this, we make it that the scanner can only see the
object as allocated after the object and the heap bits are
initialized. As the scanner uses the freeindex to determine if an
object is allocated, we delay the increment of freeindex after the
initialization.
As currently in some code path finding the next free index and
updating the free index to a new slot past it is coupled, this
needs a small refactoring. In the new code mspan.nextFreeIndex
return the next free index but not update it (although allocCache
is updated). mallocgc will update it at a later time.
Fixes#54596.
Change-Id: I6dd5ccf743f2d2c46a1ed67c6a8237fe09a71260
Reviewed-on: https://go-review.googlesource.com/c/go/+/427619
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
With net converted to libc, os/user is the last remaining
cgo code in the standard libary on macOS.
Convert it to libc too.
Now only plugin remains as a cgo-using package on macOS.
Change-Id: Ibb518b5c62ef9ec1e6ab6191f4b576f7c5a4501c
Reviewed-on: https://go-review.googlesource.com/c/go/+/449316
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Adjust PGO inlining default parameters to 99% CDF threshold and
2000 budget. Benchmark results (mostly from Sweet) show that this
set of parameters performs reasonably well, with a few percent
speedup at the cost of a few percent binary size increase.
Also rename the debug flags to start with "pgo", to make it clear
that they are related to PGO.
Change-Id: I0749249b1298d1dc55a28993c37b3185f9d7639d
Reviewed-on: https://go-review.googlesource.com/c/go/+/449477
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This copies the logic we use in runtime/cgo, when calling pthread_create,
into runtime proper, when calling newosproc.
We only do this in newosproc, not newosproc0, because in newosproc0 we
need a nosplit function literal, and we need to pass arguments to it through
newosproc, which is a pain. Also newosproc0 is only called at process
startup, when thread creation is less likely to fail anyhow.
Fixes#49438
Change-Id: Ia26813952fdbae8aaad5904c9102269900a07ba9
Reviewed-on: https://go-review.googlesource.com/c/go/+/447175
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
This package is only used by tests anyway, but might as well
remove the cgo use on macOS so that it doesn't show up as
a cgo user, as part of our overall strategy to remove cgo use
in the standard library on macOS.
Change-Id: I5a1a39ed56373385f9d43a5e17098035dc1a451a
Reviewed-on: https://go-review.googlesource.com/c/go/+/449315
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
We had a special case to zero out the Target field for package
"unsafe", which is not imported from a normal object file.
As of CL 449376 that special case has been folded into go/build's
logic for setting the PkgObj field, so the special case in
cmd/go/internal/load has become redundant.
(Noticed while investigating CL 449376.)
Updates #47257.
Updates #56687.
Change-Id: I1668123aa6230097aa75e55380d3e2c7937c4b64
Reviewed-on: https://go-review.googlesource.com/c/go/+/449515
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
CL 107196 introduced a default TCP keepalive interval for Dialer and TCPListener (used by both ListenConfig and ListenTCP). Leaving DialTCP out was likely an oversight.
DialTCP's documentation says it "acts like Dial". Therefore it's natural to also expect DialTCP to enable TCP keepalive by default.
This commit addresses this disparity by moving the enablement logic down to the newTCPConn function, which is used by both dialer and listener.
Fixes#49345
Change-Id: I99c08b161c468ed0b993d1dbd2bd0d7e803f3826
GitHub-Last-Rev: 5c2f1cb0fb
GitHub-Pull-Request: golang/go#56565
Reviewed-on: https://go-review.googlesource.com/c/go/+/447917
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
In some places of the HTTP transport errors were constructed that
wrapped other errors without providing the ability to call
`errors.Unwrap` on them to get the underlying error.
These places have been fixed to use `%w` when using `fmt.Errorf`
or to implement `Unwrap() error`.
Fixes#56435
Change-Id: Ieed3359281574485c8d0b18298e25e5f1e14555c
GitHub-Last-Rev: 504efbc507
GitHub-Pull-Request: golang/go#56451
Reviewed-on: https://go-review.googlesource.com/c/go/+/445775
Reviewed-by: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Damien Neil <dneil@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
This was the last piece of ref10 code, including the infamous "Christmas
tree" in scMulAdd, that approximately all Ed25519 implementations
inherited. Replace the whole scalar field implementation with a
fiat-crypto generated one, like those in crypto/internal/nistec/fiat.
The only complexity is the wide reduction (both for the 64-byte one and
for the clamped input). For that we do a limbed reduction suggested by
Frank Denis.
Some minor housekeeping and test changes from filippo.io/edwards25519
are included, as part of syncing with downstream.
Ignoring the autogenerated file, the diff is
268 insertions(+), 893 deletions(-)
George Tankersley signed the Individual CLA and authorized me to submit
this change on his behalf at the time he contributed it to
filippo.io/edwards25519.
Co-authored-by: George Tankersley <george.tankersley@gmail.com>
Change-Id: I4084b4d3813f36e16b3d8839df75da1b4fd7846b
Reviewed-on: https://go-review.googlesource.com/c/go/+/420454
Run-TryBot: Filippo Valsorda <filippo@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
The ResponseController type provides a discoverable interface
to optional methods implemented by ResponseWriters.
c := http.NewResponseController(w)
c.Flush()
vs.
if f, ok := w.(http.Flusher); ok {
f.Flush()
}
Add the ability to control per-request read and write deadlines
via the ResponseController SetReadDeadline and SetWriteDeadline
methods.
For #54136
Change-Id: I3f97de60d4c9ff150cda559ef86c6620eee665d2
Reviewed-on: https://go-review.googlesource.com/c/go/+/436890
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
Package "builtin" is not a real, importable package; it exists only
for documentation. Package "unsafe" is not compiled into an object
file from its source code; instead, imports of "unsafe" are handled
specially by the compiler.
(In Go 1.19.3, package "unsafe" did not have an install target, while
package "builtin" did but that target was never written.)
Fixes#56687.
Updates #47257.
Change-Id: I1d1e90ff9e1629b80e0df93e1f7e17242c8dab69
Reviewed-on: https://go-review.googlesource.com/c/go/+/449376
Reviewed-by: Michael Matloob <matloob@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
As of CL 448803, packages in GOROOT only have install targets when
they have cgo source files. When cgo is not enabled, that condition
is necessarily false, and no install target will exist.
For #47257.
Change-Id: I653a9c5f89d18a5841810f3de8d490bd7cb7e922
Reviewed-on: https://go-review.googlesource.com/c/go/+/449375
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Modified the fmahash gc debug flag to use this, and modified the
test to check for a hash match that includes inlining. Also
made the test non-short to ensure portability.
Note fma.go has been enhanced into an FMA test that requires
two separate FMAs in order to "fail"; if either one is 2-rounding,
then it "passes". (It neither passes nor fails here; its role
is to demonstrate that the FMAs are correctly reported; the
enhanced failure mode was discovered while testing the search
tool.)
Change-Id: I4e328e3654f442d498eac982135420abb59c5434
Reviewed-on: https://go-review.googlesource.com/c/go/+/448358
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: David Chase <drchase@google.com>
The new "InvalidSyntaxTree" node in the error code declaration
inadvertently incremented the value of iota by 1. Fix this by moving it
to its own declaration.
Change-Id: I34b33a8caddbbb9e41f431321ec0e5863dc15055
Reviewed-on: https://go-review.googlesource.com/c/go/+/449475
Run-TryBot: Robert Findley <rfindley@google.com>
Auto-Submit: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
When writing markdown for godoc, we can reference a method M of
a type T as [T.M]. This doesn't currently work for methods on generic
types because the declaration of the type parameter gets in the way.
(You'd have to write [T[P].M] and that doesn't parse, and even if it
did you'd have to spell "P" correctly.)
Get rid of the type parameter when building the list of Funcs so
[T.M] works in godoc if T is generic.
Change-Id: I8ef5264124a944967df3ce20ddd40a2447ff4187
Reviewed-on: https://go-review.googlesource.com/c/go/+/449236
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Since the first implementation of os/user, it's called C malloc
to allocate memory for buffers. However, the buffers are just
used for temporary storage, and we can just a []byte instead.
To make this work without causing cgo pointer errors, we move
the pwd and grp structs into C stack memory, and just return them.
It's OK to store a Go pointer on the C stack temporarily.
Change-Id: I9f8ffb6e51df1e585276c259fe99359d7835df87
Reviewed-on: https://go-review.googlesource.com/c/go/+/449335
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
When you run 'go env' or any command that needs to consider
what the default gcc flags are (such as 'go list net' or
'go list <any package with net as a dependency>'),
the go command runs gcc (or clang) a few times to see what
flags are available.
These runs can be quite expensive on some systems, particularly
Macs that seem to need to occasionally cache something before
gcc/clang can execute quickly.
To fix this, cache the derived information about gcc under a cache
key derived from the size and modification time of the compiler binary.
This is not foolproof, but it should be good enough.
% go install cmd/go
% time go env >/dev/null
0.22 real 0.01 user 0.01 sys
% time go env >/dev/null
0.03 real 0.01 user 0.01 sys
%
Fixes#50982.
Change-Id: Iba7955dd10f610f2793e1accbd2d06922f928faa
Reviewed-on: https://go-review.googlesource.com/c/go/+/392454
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
When people add new API and get an all.bash failure,
they often don't know about the API checker at all.
Point to the README in the failure message, to try to
help them find what they need to know.
Change-Id: I6b148ec414d212033b371357a5e8c6ab79bb50a4
Reviewed-on: https://go-review.googlesource.com/c/go/+/449015
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
crypto/rsa assumes RSA OAEP uses the same hash to be used for both the label
and the mask generation function. However, implementations in other languages,
such as Java and Python, allow these parameters to be specified independently.
This change allows the MGF hash to be specified independently for decrypt
operations in order to allow decrypting ciphertexts generated in other
environments.
Fixes: #19974
Change-Id: If453d628f0da354ceb3b52863f30087471670f7b
Reviewed-on: https://go-review.googlesource.com/c/go/+/418874
Auto-Submit: Andrew Bonventre <andybons@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Andrew Bonventre <andybons@golang.org>
Run-TryBot: Andrew Bonventre <andybons@golang.org>
This reverts commit d154ef60a0.
This change made IsAbs return true for certain reserved filenames,
but does not consistently detect reserved names. For example,
"./COM1", "//./COM1", and (on some Windows versions) "COM1.txt"
all refer to the COM1 device, but IsAbs detects none of them.
Since NUL is not an absolute path, do not attempt to detect it
or other device paths in IsAbs. See #56217 for more discussion
of IsAbs and device paths.
For #56217.
Change-Id: If4bf81c7e1a2e8842206c7c5268555102140dae8
Reviewed-on: https://go-review.googlesource.com/c/go/+/448898
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Quim Muntal <quimmuntal@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Rob Pike <r@golang.org>
Some file operations, notably Stat and Mkdir, special cased their
behavior when operating on a file named "NUL" (case-insensitive).
This check failed to account for the many other names of the NUL
device, as well as other non-NUL device files: "./nul", "//./nul",
"nul.txt" (on some Windows versions), "con", etc.
Remove the special case.
os.Mkdir("NUL") now returns no error. This is consonant with the
operating system's behavior: CreateDirectory("NUL") succeeds, as
does "MKDIR NUL" on the command line.
os.Stat("NUL") now follows the existing path for FILE_TYPE_CHAR devices,
returning a FileInfo which correctly reports the file as being a
character device.
os.Stat and os.File.Stat have common elements of their logic unified.
For #24482.
For #24556.
For #56217.
Change-Id: I7e70f45901127c9961166dd6dbfe0c4a10b4ab64
Reviewed-on: https://go-review.googlesource.com/c/go/+/448897
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Quim Muntal <quimmuntal@gmail.com>
Previously, the volumeNameLen function checked for UNC paths starting
with two slashes, a non-'.' character, and another slash. This misses
volume names such as "\\.\C:\".
The previous check for volume names rejects paths beginning
with "\\.". This is incorrect, because while these names are not
UNC paths, "\\.\C:\" is a DOS device path prefix indicating the
C: device. It also misses UNC path prefixes in the form
"\\.\UNC\server\share\".
The previous check for UNC paths also rejects any path with an
empty or missing host or share component. This leads to a number
of possibly-incorrect behaviors, such as Clean(`\\a`) returning `\a`.
Converting the semantically-significant `\\` prefix to a single `\`
seems wrong.
Consistently treat paths beginning with two separators as having
a volume prefix.
Update VolumeName to detect DOS device paths (`\\.\` or `\\?\`),
DOS device paths linking to UNC paths (`\\.\UNC\Server\Share`
or `\\?\UNC\Server\Share`), and UNC paths (`\\Server\Share\`).
Clean(`\\a`) = `\\a`
Join(`\\`, `a`, `b`) = `\\a\b`
In addition, normalize path separators in VolumeName for consistency
with other functions which Clean their result.
Fixes#56336
Change-Id: Id01c33029585bfffc313dcf0ad42ff6ac7ce42fd
Reviewed-on: https://go-review.googlesource.com/c/go/+/444280
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Quim Muntal <quimmuntal@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
The "go test" and "go build" commands have special-case behavior when
passed "-o /dev/null". These checks are case-sensitive and assume that
os.DevNull is an absolute path. Windows filesystems are case-insensitive
and os.DevNull is NUL, which is not an absolute path.
CL 145220 changed filepath.IsAbs to report "NUL" as absolute to work
around this issue; that change is being rolled back and a better fix here
is to compare the value of -o against os.DevNull before attempting to
merge it with a base path. Make that fix.
On Windows, accept any capitilization of "NUL" as the null device.
This change doesn't cover every possible name for the null device, such
as "-o //./NUL", but this test is for efficiency rather than correctness.
Accepting just the most common name is fine.
For #56217.
Change-Id: I60b59b671789fc456074d3c8bc755a74ea8d5765
Reviewed-on: https://go-review.googlesource.com/c/go/+/449117
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Instead of hardcoding the set of five packages that depend on cgo to
decide whether a package should have an install target, make the
decision based on whether the package has any CgoFiles. This means that
in nocgo configurations, there will be no installed packages, and that
if an GOOS/GOARCH combination doesn't have cgo files we don't
unnecessarily install a .a.
Because the determination of whether a file is a CgoFile is made later
in the Import functions, the choice of whether to add a PkgObj for teh
case there are CgoFiles is moved later. One concern here is that in some
cases, PkgObj may be set differently in the case of the FindOnly mode,
since the determination is moved across the boundary. We might want
to always set PkgObj after the FindOnly boundary for consistency? cmd/go
doesn't seem to use it when calling Import with FindOnly.
Also remove internal/buildinternal/needs_install.go because we will be
checking whether to install based on the number of cgo files and it
might be overkill to make the NeedsInstalledDotA function be the
equivalent of len(input) > 0.
For #47257
Change-Id: I5f7f2137dc99aaeb2e2695c14e0222093a6b2407
Reviewed-on: https://go-review.googlesource.com/c/go/+/448803
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Matloob <matloob@golang.org>
An upcoming change to the filepath package to make IsAbs("NUL")==false
on Windows will cause this test to fail, since it sets GOPATH=NUL and
GOPATH must be an absolute path.
Set GOPATH to the name of a text file instead. (The intent is that GOPATH
be set to a path that is not writable.)
For #56217.
Change-Id: I18e645fe11547d02d1a2e0e580085e6348c4009a
Reviewed-on: https://go-review.googlesource.com/c/go/+/448896
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
If fsys.Walk is called with a root directory that is a symlink, follow
the symlink when doing the walk. This allows for users setting their
current directory to a symlink to a module.
Fixes#50807
Change-Id: Ie65a7cb804b87dea632ea6c758c20adcfa62fcd4
Reviewed-on: https://go-review.googlesource.com/c/go/+/448360
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
If GOROOT is stale, test fail when commands unexpectedly write to GOROOT.
Include an message in the test failure indicating that this is a possible
and expected reason for the failure, and how to fix it.
For #48698.
Change-Id: I057c20260bab09aebf684e8f20794ab8fc0ede1e
Reviewed-on: https://go-review.googlesource.com/c/go/+/448895
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
Add a new "Action":"start" test2json event to mark the
start of the test binary execution. This adds useful information
to the JSON traces, and it also lets programs watching test
execution see the order in which the tests are being run,
because we arrange for the starts to happen sequentially.
Change-Id: I9fc865a486a55a7e9315f8686f59a2aa06455884
Reviewed-on: https://go-review.googlesource.com/c/go/+/448357
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Fixes#53658. io.Copy() uses sendfile(2) to avoid allocating extra buffers when src is a file and dst is a TCPConn. However if src returns no bytes current logic treats it as failure and falls back to copying via user space. The following is a benchmark that illustrates the bug.
Benchmark: https://go.dev/play/p/zgZwpjUatSq
Before:
BenchmarkCopy-16 541006 2137 ns/op 4077 B/op 0 allocs/op
After:
BenchmarkCopy-16 490383 2365 ns/op 174 B/op 8 allocs/op
Change-Id: I703376d53b20e080c6204a73c96867cce16b24cf
GitHub-Last-Rev: 3a50be4f16
GitHub-Pull-Request: golang/go#53659
Reviewed-on: https://go-review.googlesource.com/c/go/+/415834
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Comparing opening and closing tag is done using the prefix when available.
Documentation states that Token returns URI in the Space part of the Name.
Translation has been moved for the End tag before the namespace is removed
from the stack.
After closing a tag using a namespace, the valid namespace must be taken
from the opening tag. Tests added.
Fixes#20685
Change-Id: I4d90b19f7e21a76663f0ea1c1db6c6bf9fd2a389
Reviewed-on: https://go-review.googlesource.com/c/go/+/107255
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
The linux-amd64-wsl builder was failing because the res_nsearch
implementation was storing pointer to the res_state's own fields
in other fields in the res_state. If the res_state is Go memory, this
looks like pointers to Go pointers. Moving the res_state to C memory
avoids the problem.
The linux-amd64-wsl builder has been fixed a different way by
replacing res_nsearch with res_search on Linux, where it is thread-safe.
But other systems that still need to use res_nsearch (such as macOS)
may run into the same kind of problem, so it is probably still worth
arranging for the res_state to live entirely in C memory.
Fixes#56658 (again).
Change-Id: I58a14e72c866eaceb02ad828854a1f626b9b8e73
Reviewed-on: https://go-review.googlesource.com/c/go/+/448798
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
If an imported, non-generic function F transitively calls a generic
function G[T], we may need to call CanInline on G[T].
While here, we can also take advantage of the fact that we know G[T]
was already seen and compiled in an imported package, so we don't need
to call InlineCalls or add it to typecheck.Target.Decls. This saves us
from wasting compile time re-creating DUPOK symbols that we know
already exist in the imported package's link objects.
Fixes#56280.
Change-Id: I3336786bee01616ee9f2b18908738e4ca41c8102
Reviewed-on: https://go-review.googlesource.com/c/go/+/443535
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
On Android and Dragonfly, don't use -lresolv. It doesn't exist there.
On Linux, use res_search instead of res_nsearch.
glibc makes res_search thread-safe by having a per-thread __res.
It still also provides res_nsearch.
musl makes res_search thread-safe by ignoring __res completely.
It does not provide res_nsearch at all.
Changing to res_search on Linux will fix builds on musl-based systems
while keeping glibc-based systems working.
Fixes#56657.
Fixes#56660.
Change-Id: Id87dde6c8bbf6c0d34543c09782f3871489c8712
Reviewed-on: https://go-review.googlesource.com/c/go/+/448797
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
The current code passes a Go pointer to a NUL-terminated C string
to the C function res_nsearch (or res_search), but that function may
in turn store the pointer into the res_state, which is a violation of the
cgo pointer rules and is being detected on the linux-amd64-wsl builder.
Allocating the string in C memory is safer and should resolve
the cgo pointer check. When using libc/syscall mode, the memory
is still allocated Go-side, which could potentially be a problem
if we ever add a moving collector. For now it is OK.
Fixes#56658.
Change-Id: Ibd84a9665be16c71994ddb1eedf09d45a6553a3e
Reviewed-on: https://go-review.googlesource.com/c/go/+/448795
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Bypass: Russ Cox <rsc@golang.org>
Return an error when parsing a MIME header containing bytes in the
key or value outside the set allowed by RFC 7230.
For historical compatibility, accept spaces in keys (but do not
canonicalize the key in this case).
For #53188.
Change-Id: I195319362a2fc69c4e506644f78c5026db070379
Reviewed-on: https://go-review.googlesource.com/c/go/+/410714
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
Add a -pgo flag for "go build" (and other build actions), to
specify the file path of a profile used for PGO. Special name
"off" turns off PGO.
The given profile path is passed to the compiler.
The build cache is sensitive to the content of the given PGO
profile.
TODO: auto mode.
For #55022.
Change-Id: Ieee1b131b4c041f9502fd0a1acf112f3e44246be
Reviewed-on: https://go-review.googlesource.com/c/go/+/438736
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
This change modifies the pacer in two ways:
* It replaces the PI controller used as a smoothing function with a
simple two-cycle moving average.
* It makes the pacer use the actual GC trigger point for cons/mark (and
other) calculations instead of the precomputed one.
The second part of this change was attempted in the Go 1.19 release
cycle, but could not be done because although it resulted in a
better-behaved pacer, it exploited the PI controller's sensitivity to
history in a way that was ultimately unfavorable for most applications.
This sensitivity is complex to reason about, and forces us into choices
that don't really make sense (like using the precomputed trigger over
the actual one -- that's really a bug fix).
The net effect of this change is intended to be:
* The pacer computes a more stable estimate of the actual cons/mark
ratio, making it easier to understand.
* The pacer is much more accurate at hitting the heap goal, so the GC
respects GOGC much more often and reliably.
* The pacer forces a more stable rate of GC assists in the
steady-state overall.
See https://perf.golang.org/search?q=upload:20221106.10 for benchmark
results and #53892 for complete context. The benchmarks that regress
in memory use appear to be not worth worrying about. In all cases, it appears that the GC was triggering super early, resulting in a lack
of adherence to rule of GOGC.
The fogleman benchmarks just have a single final GC that triggers
early and also happens to be the peak heap size. The tile38
WithinCircle benchmark only has 4 GC cycles in the benchmarked region,
so it's very sensitive to pacing. In this case, the old smoothing
function is getting lucky by starting way too early, avoiding assists. Meanwhile the 2-cycle moving average is more accurate on the heap
goal, but the 1st and 2nd cycle after the initialization phase are operating on a cons/mark from the initialization phase which is much
less active, resulting in a cons/mark that's too low, causing the GC
to start too late, increasing assists, and therefore latency (but
only transiently, at this phase change). I really do think the PI
controller is just getting lucky here with a particular history,
because I've definitely observed it oscillating wildly in response to
a phase change.
This change also moves the PI controller out of mgcpacer.go, because
it's no longer used there. It now lives in mgcscavenge.go, where it's
actually used.
Fixes#53892.
Change-Id: I3f875a2e40f31f381920f91d8b090556b17a2b16
Reviewed-on: https://go-review.googlesource.com/c/go/+/417558
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Passing a profile with no sample is arguably not a user error.
Accept such a profile, and ignore it as it doesn't indicate any
optimizations. This also makes testing easier.
Change-Id: Iae49a4260e20757419643153f50d8d5d51478411
Reviewed-on: https://go-review.googlesource.com/c/go/+/448495
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Unix and Windows differ in how they handle LookupCNAME(name).
If name exists in DNS with an A or AAAA record but no CNAME,
then on all operating systems and in the pure Go resolver,
LookupCNAME returns the name associated with the A/AAAA record
(the original name).
TestLookupCNAME has been testing this, because www.google.com
has no CNAME. I think it did at one point, but not anymore, and the
tests are still passing. Also added google.com as a test, since
top-level domains are disallowed from having CNAMEs.
If name exists in DNS with a CNAME record pointing to a record that
does not exist or that has no A or AAAA record,
then Windows has always reported the CNAME value,
but Unix-based systems and the pure Go resolver have reported
a failure instead. cname-to-txt.go4.org is an test record that is
a CNAME to a non-A/AAAA target (the target only has a TXT record).
This CL changes the Unix-based systems and the pure Go resolver
to match the Windows behavior, allowing LookupCNAME to succeed
whenever a CNAME exists.
TestLookupCNAME nows tests the new behavior by looking up
cname-to-txt.go4.org (run by bradfitz).
Fixes#50101.
Change-Id: Ieff5026c8535760e6313c7a41ebd5ff24de6d9be
Reviewed-on: https://go-review.googlesource.com/c/go/+/446179
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
The interface will let us store actor-specific state in the
interface implementation instead of continuing to grow
the Action struct. In the long term we should remove fields
from the struct that are not used by all Actions.
Change-Id: I8ac89eda7a91d742cee547a1f779e9f254dfd2f0
Reviewed-on: https://go-review.googlesource.com/c/go/+/448356
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
Extend the context package to allow users to specify why a context was
canceled in the form of an error, the "cause". Users write the cause
by calling WithCancelCause to construct a derived context, then
calling cancel(cause) to cancel the context with the provided cause.
Users retrieve the cause by calling context.Cause(ctx), which returns
the cause of the first cancelation for ctx or any of its parents.
The cause is implemented as a field of cancelCtx, since only cancelCtx
can be canceled. Calling cancel copies the cause to all derived (child)
cancelCtxs. Calling Cause(ctx) finds the nearest parent cancelCtx by
looking up the context value keyed by cancelCtxKey.
API changes:
+pkg context, func Cause(Context) error
+pkg context, func WithCancelCause(Context) (Context, CancelCauseFunc)
+pkg context, type CancelCauseFunc func(error)
Fixes#26356Fixes#51365
Change-Id: I15b62bd454c014db3f4f1498b35204451509e641
Reviewed-on: https://go-review.googlesource.com/c/go/+/375977
Reviewed-by: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Sameer Ajmani <sameer@golang.org>
Auto-Submit: Sameer Ajmani <sameer@golang.org>
Resend of CL 432735 (with one additional conversion that the original CL
missed) after it broke the longtest builder on x/tools and was reverted
in CL 433478. Now that x/tools/go/ssa has support for this, the longtest
x/tools build passes as well.
Use slice-to-array conversions in AddrFromSlice and
(*Addr).UnmarshalBinary. This allows using AddrFrom16 and drop the
redundant ipv6Slice helper.
For #46505
Change-Id: I4d8084b7a97f162e4f7d685c86aac56d960ff693
Reviewed-on: https://go-review.googlesource.com/c/go/+/448396
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
This gets the Go command out of the business of thinking it understands
compiler debug flags, and allows the compiler to turn down its worker
concurrency instead of failing and forcing the user to do the very
same thing. Debug flags that are obviously safe for concurrency
(at least to me) are tagged; probably there's more.
Change-Id: I59bb19861d8a654a9cfd2364ee78c8628212f82e
Reviewed-on: https://go-review.googlesource.com/c/go/+/448359
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Alpine has a known issue where setgid clobbers the Go stack (#39857).
misc/cgo/test skips other tests that use setgid on Alpine, but not
this one. It's not clear to me why this test *used to* pass, but when
I refactored misc/cgo/test in CL 447355 it started failing.
Disable this test on Alpine, like the other setgid tests.
Change-Id: I2e646ef55e2201a4f0b377319d719a011ec847f0
Reviewed-on: https://go-review.googlesource.com/c/go/+/448355
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Support for direct libc calls was added in CL 446178 but the build
tags weren't quite activating it when cgo was not enabled. Adjust them
and add a new supporting file for darwin.
This should use the new direct libc calls with both CGO_ENABLED=0 and
CGO_ENABLED=1 when building for darwin.
Updates #12524
Change-Id: Ieee4b298dee13f389ed3a63c0a4a3a18c9180163
Reviewed-on: https://go-review.googlesource.com/c/go/+/448020
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Dan Peterson <danp@danp.net>
Auto-Submit: Ian Lance Taylor <iant@google.com>
In verifyServerCertificate parse certificates using the global
certificate cache.
This should signficiantly reduce memory usage in TLS clients which make
concurrent connections which reuse certificates (anywhere in the chain)
since there will only ever be one copy of the certificate at once.
Fixes#46035
Change-Id: Icf5153d0ea3c14a0bdc8b26c794f21153bf95f85
Reviewed-on: https://go-review.googlesource.com/c/go/+/426455
Reviewed-by: Heschi Kreinick <heschi@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Roland Shoemaker <roland@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Adds a BoringSSL CRYPTO_BUFFER_POOL style reference counted intern
table for x509.Certificates. This can be used to significantly reduce
the amount of memory used by TLS clients when certificates are reused
across connections.
Updates #46035
Change-Id: I8d7af3bc659a93c5d524990d14e5254212ae70f4
Reviewed-on: https://go-review.googlesource.com/c/go/+/426454
Run-TryBot: Roland Shoemaker <roland@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Though we split traces into 100MB chunks, currently each chunk always
includes the entire stack frame map, including frames for all events in
the trace file, even if they aren't needed by events in this chunk.
This means that if the stack frame JSON alone is >100MB then there is no
space at all for events. In that case, we'll generate splits each
containing 1 event, which is effectively useless.
Handle this more efficiently by only including stack frames referenced
by events in the chunk. Each marginal events only adds at most a few
dozen stack frames, so it should now longer be possible to only include
a tiny number of events.
Fixes#56444.
Change-Id: I58aa8f271c32678028b72d82df16e6ea762ebb39
Reviewed-on: https://go-review.googlesource.com/c/go/+/445895
Run-TryBot: Michael Pratt <mpratt@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
runtime.bgsweep contains an infinite loop. With aggressive enough
inlining, it may not perform any CALLs on a typical iteration. If
the runtime trying to preempt this goroutine, the lack of CALLs may
prevent preemption for ever occurring.
bgsweep does happen to call goschedIfBusy. Add a preempt check there to
make sure we yield eventually.
For #55022.
Change-Id: If22eb86fd6a626094b3c56dc745c8e4243b0fb40
Reviewed-on: https://go-review.googlesource.com/c/go/+/447135
Run-TryBot: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
ISEL is roughly equivalent to CMOV on PPC64. Verify ISEL generation
in all reasonable cases.
Note "ISEL test x y z" is the same as "ISEL !test y x z". test is
always one of LT (0), GT (1), EQ (2), SO (3). Sometimes x and y are
swapped if GE/LE/NE is desired.
Change-Id: Ie1bf029224064e004d855099731fe5e8d05aa990
Reviewed-on: https://go-review.googlesource.com/c/go/+/445215
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Paul Murphy <murp@ibm.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Currently locations are stored in a map and looked up by ID from
the map. The IDs are usually small sequential integers (the Go
pprof CPU profiles are so). Using a slice is more efficient (with
a fallback map to handle weirdly large IDs).
Change-Id: I9e20d5cebca3a5239636413e1bf2f0b273038031
Reviewed-on: https://go-review.googlesource.com/c/go/+/447803
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Now that most GOROOT packages do not have install targets, their
staleness depends on whether their builds have been cached. That means
that, from a clean cache, once
TestNewReleaseRebuildsStalePackagesInGOPATH builds certain targets, they
will stay non-stale for the second run of the test, becasue the will
still be in the cache. So the first run of the test from a clean cache
will pass and the second will fail. Set GOCACHE to a temporary directory
in the test to use a clean cache and avoid this.
Change-Id: I91f954138a4723d81545134441148badbfc515f5
Reviewed-on: https://go-review.googlesource.com/c/go/+/448018
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Michael Matloob <matloob@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Rather than matching calls to edges in the profile based directly on
line number in the source file, use the line offset from the start of
the function. This makes matching robust to changes in the source file
above the function containing the call.
The start line in the profile comes from Function.start_line, which is
included in Go pprof output since CL 438255.
Currently it is an error if no samples set start_line to help users
detect profiles missing this information. In the future, we should
fallback to using absolute lines, which is better than nothing.
For #55022.
Change-Id: Ie621950cfee1fef8fb200907a2a3f1ded41d04fa
Reviewed-on: https://go-review.googlesource.com/c/go/+/447315
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
When adding weights for a call stack, for recursive calls, to
avoid double counting we check if we already saw the node and the
edge. We check the node first. An edge can be repeated if the node
is repeated. Most stacks are not recursive, so check repeated edge
only conditionally.
Change-Id: I4b8f039289dcd3383ca89593d6d16d903b94c3dd
Reviewed-on: https://go-review.googlesource.com/c/go/+/447804
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
When we were first introducing module mode, CL 163418 moved many of
the tests in misc/cgo/test into their own test binary under testdata
so misc/cgo/test continued to work in both GOPATH mode and module
mode. This introduce a somewhat complicated test driver into
misc/cgo/test. Since the misc/cgo/test test had to invoke "go test" as
a subprocess, this required care to thread any build flags down into
the subprocess. The output from any failures of the sub-process was
also less than ideal.
Now that we don't have to worry about running these in GOPATH mode any
more, this CL moves all of the tests back into misc/cgo/test and drops
the test driver.
There are two slight complications:
- Test41761 was added after this split and has a C type "S" that's
also present in misc/cgo/test itself. We rename that to keep that
test working.
- TestCgo in go/internal/srcimporter now fails to import misc/cgo/test
because misc/cgo/test now contains imports of other "misc" module
packages and the importer it sets up isn't configured to allow that.
We fix this by setting up a build context that's configured for
this.
Preparation for #37486.
Change-Id: I3c4f73540e0482bbd493823cca44b0ce7fac01f3
Reviewed-on: https://go-review.googlesource.com/c/go/+/447355
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently in the pprof Graph, a Node's in/out edges are
represented as maps, keyed by the source/destination Nodes. For a
Node it usually has a very small number of edges, so linear search
would be generally faster than map operations. Use slices and
linear search instead.
Change-Id: I5ab06be0c019373cd3a845b9f3282731372e9c4c
Reviewed-on: https://go-review.googlesource.com/c/go/+/447802
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
We intentionally don't use file path so it is resilient to code
moving. OrigName and Objfile are also not used currently. Remove
them. (We can add them back if it turns out to be useful.)
Change-Id: I7975d78c874bc21475b9119301088452a4426cb9
Reviewed-on: https://go-review.googlesource.com/c/go/+/447801
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Default concurrency is now GOMAXPROCS, though this is normally
reduced to 4 by the go build commmand.
If compiler flags don't permit concurrency, silently
turn it down instead of failing the compilation.
Change-Id: Id26cc3214e0ca402d60f915e98276a58d809e097
Reviewed-on: https://go-review.googlesource.com/c/go/+/447358
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: David Chase <drchase@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Semicolon has bad copy-paste ergonomics; it requires quoting.
Slash is okay, and won't be a separator ever in debug strings
because it is already used in e.g. ssa/phase/debug=etc.
Change-Id: I493360e9282666eea1a342971a77df2ebd6c92ad
Reviewed-on: https://go-review.googlesource.com/c/go/+/447975
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: David Chase <drchase@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
With CL 447015, we identify hot callees from edge weights, but
the code only traverses edges for calls from the current package.
If the callee is in a different package, when compiling that
package, the edge was not visited, so the callee was not actually
marked inline candidate. This CL fixes it by traversing all hot
edges.
Change-Id: If668c1a16ebe34e3474376b88ab3a84be76b8562
Reviewed-on: https://go-review.googlesource.com/c/go/+/448015
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Fixes a bug that was introduced in CL 410414; in that CL, to avoid
a race condition in the initialization of the lexer, the setting
of the breakOK and continueOK options was moved to before
Tree.funcs was populated from parameters. As a result, the parser
missed the fact that 'break' and 'continue' were defined as functions.
Following CL 421883 race conditions are no longer an issue, so
the simplest fix is just to move the initialization where it was
before - in startParse, after t.funcs has been set.
Fixes#56538
Change-Id: I3b99fe9ad12255a4f6eb9a00eb3f64529ff055c0
Reviewed-on: https://go-review.googlesource.com/c/go/+/447775
Run-TryBot: Eli Bendersky <eliben@golang.org>
Auto-Submit: Eli Bendersky <eliben@golang.org>
Reviewed-by: Rob Pike <r@golang.org>
Reviewed-by: Eli Bendersky <eliben@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Plan 9 uses NUL as os.PathListSeparator, so it's almost always going
to appear in the environment variable list. Exempt GOOS=plan9 from the
check for NUL in environment variables.
For #56284.
Fixes#56544.
Change-Id: I23df233cdf20c0a9a606fd9253e15a9b5482575a
Reviewed-on: https://go-review.googlesource.com/c/go/+/447715
Reviewed-by: David du Colombier <0intro@gmail.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
There is existing code that calls flag.Lookup("test.v") and inspects
the value. That stopped working as of CL 443596. Make code like that
continue to work at least for the case where we aren't using
-test.v=test2json.
Change-Id: Idb30b149b48ee3987a201e349cf4d9bfe9ddee56
Reviewed-on: https://go-review.googlesource.com/c/go/+/447796
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Handle emitting (to ld) or resolving commonly used ELFv2 1.5
relocations. The new ISA provides PC relative addressing with
34 bit signed addresses, and many other relocations which can
replace addis + d-form type relocations with a single prefixed
instruction.
Updates #44549
Change-Id: I7d4f4314d1082daa3938f4353826739be35b0e81
Reviewed-on: https://go-review.googlesource.com/c/go/+/355149
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
Prior to Go 1.18, ineffectual //go:linkname directives (i.e.,
directives referring to an undeclared name, or to a declared type or
constant) were treated as noops. In Go 1.18, we changed this into a
compiler error to mitigate accidental misuse.
However, the x/sys repo contained ineffectual //go:linkname directives
up until go.dev/cl/274573, which has caused a lot of user confusion.
It seems a bit late to worry about now, but to at least prevent
further user pain, this CL changes the error message to only apply to
modules using "go 1.18" or newer. (The x/sys repo declared "go 1.12"
at the time go.dev/cl/274573 was submitted.)
Fixes#55889.
Change-Id: Id762fff96fd13ba0f1e696929a9e276dfcba2620
Reviewed-on: https://go-review.googlesource.com/c/go/+/447755
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
This needs to be as low as possible while not breaking priority
assumptions of other scores to correctly schedule carry chains.
Prior to the arm64 changes, it was set below ReadTuple. At the time,
this prevented the MulHiLo implementation on PPC64 from occluding
the scheduling of a full carry chain.
Memory scores can also prevent better scheduling, as can be observed
with crypto/internal/edwards25519/field.feMulGeneric.
Fixes#56497
Change-Id: Ia4b54e6dffcce584faf46b1b8d7cea18a3913887
Reviewed-on: https://go-review.googlesource.com/c/go/+/447435
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
go/build attempts to parse comments at the beginning of non-Go files
looking for //go:build or //+go build comments. Before this change,
if the beginning of the non-Go file failed to parse (perhaps because
it is in a format that isn't even meant to be built with Go code) the
file would be added to InvalidGoFiles. The comment for InvalidGoFiles
states that it contains Go files, so this is clearly incorrect
behavior.
Further, if there was a directory that only contained these unparsable
non-Go files, it would have a non-zero number of InvalidGoFiles, and
the matching code in cmd/go/internal/search/search.go in
(*Match).MatchDirs would treat it as a directory containing (invalid)
Go files and would match the directory as a Go package. This incorrect
behavior is also fixed by this CL.
Fixes#56509
Change-Id: Id0d905827c71f7927f7c2fa42b236181950af7e9
Reviewed-on: https://go-review.googlesource.com/c/go/+/447357
Run-TryBot: Michael Matloob <matloob@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
Benchmarking suggests about a 14-17% reduction in user build time,
about 3.5-7.8% reduction for wall time. This helps most builds
because small packages are common. Latest benchmarks (after the last
round of improvement):
(12 processors) https://perf.golang.org/search?q=upload:20221102.20
(GOMAXPROCS=2) https://perf.golang.org/search?q=upload:20221103.1
(48 processors) https://perf.golang.org/search?q=upload:20221102.19
(The number of compiler workers is capped at min(4, GOMAXPROCS))
An earlier, similar version of this CL at one point observed a 27%
reduction in user build time (building 40+ benchmarks, 20 times), but
the current form is judged to be the most reliable; it may be
profitable to tweak the numbers slightly later, and/or to adjust the
number of compiler workers.
We've talked about doing this in the past, the "new"(ish) metrics
package makes it a more tractable proposition.
The method here is:
1. If os.Getenv("GOGC") is empty, then increase GOGC to a large value,
calculated to grow the heap to 32 + 4 * compile_parallelism before a
GC occurs (e.g., on a >= 4 processor box, 64M). In practice,
sometimes GC occurs before that, but this still results in fewer GCs
and saved time. This is "heap goal".
2. Use a finalizer to approximately detect when GC occurs, and use
runtime metrics to track progress towards the goal heap size,
readjusting GOGC to retarget it as necessary. Reset GOGC to 100 when
the heap is "close enough" to the goal.
One feared failure mode of doing this is that the finalizer will be
slow to run and the heap will grow exceptionally large before GOGC is
reset; I monitored the heap size at reset and exit across several
boxes with a variety of processor counts and extra noise
(including several builds in parallel, including a laptop with a busy
many-tabs browser running) and overshoot effectively does not occur.
In some cases the compiler's heap grows so rapidly that estimated live
exceeds the GC goal, but this is not delayed-finalizer overshoot; the
compiler is just using that much memory. In a small number of cases
(3% of GCs in make.bash) the new goal is larger than predicted by as
much as 38%, so check for that and redo the adjustment.
I considered instead using the maximum heap size limit +
GC-detecting-finalizer + reset instead, but to me that seemed like it
might have a worse bad-case outcome; if the reset is delayed, it's
possible the GC would start running frequently, making it harder to
run the finalizer, reach 50% utilization, and the extra GCs would
lose the advantage. This might also perform badly in the case that a
rapidly growing heap outruns goal. In practice, this sort of
overshoot hasn't been observed, and a goal of 64M is small enough to
tolerate plenty of overshoot anyway.
This version of the CL includes a comment urging anyone who sees the
code and thinks it would work for them, to update a bug (to be
created if the CL is approved) with information about their
situation/experience, so that we may consider creating some more
official and reliable way of obtaining the same result.
Change-Id: I45df1c927c1a7d7503ade1abd1a3300e27516633
Reviewed-on: https://go-review.googlesource.com/c/go/+/436235
Run-TryBot: David Chase <drchase@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
The stack pointer must lie within system stack limits
when Control Flow Guard (CFG) is enabled on Windows.
This CL updates runtime.sigtramp to honor this restriction by
porting some code from the windows/arm64 version, which
already supports CFG.
Fixes#53560
Change-Id: I7f88f9ae788b2bac38aac898b2567f1bea62f8f3
Reviewed-on: https://go-review.googlesource.com/c/go/+/437559
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
This adds a -d debug flag "fmahash" for hashcode search for
floating point architecture-dependent problems. This variable has no
effect on architectures w/o fused-multiply-add.
This was rebased onto the GOSSAHASH renovation so that this could have
its own dedicated environment variable, and so that it would be
cheap (a nil check) to check it in the normal case.
Includes a basic test of the trigger plumbing.
Sample use (on arm64, ppc64le, s390x):
% GOCOMPILEDEBUG=fmahash=001110110 \
go build -o foo cmd/compile/internal/ssa/testdata/fma.go
fmahash triggered main.main:24 101111101101111001110110
GOFMAHASH triggered main.main:20 010111010000101110111011
1.0000000000000002 1.0000000000000004 -2.220446049250313e-16
exit status 1
The intended use is in conjunction with github.com/dr2chase/gossahash,
which will probably acquire a flag "-fma" to streamline its use. This
tool+use was inspired by an ad hoc use of this technique "in anger"
to debug this very problem. This is also a dry-run for using this
same technique to identify code sensitive to loop variable
lifetime/capture, should we make that change.
Example intended use, with current search tool (using old environment
variable), for a test example:
gossahash -e GOFMAHASH GOMAGIC=GOFMAHASH go run fma.go
Trying go args=[...], env=[GOFMAHASH=1 GOMAGIC=GOFMAHASH]
go failed (81 distinct triggers): exit status 1
Trying go args=[...], env=[GOFMAHASH=11 GOMAGIC=GOFMAHASH]
go failed (39 distinct triggers): exit status 1
Trying go args=[...], env=[GOFMAHASH=011 GOMAGIC=GOFMAHASH]
go failed (18 distinct triggers): exit status 1
Trying go args=[...], env=[GOFMAHASH=0011 GOMAGIC=GOFMAHASH]
Trying go args=[...], env=[GOFMAHASH=1011 GOMAGIC=GOFMAHASH]
...
Trying go args=[...], env=[GOFMAHASH=0110111011 GOMAGIC=GOFMAHASH]
Trying go args=[...], env=[GOFMAHASH=1110111011 GOMAGIC=GOFMAHASH]
go failed (2 distinct triggers): exit status 1
Trigger string is 'GOFMAHASH triggered math.qzero:427 111111101010011110111011', repeated 6 times
Trigger string is 'GOFMAHASH triggered main.main:20 010111010000101110111011', repeated 1 times
Trying go args=[...], env=[GOFMAHASH=01110111011 GOMAGIC=GOFMAHASH]
go failed (1 distinct triggers): exit status 1
Trigger string is 'GOFMAHASH triggered main.main:20 010111010000101110111011', repeated 1 times
Review GSHS_LAST_FAIL.0.log for failing run
FINISHED, suggest this command line for debugging:
GOSSAFUNC='main.main:20 010111010000101110111011' \
GOFMAHASH=01110111011 GOMAGIC=GOFMAHASH go run fma.go
Change-Id: Ifa22dd8f1c37c18fc8a4f7c396345a364bc367d5
Reviewed-on: https://go-review.googlesource.com/c/go/+/394754
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: David Chase <drchase@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Currently in PGO we use a percentage threshold to determine if a
callsite is hot. This CL uses a different method -- treating the
hottest callsites that make up cumulatively top X% of total edge
weights as hot (X=95 for now). This default might work better for
a wider range of profiles. (The absolute threshold can still be
changed by a flag.)
For #55022.
Change-Id: I7e3b6f0c3cf23f9a89dd5994c10075b498bf14ee
Reviewed-on: https://go-review.googlesource.com/c/go/+/447016
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Currently, FromDir identifies a VCS checkout directory just by checking
whether it contains a specified file. This is not enough. For example,
although there is a ".git" file (a plain file, not a directory) in a
git submodule directory, this directory is not a git repository root.
This change takes the file mode into account. As of now, the filename
and file mode for the supported VCS tools are:
- Mercurial: .hg directory
- Git: .git directory
- Bazaar: .bzr directory
- Subversion: .svn directory
- Fossil: .fslckout plain file
- Fossil: _FOSSIL_ plain file
This CL effectively reverts CL 30948 for #10322.
Fixes#53640.
Change-Id: Iea316c7e983232903bddb7e7f6dbaa55e8498685
GitHub-Last-Rev: 7a2d6ff6f9
GitHub-Pull-Request: golang/go#56296
Reviewed-on: https://go-review.googlesource.com/c/go/+/443597
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Per discussion with Roland Shoemaker, this updates
x509.CreateRevocationList to mirror the behavior of
x509.CreateCertificate, creating an internal struct for the ASN.1
encoding of the CRL. This allows us to switch the Issuer field type to
asn1.RawValue, bypassing the round-tripping issues of pkix.Name in most
scenarios.
Per linked ticket, this resolves issues where a non-Go created
certificate can be used to create CRLs which aren't correctly attested
to that certificate. In rare cases where the CRL issuer is validated
against the certificate's issuer (such as the linked JDK example), this
can result in failing to check this CRL for the candidate certificate.
Fixes#53754
Change-Id: If0adc053c081d6fb0b1ce47324c877eb2429a51f
GitHub-Last-Rev: 033115dd5e
GitHub-Pull-Request: golang/go#53985
Reviewed-on: https://go-review.googlesource.com/c/go/+/418834
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Roland Shoemaker <roland@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Roland Shoemaker <roland@golang.org>
Two edge cases that were mentioned in the docs are actually impossible:
* For NIST curves, ECDH can't fail, since the zero scalar is rejected
by NewPrivateKey, the identity point is rejected by NewPublicKey,
and NIST curves are a prime-order group.
Let's call the inputs to scalar multiplication k and P, and the
order of the group q. If k[P] is the identity, and also q[P] is the
identity by definition, then P's order is a divisor of q-k, because
k[P] + [q-k]P = q[P] = I
P's order is either 1 or q, and can only be a divisor of q-k if it's
1 (so P is the identity), or if k is zero.
* For X25519, PrivateKey.PublicKey can't return the all-zero value,
since no value is equivalent to zero after clamping.
Clamping unsets the lowest three bit, sets the second-to-highest
bit, and unsets the top bit; this means that a scalar equivalent to
zero needs to be a multiple of 8*q, and needs to be between 2**254
and 2**255-1, but 8*p > 2**255-1.
Tests for other exotic edge cases such as non-canonical point encodings,
clamping, points on the twist, and low-order components are covered by
x/crypto/wycheproof.
Change-Id: I731a878c58bd59aee5636211dc0f19ad8cfae9db
Reviewed-on: https://go-review.googlesource.com/c/go/+/425463
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
In https://build.golang.org/log/d2eb315305bf3d513c490e7f85d56e9a016aacd2,
we observe a failure in TestPipeLookPathLeak due to an additional
descriptor (7) that was open at the start of the test being closed while
the test executes.
I haven't dug much into the failure, but it seems plausible to me that the
descriptor may have been opened by libc for some reason, and may have been
closed due to some sort of idle timeout or the completion of a background
initialization routine.
Since the test is looking for a leak, and closing an existing descriptor
does not indicate a leak, let's not fail the test if an existing descriptor
is unexpectedly closed.
Updates #5071.
Change-Id: I03973ddff6592c454cfcc790d6e56accd051dd52
Reviewed-on: https://go-review.googlesource.com/c/go/+/447235
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently, with PGO, the inliner uses node weights to decide if a
function is inlineable (with a larger budget). But the actual
inlining is determined by the weight of the call edge. There is a
discrepancy that, if a callee node is hot but the call edge is not,
it would not inlined, and marking the callee inlineable would of
no use.
Instead of using two kinds of weights, we just use the edge
weights to decide inlineability. If a function is the callee of a
hot call edge, its inlineability is determined with a larger
threshold. For a function that exceeds the regular inlining budget,
it is still inlined only when the call edge is hot, as it would
exceed the regular inlining cost for non-hot call sites, even if
it is marked inlineable.
For #55022.
Change-Id: I93fa9919fc6bcbb394e6cfe54ec96a96eede08f7
Reviewed-on: https://go-review.googlesource.com/c/go/+/447015
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
CC-BY-3.0 was shiny and new back in 2009, but CC-BY-4.0 is
generally preferred now. Update our CC-BY uses to CC-BY-4.0.
Google lawyers signed off on the overall CC-BY-4.0 update
and Renee French signed off on the update of the gopher license.
See also CL 447156.
Change-Id: I3908910d6011ed733271e595f761c773351b30f7
Reviewed-on: https://go-review.googlesource.com/c/go/+/447275
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
We don't have a formatter for these files, so check here that
they are in the right form to allow 'cat next/*.txt >go1.X.txt'
at the end of each cycle.
Fix the api files that the check finds.
Change-Id: I0c5e4ab11751c7d0afce32503131d487313f41c0
Reviewed-on: https://go-review.googlesource.com/c/go/+/431335
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
net.IP states that a 16-byte slice can still be an IPv4 address.
But after netip.Addr is introduced, it requires extra care to keep
it as an IPv4 address when converting it to a netip.Addr using
netip.AddrFromSlice.
To address this issue, let's change the cgo resolver to return
4-byte net.IP for IPv4. The change will save us 12 bytes too.
Please note that the go resolver already return IPv4 as 4-byte
slice.
The test TestResolverLookupIP has been modified to cover this
behavior. So no new test is added.
Fixes#53554.
Change-Id: I0dc2a59ad785c0c67a7bc22433105529f055997f
GitHub-Last-Rev: bd7bb2f17b
GitHub-Pull-Request: golang/go#53638
Reviewed-on: https://go-review.googlesource.com/c/go/+/415580
Auto-Submit: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Damien Neil <dneil@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Randomized feature enable/disable might be something we use to
help users debug any problems with changed loop variable capture,
and there's another CL that would like to use it to help in
locating places where "fused" multiply add instructions change
program behavior.
This CL:
- adds the ability to include an integer parameter (e.g. line number)
- replumbed the environment variable into a flag to simplify go build cache management
- but added an environment variable to allow flag setting through the environment
- which adds the possibility of switching on a different variable
(if there's one built-in for variable capture, it shouldn't be GOSSAHASH)
- cleaned up the checking code
- adds tests for all the intended behavior
- removes the case for GSHS_LOGFILE; TBD whether we'll need to put that back
or if there is another way.
Change-Id: I8503e1bb3dbc4a743aea696e04411ea7ab884787
Reviewed-on: https://go-review.googlesource.com/c/go/+/443063
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
Run-TryBot: David Chase <drchase@google.com>
There does not seem to be any point to this sync.Once.
I noticed because I was surveying uses of sync.Once to
understand usage patterns. This seems to be a dreg left over
from some earlier instance of the code.
Change-Id: I99dd258d865a41d0e8f6cfa55887855e477fb9c2
Reviewed-on: https://go-review.googlesource.com/c/go/+/445755
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
It adds support for no-reload option, as specified in resolv.conf(5):
no-reload (since glibc 2.26)
Sets RES_NORELOAD in _res.options. This option
disables automatic reloading of a changed
configuration file.
Change-Id: I11182c5829434503f719ed162014f2301e3ba8d4
GitHub-Last-Rev: 7ae44be2d5
GitHub-Pull-Request: golang/go#56489
Reviewed-on: https://go-review.googlesource.com/c/go/+/446555
Reviewed-by: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
The global ListOfHotCallSites set is used to communicate between
CanInline and InlineCalls the set of call sites that InlineCalls may
increase the budget for.
CanInline clears this map on each call, thus assuming that
InlineCalls(x) is called immediately after CanInline(x). This assumption
is false, as CanInline (among other cases) is recursive (CanInline ->
hairyVisitor.doNode -> inlCallee -> CanInline).
When this assumption proves false, we will lose the opportunity to
inline hot calls.
This CL is the least invasive fix for this. ListOfHotCallSites is
actually just a subset of the candHotEdgeMap, with CallSiteInfo.Callee
cleared. candHotEdgeMap doesn't actually need to distinguish based on
Callee, so we can drop callee from candHotEdgeMap as well and just use
that directly [1].
Later CLs should do more work to remove the globals entirely.
For cmd/compile, this inceases the number of PGO inlined functions by
~50% for one set of PGO parameters. I have no evaluated performance
impact.
[1] This is something that we likely want to change in the future.
For #55022.
Change-Id: I57735958d651f6dfa9bd296499841213d20e1706
Reviewed-on: https://go-review.googlesource.com/c/go/+/446755
Auto-Submit: Michael Pratt <mpratt@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
ErrWaitDelay is not expected to occur in this test, but if it does
it indicates a failure mode very different from the “failed to start”
catchall that we log for other non-ExitError errors.
Updates #50436.
Change-Id: I3f4d87d502f772bf471fc17303d5a6b483446f8f
Reviewed-on: https://go-review.googlesource.com/c/go/+/446876
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
I noticed some test failures in the build dashboard after CL 445597
that made me realize the grace period should be based on the test
timeout, not the Context timeout: if the test itself sets a short
timeout for a command, we still want to give the test process enough
time to consume and log its output.
I also put some more thought into how one might debug a test hang, and
realized that in that case we don't want to set a WaitDelay at all:
instead, we want to leave the processes in their stuck state so that
they can be investigated with tools like `ps` and 'lsof'.
Updates #50436.
Change-Id: I65421084f44eeaaaec5dd2741cd836e9e68dd380
Reviewed-on: https://go-review.googlesource.com/c/go/+/446875
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Otherwise given a file like defs_nacl_amd64p32.go.~1~ we will add
"nacl" and "amd64p32" to AllTags. This was causing the
cmd/go/internal/modindex tests to fail on my system, since I had
an old editor backup file lying around.
Change-Id: Ib1c5d835e4871addae6dc78cee07c9839bb880e2
Reviewed-on: https://go-review.googlesource.com/c/go/+/446395
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
The go1 benchmark suite does a lot of work at package init time, which
makes it take quite a while to run even if you're not running any of
the benchmarks, or if you're only running a subset of them. This leads
to an awkward workaround in dist test to compile but not run the
package, unlike roughly all other packages. It also reduces isolation
between benchmarks by affecting the starting heap size of all
benchmarks.
Fix this by initializing all data required by a benchmark when that
benchmark runs, and keeping it local so it gets freed by the GC and
doesn't leak between benchmarks. Now, none of the benchmarks depend on
global state.
Re-initializing the data on each benchmark run does add overhead to an
actual benchmark run, as each benchmark function is called several
times with different values of b.N. A full run of all benchmarks at
the default -benchtime=1s now takes ~10% longer; higher -benchtimes
would be less. It would be quite difficult to cache this data between
invocations of the same benchmark function without leaking between
different benchmarks and affecting GC overheads, as the testing
package doesn't provide any mechanism for this.
This reduces the time to run the binary with no benchmarks from 1.5
seconds to 10 ms, and also reduces the memory required to do this from
342 MiB to 17 MiB.
To make sure data was not leaking between different benchmarks, I ran
the benchmarks with -shuffle=on. The variance remained low: mostly
under 3%. A few benchmarks had higher variance, but in all cases it
was similar to the variance between this change.
This CL naturally changes the measured performance of several of the
benchmarks because it dramatically changes the heap size and hence GC
overheads. However, going forward the benchmarks should be much better
isolated.
For #37486.
Change-Id: I252ebea703a9560706cc1990dc5ad22d1927c7a0
Reviewed-on: https://go-review.googlesource.com/c/go/+/443336
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Austin Clements <austin@google.com>
If TestArenaCollision cannot reserve the address range it expects to
reserve, it currently fails somewhat mysteriously. Detect this case
and skip the test. This could lead to test rot if we wind up always
skipping this test, but it's not clear that there's a better answer.
If the test does fail, we now also log what it thinks it reserved so
the failure message is more useful in debugging any issues.
Fixes#49415Fixes#54597
Change-Id: I05cf27258c1c0a7a3ac8d147f36bf8890820d59b
Reviewed-on: https://go-review.googlesource.com/c/go/+/446877
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Track the running tests and when they started,
so that we can report the running tests on a test timeout.
% go test -timeout=5s
panic: test timed out after 5s
running tests:
TestTCPSpuriousConnSetupCompletion (4s)
... stack traces as usual ...
% go test -run=Script -timeout=10s cmd/go
vcs-test.golang.org rerouted to http://127.0.0.1:65168https://vcs-test.golang.org rerouted to https://127.0.0.1:65169
go test proxy running at GOPROXY=http://127.0.0.1:65170/mod
panic: test timed out after 10s
running tests:
TestScript (10s)
TestScript/mod_get_patchcycle (0s)
TestScript/mod_get_prefer_incompatible (0s)
TestScript/mod_get_promote_implicit (0s)
TestScript/mod_get_pseudo (0s)
TestScript/mod_get_pseudo_other_branch (0s)
TestScript/mod_get_pseudo_prefix (0s)
TestScript/mod_get_test (0s)
TestScript/mod_get_trailing_slash (0s)
TestScript/mod_get_update_unrelated_sum (0s)
TestScript/mod_gobuild_import (0s)
TestScript/mod_gomodcache (0s)
TestScript/mod_gonoproxy (0s)
TestScript/mod_load_badchain (0s)
TestScript/mod_overlay (0s)
TestScript/test_fuzz_minimize (6s)
TestScript/test_fuzz_minimize_dirty_cov (7s)
... stack traces as usual ...
Change-Id: I3a6647c029097becc06664ebd76a2597c7ed7b8e
Reviewed-on: https://go-review.googlesource.com/c/go/+/446176
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
Use a slightly different line number pragma when emitting instrumented
code, so as to ensure that we don't get any changes in the
"-gcflags=-m" output for coverage vs non-coverage.
Fixes#56475.
Change-Id: I3079171fdf83c0434ed6ea0ce3eb2797c2280c55
Reviewed-on: https://go-review.googlesource.com/c/go/+/446259
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Change the 'PrepareForCoverageBuild' helper function to provide more
sensible defaults in the case where Go packages are listed on the
command line (e.g. cases such as "go run -cover mumble.go").
With the old implementation, if module mode was enabled, we would only
instrument packages in the main module, meaning that if you did
something like this:
$ ls go.mod
go.mod
$ GOCOVERDATA=/tmp/cov go run -cover testdata/somefile.go
$
no coverage profiles would be generated at all. This is due to the
fact that the pseudo-package created by the Go command internally when
building "somefile.go" is not considered part of the main module.
This patch moves the default to "packages explicitly mentioned on the
command line, plus packages in the main module", which will make more
sense to users passing specific packages and *.go files on the command
line. Examples:
// Here cmd/compile is part the Go standard library + commands
// (which we exclude from instrumentation by default), but since
// 'cmd/compile' is mentioned on the command line, we will instrument
// just that single package (not any of its deps).
$ cd $GOROOT/src
$ go build -o gc.exe -cover cmd/compile
$ GOCOVERDATA=/tmp/cov ./gc.exe ...
...
$
// Here we're running a Go file named on the command line, where
// the pseudo-package for the command line is not part of the
// main module, but it makes sense to instrument it anyhow.
$ cd ~/go/k8s.io/kubernetes
$ GOCOVERDATA=/tmp/cov go run -cover test/typecheck/testdata/bad/bad.go
...
$
This patch also simplifies the logic and improves flow/comments in
in the helper function PrepareForCoverageBuild.
Change-Id: Id8fc8571157dac8c09e44cc73baa05aeba1640ca
Reviewed-on: https://go-review.googlesource.com/c/go/+/445918
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
This patch fixes a typo/bug introduced in CL 441858 where when pattern
matching a coverage counter access we were looking at an assingment
node instead of the assignment LHS, and fixes a similar problem in
atomic counter update pattern matching introduced in CL 444835. In
both of these cases the bug was not caught because the test intended
to lock down the behavior was written incorrectly (wasn't
instrumenting what the test author thought it was instrumenting,
ouch).
Change-Id: I6e6ac3beacf12ef1a817de5527340b639f0bb044
Reviewed-on: https://go-review.googlesource.com/c/go/+/446258
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Than McIntosh <thanm@google.com>
When coverage testing a local package (defined by a relative import
path such as "./foo/bar") the convention when "-coverprofile" is used
has been to capture source files by full pathname, as opposed to
recording the full import path or the invented import path
("command-line-arguments/") created by the go command in the case of
building named Go files. Doing this makes it much easier to use
collected profiles with "go tool -cover -html=<profile>".
The support for this feature/convention wound up being inadvertantly
dropped during the GOEXPERIMENT=coverageredesign implementation; this
patch restores it.
Fixes#56433.
Change-Id: Ib9556fdc86011b00c155caa614ab23e5148f3eb4
Reviewed-on: https://go-review.googlesource.com/c/go/+/445917
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Change the macOS implementation to use libc calls.
Using libc calls directly is what we do for all the runtime and os syscalls.
Doing so here as well improves consistency and also makes it possible
to cross-compile (from non-Mac systems) macOS binaries that use the
native name resolver.
Fixes#12524.
Change-Id: I011f4fcc5c50fbb5396e494889765dcbb9342336
Reviewed-on: https://go-review.googlesource.com/c/go/+/446178
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
This an end-to-end test that sets GOROOT to a symlink
to the distribution, approximating copying it to a new
location, and checks that packages in the standard library
are not stale, as they would be if paths were embedded
in artifacts.
For #47257
Change-Id: I4ed04df36656ad946a2f6f5ce3194e76e06372e3
Reviewed-on: https://go-review.googlesource.com/c/go/+/445358
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
Run-TryBot: Michael Matloob <matloob@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
We kind of have this mechanism already, just normalizing it and
using it in a bunch of places. Previously a bunch of places cached
slices only for the duration of a single function compilation. Now
we can reuse slices across a whole compiler run.
Use a sync.Pool of powers-of-two sizes. This lets us use not
too much memory, and avoid holding onto memory we're no longer
using when a GC happens.
There's a few different types we need, so generate the code for it.
Generics would be useful here, but we can't use generics in the
compiler because of bootstrapping.
Change-Id: I6cf37e7b7b2e802882aaa723a0b29770511ccd82
Reviewed-on: https://go-review.googlesource.com/c/go/+/444820
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
The recently added rule only works before decomposing slices.
Add a rule that works after decomposing slices.
The reason we need the latter is because although the length may
be a constant, it can be hidden inside a slice that is not constant
(its pointer or capacity might be changing). By applying this
optimization after decomposing slices, we can find more cases
where it applies.
Fixes#56440
Change-Id: I0094e59eee3065ab4d210defdda8227a6e897420
Reviewed-on: https://go-review.googlesource.com/c/go/+/446277
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Since pgo is a new package, it is reasonably straightforward to
encapsulate its state into a non-global object that we pass around,
which will help keep it isolated.
There are no functional changes in this CL, just packaging up the
globals into a new object.
There are two major pieces of cleanup remaining:
1. reflectdata and noder have separate InlineCalls calls for method
wrappers. The Profile is not plumbed there yet, but this is not a
regression as the globals were previously set only right around the
main inlining pass in gc.Main.
2. pgo.ListOfHotCallSites is still global, as it will require more work
to clean up. It is effectively a local variable in InlinePackage,
except that it assumes that InlineCalls is immediately preceded by a
CanInline call for the same function. This is not necessarily true
due to the recursive nature of CanInline. This also means that some
InlineCalls calls may be missing the list of hot callsites right now.
For #55022.
Change-Id: Ic1fe41f73df96861c65f8bfeecff89862b367290
Reviewed-on: https://go-review.googlesource.com/c/go/+/446303
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Pratt <mpratt@google.com>
Parts of package pgo fetch the line number of a node by parsing the
number out of the string returned from ir.Line().
This is indirect and inefficient, so it should be replaced with a more
direct lookup. It is also potentially buggy: ir.Line uses
ctxt.OutermostPos, i.e., the line number where an inlined node in
inlined. We want ctxt.InnermostPos, because that is the line number used
in pprof profiles that we are matching against (See comments on
OutermostPos and InnermostPos).
I'm not sure whether this was an active, as we use ir.Line before and
during inlining. I think we could see CALL nodes with OutermostPos !=
InnermostPos during midstack inlining, but I am not sure. Regardless,
explicitly using the desired position is clearer.
For #55022.
Change-Id: Ic640761c9e1d01cacbf91f3aaeaf284ad7e38dbd
Reviewed-on: https://go-review.googlesource.com/c/go/+/446302
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
For most tests, the test's deadline itself is more appropriate than an
arbitrary timeout layered atop of it (especially once #48157 is
implemented), and testenv.Command already adds cleaner timeout
behavior when a command would run too close to the test's deadline.
That makes RunWithTimeout something of an attractive nuisance. For
now, migrate the two existing uses of it to testenv.CommandContext,
with a shorter timeout implemented using context.WithTimeout.
As a followup, we may want to drop the extra timeouts from these
invocations entirely.
Updates #50436.
Updates #37405.
Change-Id: I16840fd36c0137b6da87ec54012b3e44661f0d08
Reviewed-on: https://go-review.googlesource.com/c/go/+/445597
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This adds a testenv.CommandContext function, with timeout behavior
based on the existing logic in cmd/go.TestScript: namely, the command
is terminated with SIGQUIT (if supported) with an arbitrary grace
period remaining until the test's deadline.
If the test environment does not support executing subprocesses,
CommandContext skips the test.
If the command is terminated due to the timout expiring or the test
fails to wait for the command after starting it, CommandContext marks
the test as failing.
For tests where a shorter timeout is desired (such as for fail-fast
behavior), one may be supplied using context.WithTimeout.
The more concise Command helper is like CommandContext but without
the need to supply an explicit Context.
Updates #50436.
Change-Id: Ifd81fb86c402f034063c9e9c03045b4106eab81a
Reviewed-on: https://go-review.googlesource.com/c/go/+/445596
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Austin Clements <austin@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Make linkgetlineFromPos and getFileIndexAndLine methods on Link, and
give the former a more descriptive name.
The docs are expanded to make it more clear that these are final
file/line visible in programs.
In getFileSymbolAndLine use ctxt.InnermostPos instead of ctxt.PosTable
direct, which makes it more clear that we want the semantics of
InnermostPos.
Change-Id: I7c3d344dec60407fa54b191be8a09c117cb87dd0
Reviewed-on: https://go-review.googlesource.com/c/go/+/446301
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The most important change here is to log output from the child, making
it easier to diagnose problems when the child 'go test' fails.
We can also eliminate the cmd.Wait goroutine by using an os.Pipe, whose
reader will return io.EOF when the child exits.
For #55022.
Change-Id: I1573ea444407d545bdca8525c9ff7b0a2baebf5e
Reviewed-on: https://go-review.googlesource.com/c/go/+/446300
Run-TryBot: Michael Pratt <mpratt@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The test sleeps for 1 millisecond to give the cancellation a moment
to take effect. This is flaky because the request can finish before
the cancellation of the context is seen. It's easy to verify by adding
time.Sleep(2*time.Millisecond)
after 0a6c4c8740/src/net/http/transport.go (L2619).
With this modification, the test fails about 5 times out of 10 runs.
The fix is easy. We just need to block the handler of the second
request until this request is cancelled. I have verify that the
updated test can uncover the issue fixed by CL 257818.
Fixes#55226.
Change-Id: I81575beef1a920a2ffaa5c6a5ca70a4008bd5f94
GitHub-Last-Rev: 99cb1c2eae
GitHub-Pull-Request: golang/go#56500
Reviewed-on: https://go-review.googlesource.com/c/go/+/446676
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Wasm can't handle the recusion for XML nested to depth 10,000.
Cut it off at 5,000 instead. This fixes TestCVE202228131 on trybots
in certain conditions.
Also disable TestCVE202230633 to fix 'go test -v encoding/xml' on gomotes.
Also rename errExeceededMaxUnmarshalDepth [misspelled and unwieldy]
to errUnmarshalDepth.
For #56498.
Change-Id: I7cc337ccfee251bfd9771497be0e5272737114f9
Reviewed-on: https://go-review.googlesource.com/c/go/+/446639
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
The js/wasm builder started failing as of CL 432535 due to needing
'go build' to import standard-library packages that are no longer
installed to GOROOT/pkg. Since js/wasm can't exec subprocesses,
it cannot run 'go build' to generate the export data needed for
these tests.
For #47257.
Change-Id: I804235fd725faf00b27cbed79ee1f43dea8ab734
Reviewed-on: https://go-review.googlesource.com/c/go/+/446635
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
The index function was not handling certain corner cases where there
were two more bytes to be examined in the tail end of the string to
complete the comparison. Fix code to ensure that when the string has
to be shifted two more times the correct bytes are examined.
Also hoisted vsplat to V10 so that all paths use the correct value.
Some comments had incorrect register names and corrected the same.
Added the strings that were failing to strings test for verification.
Fixes#56457
Change-Id: Idba7cbc802e3d73c8f4fe89309871cc8447792f5
Reviewed-on: https://go-review.googlesource.com/c/go/+/446135
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Archana Ravindar <ravindararchana@gmail.com>
Address folding is disabled in CL42172, the commit message of which
said that "In shared library, load/store of global is rewritten to
using GOT and temp register, which conflicts with the use of temp
register for assembling large offset.". Actually this doesn't happen
because the sequence of instructions when rewritten to use Got looks
like this:
MOVD $sym, Rx becomes
MOVD sym@GOT, Rx
If there is an offset off, there will be one more instruction:
ADD $off, Rx, Rx
And MOVD sym, Rx becomes
MOVD sym@GOT, REGTMP
MOVx (REGTMP), Ry
If there is a small offset off, it becomes:
MOVD sym@GOT, REGTMP
MOVx (REGTMP)off, Ry
If off is very large, it becomes:
MOVD sym@GOT, REGTMP
MOVD $off, Rt
ADD Rt, REGTMP
MOVx (REGTMP), Ry
We can see that the address can be calculated correctly, and testing
on darwin/arm64 confirms this.
Removing this restriction is beneficial to further optimize the sequence
of "ADRP+ADD+LD/ST" to "ADRP+LD/ST(offset), so this CL removes it.
Change-Id: I0e9f7bc1723e0a027f32cf0ae2c41cd6df49defe
Reviewed-on: https://go-review.googlesource.com/c/go/+/445535
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Eric Fang <eric.fang@arm.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This change improves the performance of Binomial by implementing an
algorithm that produces smaller intermediate values at each step.
Working with smaller big.Int values has the advantage that fewer allocations
and computations are required for each mathematical operation.
The algorithm used is the Multiplicative Formula, which is a well known
way of calculating the Binomial coefficient and is described at:
https://en.wikipedia.org/wiki/Binomial_coefficient#Multiplicative_formulahttps://en.wikipedia.org/wiki/Binomial_coefficient#In_programming_languages
In addition to that, an optimization has been made to remove a
redundant computation of (i+1) on each loop which has a measurable
impact when using big.Int.
Performance improvement measured on an M1 MacBook Pro
running the existing benchmark for Binomial:
name old time/op new time/op delta
Binomial-8 589ns ± 0% 435ns ± 0% -26.05% (p=0.000 n=10+10)
name old alloc/op new alloc/op delta
Binomial-8 1.02kB ± 0% 0.08kB ± 0% -92.19% (p=0.000 n=10+10)
name old allocs/op new allocs/op delta
Binomial-8 38.0 ± 0% 5.0 ± 0% -86.84% (p=0.000 n=10+10)
Change-Id: I5a830386dd42f062e17af88411dd74fcb110ded9
GitHub-Last-Rev: 6b2fca07de
GitHub-Pull-Request: golang/go#56339
Reviewed-on: https://go-review.googlesource.com/c/go/+/444315
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Packages in GOROOT that don't use cgo will not be installed in
GOROOT/pkg, and will instead be cached as usual like other Go
packages.
- add a internal/buildinternal package to hold the identities of the
five packages that use cgo
- update dist's test code to do a go build std cmd before checking
staleness on builders. Because most of those packages no longer have
install locations, and have dependencies that don't either, the
packages need to be cached to not be stale.
- fix index_test to import packages with the path "." when preparing
the "want" values to compare the indexed data to. (the module index
matches the behavior of build.ImportDir, which always passes in "."
as the path.
- In both the index and go/build Importers, don't set
PkgObj for GOROOT packages which will no longer have install
targets. PkgTargetRoot will still be set to compute target paths,
which will still be needed in buildmode=shared.
- "downgrade" all install actions that don't have a target to build
actions. (The target should already not be set for packages that
shouldn't be installed).
For #47257
Change-Id: Ia5aee6b3b20b58e028119cf0352a4c4a2f10f6b8
Reviewed-on: https://go-review.googlesource.com/c/go/+/432535
Run-TryBot: Michael Matloob <matloob@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
When writing the set of commands excuted for go build -x, $WORK is
substituted for the work directory in all the commnands. But this
includes the cat <<EOF commands used to create a file with the given
contents. While we can expect the shell to substitute $WORK properly,
commands that read input files, such as importcfgs won't do that
substitution.
This is necessary to fix the build_dash_x script test for CL 432535
because it removes .a files from the traditional stdlib install
locations. The test can pass even with importcfg packagefiles in $WORK
because all transitive imports are in the stdlib, and the compiler can
fall back to finding stdlib .a files in their traditional places, but
once they're gone the packagefile paths in $WORK will have paths that
contain the string $WORK, and os.Open will fail to open them for
reading. And since the fallback is gone the test will fail.
For #47257
Change-Id: I5db0066de6ed3ccf97927a78ce0939e3eb14aebe
Reviewed-on: https://go-review.googlesource.com/c/go/+/446116
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Matloob <matloob@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
Before this change, the .a files for the intermediate go toolchains
were produced in the same location as the install target. This change
has them produced in a temporary location instead. This change,
combined with go install not producing .a files for most stdlib
packages by default results in a build of the distribution only
producing .a's for the five packages in std that require cgo. (Before
this change, the .a's for the final build were not being produced, but
stale ones from the intermediate builds were left behind.)
For #47257
Change-Id: I91b826cd1ce9aad9c492fb865e36d34dc8bb188e
Reviewed-on: https://go-review.googlesource.com/c/go/+/436135
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Matloob <matloob@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
In the previous, we will only use the slicing by 8 look up table when
the data size is greater than 16k, but the threshold seems to be too
large. We may lose some performance for the small data size.
In this CL, we change the threshold to 2k, it shows that the performance
is improved greatly when the data size is 2k ~ 16k.
We did some tests for the Random2K ~ Random16K between various
cores(mostly on x86 and arm architecture). Here are the benchmark and
some results:
1. The benchmark for testing:
func BenchmarkRandom(b *testing.B) {
poly := 0x58993511
b.Run("Random256", func(b *testing.B) {
bench(b, uint64(poly), 256)
})
b.Run("Random512", func(b *testing.B) {
bench(b, uint64(poly), 512)
})
b.Run("Random1KB", func(b *testing.B) {
bench(b, uint64(poly), 1<<10)
})
b.Run("Random2KB", func(b *testing.B) {
bench(b, uint64(poly), 2<<10)
})
b.Run("Random4KB", func(b *testing.B) {
bench(b, uint64(poly), 4<<10)
})
b.Run("Random8KB", func(b *testing.B) {
bench(b, uint64(poly), 8<<10)
})
b.Run("Random16KB", func(b *testing.B) {
bench(b, uint64(poly), 16<<10)
})
}
2. Some results:
Apple silicon M1:
Benchmark old new delta
Random/Random2KB-10 362MB/s 801MB/s +121.41%
Random/Random4KB-10 360MB/s 1083MB/s +200.93%
Random/Random8KB-10 359MB/s 1309MB/s +264.88%
Random/Random16KB-10 358MB/s 1466MB/s +309.79%
Neoverse N1:
Benchmark old new delta
Random/Random2KB-160 397MB/s 493MB/s +24.23%
Random/Random4KB-160 397MB/s 742MB/s +86.86%
Random/Random8KB-160 398MB/s 995MB/s +150.12%
Random/Random16KB-160 398MB/s 1196MB/s +200.58%
Silver 4116:
Benchmark old new delta
Random/Random2KB-48 252MB/s 418MB/s +65.79%
Random/Random4KB-48 253MB/s 621MB/s +145.72%
Random/Random8KB-48 254MB/s 796MB/s +213.07%
Random/Random16KB-48 258MB/s 929MB/s +260.46%
EPYC 7251:
Benchmark old new delta
Random/Random2KB-32 255MB/s 380MB/s +48.88%
Random/Random4KB-32 255MB/s 561MB/s +119.73%
Random/Random8KB-32 255MB/s 738MB/s +189.18%
Random/Random16KB-32 255MB/s 877MB/s +243.80%
Change-Id: Ib7b4f6826c3edd6f315cac8057d52b6da252a652
Reviewed-on: https://go-review.googlesource.com/c/go/+/445475
Run-TryBot: Eric Fang <eric.fang@arm.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Auto-Submit: Keith Randall <khr@golang.org>
The size of gnuVersym should be multiples of 2. If not, the input is
invalid. No Library and Version information is added to sym in this
case. The current implementation of gnuVersion does not report errors
for invalid input.
While at here, bring back the comment that states that the undef entry
at the beginning is skipped. This is not an off-by-one error.
No test case because the problem can only happen for invalid data. Let
the fuzzer find cases like this.
Fixes#56429.
Change-Id: Ia39ad8bd509088a81cc77f7a76e23185d40a5765
GitHub-Last-Rev: 3be0cc1b15
GitHub-Pull-Request: golang/go#56431
Reviewed-on: https://go-review.googlesource.com/c/go/+/445555
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Run-TryBot: Meng Zhuo <mzh@golangcn.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Test2json is parsing the output stream from the test, which includes
package testing's own framing lines intermingled with other output,
in particular any output printed via fmt.Printf, println, and so on.
We have had recurring problems with unexpected partial output lines
causing a framing line to be missed.
A recent talk at GopherCon gave an example of an integration test
involving Docker that happened to print \r-terminated lines instead
of \n-terminated lines in some configurations, which in turn broke
test2json badly. (https://www.gophercon.com/agenda/session/944259)
There are also a variety of open reported issues with similar problems,
which this CL also addresses. The general approach is to add a new
testing flag -test.v=json that means to print additional output to help
test2json. And then test2json takes advantage of that output.
Among the fixes:
- Identify testing framing more reliably, using ^V
(#23036, #26325, #43683, GopherCon talk)
- Test that output with \r\n endings is handled correctly
(#43683, #34286)
- Use === RUN in fuzz tests (#52636, #48132)
- Add === RUN lines to note benchmark starts (#27764, #49505)
- Print subtest --- PASS/FAIL lines as they happen (#29811)
- Add === NAME lines to emit more test change events,
such as when a subtest stops and the parent continues running.
- Fix event shown in overall test failure (#27568)
- Avoid interleaving of writes to os.Stdout and os.Stderr (#33419)
Fixes#23036.
Fixes#26325.
Fixes#27568.
Fixes#27764.
Fixes#29811.
Fixes#33419.
Fixes#34286.
Fixes#43683.
Fixes#49505.
Fixes#52636.
Change-Id: Id4207b746a20693f92e52d68c6e4a7f8c41cc7c6
Reviewed-on: https://go-review.googlesource.com/c/go/+/443596
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
flag.Example() has this comment:
... one must execute, typically at the start of main (not init!):
flag.Parse()
We don't run it here because this is not a main function
This example function will be renamed to "main" at pkg.go.dev, which
makes the comment confusing.
See https://pkg.go.dev/flag#example-package.
This change modify the comment to clarify this situation.
Change-Id: I17357fdaaefe54791fff8fbbf6a33003af207f88
GitHub-Last-Rev: eeea8ce39c
GitHub-Pull-Request: golang/go#56411
Reviewed-on: https://go-review.googlesource.com/c/go/+/445315
Run-TryBot: Ian Lance Taylor <iant@google.com>
Run-TryBot: Rob Pike <r@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Rob Pike <r@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The Cancel and WaitDelay fields recently added to exec.Cmd are
intended to support exactly the sort of cancellation behavior that we
need for script tests. Use them, and simplify the cmd/go tests
accordingly.
The more robust implementation may also help to diagose recurring test
hangs (#50187).
For #50187.
Updates #27494.
Change-Id: I7817fca0dd9a18e18984a252d3116f6a5275a401
Reviewed-on: https://go-review.googlesource.com/c/go/+/445357
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
To pick up CL 443575.
Note: This CL also update other x repositories because CL 444536 updates
x/tools dependencies to latest tagged version.
Done by
go get -d golang.org/x/tools@8166dca1cec9
go mod tidy
go mod vendor
Change-Id: Ie2836bb4ebc1db0150ba240af4e2b750dbf1e0b4
Reviewed-on: https://go-review.googlesource.com/c/go/+/445055
Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Add the experimental compiler flag -altcomparable. If set, the
compiler uses alternative comparable semantics: any ordinary
(non-type parameter) interface implements the comparable
constraint.
This permits experimenting with this alternative semantics
akin to what is proposed in #52509.
For #52509.
Change-Id: I64192eee6f2a550eeb50de011079f2f0b994cf94
Reviewed-on: https://go-review.googlesource.com/c/go/+/444636
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Implement proposal #54880, to automatically seed the global source.
The justification for this not being a breaking change is that any
use of the global source in a package's init function or exported API
clearly must be valid - that is, if a package changes how much
randomness it consumes at init time or in an exported API, that
clearly isn't the kind of breaking change that requires issuing a v2
of that package. That kind of per-package change in the position
of the global source is indistinguishable from seeding the global
source differently. So if the per-package change is valid, so is auto-seeding.
And then, of course, auto-seeding means that packages will be
far less likely to depend on the specific results of the global source
and therefore not break when those kinds of per-package changes
happen in the future.
Seed(1) can be called in programs that need the old sequence from
the global source and want to restore the old behavior.
Of course, those programs will still be broken by the per-package
changes just described, and it would be better for them to allocate
local sources rather than continue to use the global one.
Fixes#54880.
Change-Id: Ib9dc3307b97f7a45587a9cc50d81f919d3edc7ae
Reviewed-on: https://go-review.googlesource.com/c/go/+/443058
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
The scripts added in CL 421455 passed on the TryBots, but failed on
the "-stretch" builders, which supply Mercurial 4.0
(released 2016-11-01).
Debian 9 “Stretch” has been at end-of-life since June 30, 2022, but
until we can turn down the outdated builders (#56414) we should keep
them passing tests.
For #27494.
Updates #56414.
Change-Id: I9df0ed452dfbfaeb1b4c0d869d02dd9ed21b3ff6
Reviewed-on: https://go-review.googlesource.com/c/go/+/445356
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
GOFLAGS didn't split on quotes because no other env vars
(such as CC, CXX, ...) did either. This kept them all consistent.
CL 341936 changed everything but GOFLAGS, making them inconsistent.
Split GOFLAGS the same way as the other environment variables.
Fixes#26849.
Change-Id: I99bb450fe30cab949da48af133b6a36ff320532f
Reviewed-on: https://go-review.googlesource.com/c/go/+/443956
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
The scripts added in CL 421455 passed on the TryBots, but failed on a
subset of the builders that have older 'git' binaries installed.
Notably, the older versions of git do not support:
- 'git branch -m' before the current branch has a commit
- 'init.defaultBranch' in the '.gitconfig' file, and
- 'git branch -c'.
We address those by, respectively:
- waiting to run 'git branch -m' until after the first commit
- always running 'git branch -m' explicitly to set the branch name, and
- using 'git checkout' instead of 'git branch -c' to set branch parents.
Updates #27494.
Change-Id: I42f012f5add8f31e41d077d752d8268aacbce8a4
Reviewed-on: https://go-review.googlesource.com/c/go/+/445355
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
After this CL, the only test requests that should still reach
vcs-test.golang.org are for Subversion repos, which are not yet handled.
The interceptor implementation should also allow us to redirect other
servers (such as gopkg.in) fairly easily in a followup change if
desired.
For #27494.
Change-Id: I8cb85f3a7edbbf0492662ff5cfa779fb9b407136
Reviewed-on: https://go-review.googlesource.com/c/go/+/427254
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
The new server reconstructs the vcs-test repos on the fly using
scripts that run the actual version-control binaries.
This allows those repos to be code-reviewed using our normal tools —
and, crucially, allows contributors to add new vcs-test contents
as part of a contributed CL.
It also prevents failures due to network errors reaching
vcs-test.golang.org (such as when developing offline), and allows us
to iterate on the repo contents without dealing with annoying and
unpredictable GCS caching behavior.
We can't quite turn down vcs-test.golang.org yet — this server doesn't
yet handle "go-import" metadata (and related authentication behaviors),
and doesn't serve Subversion repos.
But we're getting much closer!
For #27494.
Change-Id: I233fc718617aed287b0f7248bd8cfe1e5cebe96b
Reviewed-on: https://go-review.googlesource.com/c/go/+/421455
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
This replaces a large set of individual GOOS and GOARCH conditions
with a smaller set of more verbose conditions. On balance, the more
uniform structure and more concise documentation seem worth the
verbosity.
For #27494.
Change-Id: I73fdf9e7180a92cb1baf5d4631aeecb26ce15181
Reviewed-on: https://go-review.googlesource.com/c/go/+/420054
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Previously, the script engine implicitly escaped the path in the
$WORK environment variable to be the literal string '$WORK', which
produces somewhat better error messages in case of failure.
However, for a general-purpose script engine that implicit behavior is
surprising, and it isn't really necessary.
For #27494.
Change-Id: Ic1d5b8801bbd068157315685539e7cc2795b3aa5
Reviewed-on: https://go-review.googlesource.com/c/go/+/426854
Reviewed-by: Russ Cox <rsc@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
This change decouples the script engine from both cmd/go and the
testing package; I intend to reuse it in the replacement for the
vcs-test.golang.org server.
This change also adds a few new script commands:
- 'echo' echoes its arguments, useful for verifying argument expansion.
- 'cat' prints the contents of files, useful for debugging failing script tests.
- 'help' displays information about script commands and conditions,
reducing the toil of maintaining lists in the README file.
The 'cmp' and 'cmpenv' commands now use internal/diff instead of their
own separate diff implementation.
The 'env' command now writes to the script log instead of the stdout
buffer. (This makes it more consistent with the behavior of other
synchronous builtins.)
The 'stale' command no longer logs output when a target is
unexpectedly non-stale. (However, the ouput of the 'stale' command is
not usually very useful anyway.)
The 'grep', 'stdout', and 'stderr' commands now display matching lines
(like Unix 'grep'), making their negation behavior more consistent
with running real commands on a command-line.
Likewise, the 'cmp' command now always displays differences. That
makes it useful with the '?' prefix to produce diffs for informational
purposes while debugging.
For #27494.
Change-Id: If49fd81d9b922d07c20618a8e2cef908191f9ef6
Reviewed-on: https://go-review.googlesource.com/c/go/+/419875
Reviewed-by: Russ Cox <rsc@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The appendInt function previously performed a double pass
over the formatted integer. We can avoid the second pass
if we knew the exact length of formatted integer,
allowing us to directly serialize into the output buffer.
Rename formatNano to appendNano to be consistent with
other append-like functionality.
Performance:
name old time/op new time/op delta
FormatRFC3339Nano 109ns ± 1% 72ns ± 1% -34.06% (p=0.000 n=10+10)
Change-Id: Id48f77eb4976fb1dcd6e27fb6a02d29cbf0c026a
Reviewed-on: https://go-review.googlesource.com/c/go/+/444278
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: David Chase <drchase@google.com>
This is an experiment to see the impact of a potential spec change:
As an exception to the rule that constraint satisfaction is the same
as interface implementation, if the flag Config.AltComparableSemantics
is set, an ordinary (non-type parameter) interface satisfies the
comparable constraint. (In go/types, the flag is not exported to
avoid changing the API.)
Disabled by default. Test files can set the flag by adding
// -altComparableSemantics
as the first line in the file.
For #52509.
Change-Id: Ib491b086feb5563920eaddefcebdacb2c5b72d61
Reviewed-on: https://go-review.googlesource.com/c/go/+/444635
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
The existing implementation fails to convert the remainder
microseconds to nanoseconds. This causes sysmon to consume
much more cpu, and generate lots of context switches.
We can also do a little better here to avoid division by a
constant. I used go to determine the magic numbers.
Fixes#56374
Change-Id: I2e37ec218b9027efab6db4634eed1504c0c1b3c8
Reviewed-on: https://go-review.googlesource.com/c/go/+/444735
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
The right-hand side SLLI always contains valid content in the high 32 bits,
so we should use the 64 bit integer type. Using wrong type may lead to wrong
optimizations in cse pass.
Should fix x/text test failures.
Change-Id: I972dd913b8fb238d180bb12f8b1801adc8503fc0
Reviewed-on: https://go-review.googlesource.com/c/go/+/443875
Reviewed-by: Dmitri Goutnik <dgoutnik@gmail.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org>
Reviewed-by: David Chase <drchase@google.com>
Packages in modules don't have a Target set for them, so the current
logic for determining staleness always reports them as stale. Instead it
should be reporting whether "go install" would do anything, and sholud
be false after a go install. If a package does not have a Target,
instead check to see whether we've cached its build artifact.
Change-Id: Ie7bdb234944353f6c2727bd8bf939cc27ddf3f18
Reviewed-on: https://go-review.googlesource.com/c/go/+/444619
Run-TryBot: Michael Matloob <matloob@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
This patch has a couple of minor fixes to new-style counter insertion
(noticed these problems while working on the fix for issue 56370).
First, make sure that the function registration sequence (writing of
nctrs, pkgid, funcid to counter var prolog) comes prior to the first
counter update (they were reversed up to this point, due to an
artifact of the way cmd/internal/edit operates).
Second, fix up "per function" counter insertion mode (an experimental
feature disabled by default that adds just a single counter to each
function as opposed to one per basic block), which was failing to
insert the single counter in the right place.
Change-Id: Icfb613ca385647f35c0e52da2da8edeb2a506ab7
Reviewed-on: https://go-review.googlesource.com/c/go/+/444835
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Than McIntosh <thanm@google.com>
This patch fixes a problem in which we can get a data race on a
coverage counter function registration sequence. The scenario is that
package P contains a function F that is built with coverage, then F is
inlined into some other package that isn't being instrumented. Within
F's exported function body counter updates were being done with
atomics, but the initial registration sequence was not, which had the
potential to trigger a race. Fix: if race mode is enabled and we're
using atomics for counter increments, also use atomics in the
registration sequence.
Fixes#56370.
Change-Id: If274b61714b90275ff95fc6529239e9264b0ab0c
Reviewed-on: https://go-review.googlesource.com/c/go/+/444617
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Than McIntosh <thanm@google.com>
If a user is running a go list command that wouldn't trigger a build
(for example if -export was passed), don't print the cached stdout
outputs for previous builds of the artifacts.
Fixes#56375
Change-Id: I1d3e6c01d0eb3dada941bb2783ce2ac69aa3d5d2
Reviewed-on: https://go-review.googlesource.com/c/go/+/444836
Run-TryBot: Michael Matloob <matloob@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
It was noted in the go1.9 release notes that functions in math/bits
may be implemented by compiler intrinsics, but this never made it to
the documentation.
This change adapts the wording of the release notes and puts it in the
documentation for math/bits.
Change-Id: Ibeea88eaf7df10952cbe670885e910ac30b49d55
Reviewed-on: https://go-review.googlesource.com/c/go/+/444035
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
windows-amd-2012 builder seems to have some problems handling
exception thrown in external C code which is affecting
TestVectoredHandlerExceptionInNonGoThread.
The issue is known and discussed in #49681.
This Cl skips the offending test on windows-amd-2012.
Change-Id: I7ca4353c9e531f0d75ac6a8dbd809acfa1f15bf9
Reviewed-on: https://go-review.googlesource.com/c/go/+/444616
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
Run-TryBot: Quim Muntal <quimmuntal@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This minimizes addi usage inside vector heavy loops. This
results in a small performance uptick on P9 ppc64le/linux.
Likewise, cleanup some minor whitespace issues around comments.
The implementation from crypto/sha256 is also shared with notsha256.
It is copied, but preserves notsha256's go:build directives. They are
otherwise identical now. Previously, bootstrap restrictions required
workarounds to support XXLOR on older toolchains. This is not needed
anymore as the minimum bootstrap (1.17) compiler will support XXLOR.
name old speed new speed delta
Hash8Bytes/New 28.8MB/s ± 0% 30.5MB/s ± 0% +5.98%
Hash8Bytes/Sum224 29.5MB/s ± 0% 31.3MB/s ± 0% +6.17%
Hash8Bytes/Sum256 29.5MB/s ± 0% 31.2MB/s ± 0% +5.80%
Hash1K/New 287MB/s ± 0% 312MB/s ± 0% +8.60%
Hash1K/Sum224 289MB/s ± 0% 312MB/s ± 0% +7.99%
Hash1K/Sum256 289MB/s ± 0% 312MB/s ± 0% +7.98%
Hash8K/New 313MB/s ± 0% 338MB/s ± 0% +8.12%
Hash8K/Sum224 313MB/s ± 0% 338MB/s ± 0% +8.20%
Hash8K/Sum256 313MB/s ± 0% 338MB/s ± 0% +8.12%
Change-Id: Ib386d6306673b4e6553ee745ec2e1b53a9722df1
Reviewed-on: https://go-review.googlesource.com/c/go/+/441815
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Archana Ravindar <aravind5@in.ibm.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
When a function type is copied (e.g. for substituting type
parameters), we make copies of its parameter ir.Name nodes, so
they are not shared with the old function type. But currently a
blank (_) identifier is not copied but shared. The parameter
node's frame offset is assigned (in ABI analysis) and then used in
the concurrent backend. Shared node can cause a data race. Make a
new blank parameter node to avoid sharing. (Unified IR does already
not have this problem. This fixes non-unified-IR mode.)
This seems to fix#55357.
Change-Id: Ie27f08e5589ac7d5d3f0d0d5de1a21e4fd2765c9
Reviewed-on: https://go-review.googlesource.com/c/go/+/443158
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Error messages currently print floats with %.6g, which means that if
you tried to convert something close to, but not quite, an integer, to
an integer, the error you get looks like "cannot convert 1 to type
int", when really you want "cannot convert 0.9999999 to type int".
Add more digits to floats when printing them, to make it clear that they
aren't quite integers. This helps for errors which are the result of not
being an integer. For other errors, it won't hurt much.
Fixes#56220
Change-Id: I7f5873af5993114a61460ef399d15316925a15a5
Reviewed-on: https://go-review.googlesource.com/c/go/+/442935
Reviewed-by: Rob Pike <r@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
We add strict checking to marshal and unmarshal methods,
rather than Parse to maintain compatibility in Parse behavior.
Also, the Time.Format method has no ability to report errors.
The Time.Marshal{Text,JSON} and Time.Unmarshal{Time,JSON} methods
are already documented as complying with RFC 3339, but have
edge cases on both marshal and unmarshal where it is incorrect.
The Marshal methods already have at least one check to comply
with RFC 3339, so it seems sensible to expand this to cover
all known violations of the specification.
This commit fixes all known edge cases for full compliance.
Two optimizations are folded into this change:
1. parseRFC3339 is made generic so that it can operate
directly on a []byte as well as string.
This avoids allocating or redundant copying
when converting from string to []byte.
2. When marshaling, we verify for correctness based
on the serialized output, rather than calling
attribute methods on the Time type. For example,
it is faster to check that the 5th byte is '-'
rather than check that Time.Year is within [0,9999],
since Year is a relatively expensive operation.
Performance:
name old time/op new time/op delta
MarshalJSON 109ns ± 2% 99ns ± 1% -9.43% (p=0.000 n=10+10)
UnmarshalText 158ns ± 4% 143ns ± 1% -9.17% (p=0.000 n=9+9)
Updates #54580
Updates #54568
Updates #54571
Change-Id: I1222e45a7625d1ffd0629be5738670a84188d301
Reviewed-on: https://go-review.googlesource.com/c/go/+/444277
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently, the lowerBlock function is reused with lateLowerValue, meaning
that any block rewriting rules in the late lower pass are silently ignored.
Change the late lower pass to actually use the lateLowerBlock function with
the lateLowerValue function.
Change-Id: Iaac1c2955bb27078378cac50cde3716e79a7d9f8
Reviewed-on: https://go-review.googlesource.com/c/go/+/444335
Run-TryBot: Joel Sing <joel@sing.id.au>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Replace even in places where before we have a specific error message
or different control-flow (except in TestTypeString or TestObjectString)
because failing to type-check in virtually all cases represents an error
in the test itself.
Change-Id: I9f1e6d25bddd92c168353409b281b5a3f29a747c
Reviewed-on: https://go-review.googlesource.com/c/go/+/443915
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Instead of having various inconsistent helper functions, rely on
4 helper functions with consistent naming and parameters:
- parse and mustParse
- typecheck and mustTypecheck
Panic rather than call t.Fatal in the mustX functions to simplify
their use.
Use the new functions in tests consistently.
Change-Id: Ib19dc5cc470b51512c23c09df32c379dc3eb8f4b
Reviewed-on: https://go-review.googlesource.com/c/go/+/443757
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
If there is no current G while handling an exception it means
the exception was originated in a non-Go thread.
The best we can do is ignore the exception and let it flow
through other vectored and structured error handlers.
I've removed badsignal2 from sigtramp because we can't really know
if the signal is bad or not, it might be handled later in the chain.
Fixes#50877
Updates #56082
Change-Id: Ica159eb843629986d1fb5482f0b59a9c1ed91698
Reviewed-on: https://go-review.googlesource.com/c/go/+/442896
Reviewed-by: Alex Brainman <alex.brainman@gmail.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Quim Muntal <quimmuntal@gmail.com>
Reviewed-by: David Chase <drchase@google.com>
These tests previously had a “skipSpecialPlatforms” function, added in
CL 8611 to skip tests on (apparently) NaCL and iOS. The iOS builders
no longer match the condition (GOOS=ios is its own thing now), and the
NaCL port no longer exists.
The name of the function also isn't very evocative, since it doesn't
say what is “special” about the platforms to cause them to be
skipped.m
Since the check is intending to run the tests only on platforms where
gc export data is available, and to a first approximation
“gc export data is available” on exactly the platforms that can run
“go build” to produce that export data, we can use the testenv function
instead of a one-off.
Updates #38485.
Change-Id: I8f38b9604300d165147f8942e945ab762419fad7
Reviewed-on: https://go-review.googlesource.com/c/go/+/444155
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Several OSes don't ever reach exitThread, On AIX, Plan9, Solaris, and
Windows, we throw if this function is accidentally reached. Do the same
on Darwin and OpenBSD for consistency.
Change-Id: Icd189b11179755a28b3ec48b267349c57facbf24
Reviewed-on: https://go-review.googlesource.com/c/go/+/443717
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Pratt <mpratt@google.com>
Ms are allocated via standard heap allocation (`new(m)`), which means we
must keep them alive (i.e., reachable by the GC) until we are completely
done using them.
Ms are primarily reachable through runtime.allm. However, runtime.mexit
drops the M from allm fairly early, long before it is done using the M
structure. If that was the last reference to the M, it is now at risk of
being freed by the GC and used for some other allocation, leading to
memory corruption.
Ms with a Go-allocated stack coincidentally already keep a reference to
the M in sched.freem, so that the stack can be freed lazily. This
reference has the side effect of keeping this Ms reachable. However, Ms
with an OS stack skip this and are at risk of corruption.
Fix this lifetime by extending sched.freem use to all Ms, with the value
of mp.freeWait determining whether the stack needs to be freed or not.
Fixes#56243.
Change-Id: Ic0c01684775f5646970df507111c9abaac0ba52e
Reviewed-on: https://go-review.googlesource.com/c/go/+/443716
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Pratt <mpratt@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
go/build and cmd/go will stop returing Targets for stdlib .a files, and
stop producing the .a files is pkg/GOOS_GOARCH. update tests to
anticipate that and to pass in importcfgs instead of expecting the
compiler can find .a files in their old locations.
Adds code to determine locations of .a files to internal/goroot. Also
adds internal/goroot to dist's bootstrap directories and changes
internal/goroot to build with a bootstrap version of Go.
Change-Id: Ie81e51105bddb3f0e374cbf47e81c23edfb67fa5
Reviewed-on: https://go-review.googlesource.com/c/go/+/442303
Reviewed-by: Michael Matloob <matloob@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Matloob <matloob@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
When computing package initialization order, special case the counter
variables inserted by "cmd/cover" for coverage instrumentation, since
their presence can perturb the order in which variables are
initialized in ways that are user-visible and incorrect with respect
to the original (uninstrumented) program.
Fixes#56293.
Change-Id: Ieec9239ded4f8e2503ff9bbe0fe171afb771b335
Reviewed-on: https://go-review.googlesource.com/c/go/+/443715
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
CL 428157 and CL 428759 switched debuglog to using unsafe.String and
unsafe.Slice, which broke the build with -tags=debuglog because this is
a no write barrier context, but runtime.unsafeString and unsafeSlice can
panic, which includes write barriers.
We could add a panicCheck1 path to these functions to reallow write
barriers, but it is a big mess to pass around the caller PC,
particularly since the compiler generates calls. It is much simpler to
just avoid unsafe.String and Slice.
Also add a basic test to build the runtime with -tags=debuglog to help
avoid future regressions.
For #54854.
Change-Id: I702418b986fbf189664e9aa4f40bc7de4d9e7781
Reviewed-on: https://go-review.googlesource.com/c/go/+/443380
Run-TryBot: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
The immediate reason is that we want to use godebug from math/rand,
and math/rand importing godebug importing os causes an import cycle
in package testing.
More generally, the new approach to backward compatibility outlined
in discussion #55090 will require using this package from other similarly
sensitive places, perhaps even package os itself. Best to remove all
dependencies.
Preparation for #54880.
Change-Id: Ia01657a2d90e707a8121a336c9db3b7247c0198f
Reviewed-on: https://go-review.googlesource.com/c/go/+/439418
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
UnsafePointer.Store, UnsafePointer.CompareAndSwap were missing,
although .StoreNoWB and .CompareAndSwapNoWB existed.
Same for Pointer[T}.
Do the linkname tricks necessary to add those methods.
Change-Id: I925ee27673288accb15ebe93898f9eb01ab46a98
Reviewed-on: https://go-review.googlesource.com/c/go/+/443379
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Previously we used Go 1.17, but we realized thanks to tickling
a pre-Go1.17.3 bug that if we are going to change the bootstrap
toolchain that we should default to the latest available point release
at the time we make the switch, not the initial major release, so as
to avoid bugs that were fixed in the point releases.
This CL updates the default search locations and the release notes.
Users who run make.bash and depend on finding $HOME/sdk/go1.17
may need to run
go install golang.org/dl/go1.17.13@latest
go1.17.13 download
to provide a Go 1.17.13 toolchain to their builds.
Change-Id: I3a2511f088cf852470a7216a5a41ae775fb561b6
Reviewed-on: https://go-review.googlesource.com/c/go/+/439419
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The only purpose of using the SHA256 in the file name is
collision avoidance. Using just the first 64 bits (16 hex digits)
will be more than enough, unless people start storing billions
of test cases in their corpora.
The shorter names are nicer for just about everything:
command lines, repository listings, and so on.
Change-Id: I67c760023bed85ba3ffd4f8058f86ef778322ba7
Reviewed-on: https://go-review.googlesource.com/c/go/+/443335
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Peter Weinberger <pjw@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
The src.NoXPos in fake receivers was leaking, through a series of
mishaps, all the way to logopt. If done just so, this can lead to
a compiler crash. This makes logopt crash-proof and eliminates the
root cause as well.
I'm reluctant to write a test for this because it's kinda slow
and involved; my working test is "compile something that mentions
the flag package with -json=0,$TMPDIR flag, then be sure that
$TMPDIR/flag/__unnamed__.json was not created".
Change-Id: I384b717c0e7522953d22d61f7e06319e11192d7d
Reviewed-on: https://go-review.googlesource.com/c/go/+/443156
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Currently, the entry-point to this test is a Bash script that smoke
tests the FORTRAN compiler and then runs a FORTRAN-containing Go test.
This CL rearranges things so a pure Go Go test smoke tests the FORTRAN
compiler and then runs a non-test FORTRAN-containing Go binary.
While we're here, we fix a discrepancy when the host is GOARCH=amd64,
but the target is GOARCH=386. Previously, we would pick the wrong
libgfortran path because we didn't account for the cross-compilation,
causing the link to fail. Except for some reason this was ignored and
the test nevertheless "passed". In the new test we're a little more
strict, so this build failure will cause the test to fail, so we add a
little logic to account for cross-compilation with the host toolchain.
For #37486.
Change-Id: Ie6f70066885d6fbb4e1b5a2b1e13b85dee5b359b
Reviewed-on: https://go-review.googlesource.com/c/go/+/443069
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Austin Clements <austin@google.com>
Currently, TestCgoSigfwd will pass incorrectly if the SIGSEGV that
originates in Go mistakenly goes to the C SIGSEGV handler. Fix this by
adding a signal-atomic variable that tracks what the expected behavior
is.
Change-Id: Id2a9fa3b209299dccf90bb60720b89ad96838a9c
Reviewed-on: https://go-review.googlesource.com/c/go/+/443072
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This migrates testsigfwd, which uses some one-off build
infrastructure, to be part of the runtime's testprogcgo.
The test is largely unchanged. Because it's part of a larger binary,
this CL renames a few things and gates the constructor-time signal
handler registration on an environment variable. This CL also replaces
an errant fmt.Errorf with fmt.Fprintf.
For #37486, since it eliminates a non-go-test from dist.
Change-Id: I0efd146ea0a0a3f0b361431349a419af0f0ecc61
Reviewed-on: https://go-review.googlesource.com/c/go/+/443068
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
On newer amd64 cpus 3 operand LEA instructions are slow, CL 114655 split
them to 2 LEA instructions in genssa.
This CL make late lower pass run after addressing modes, and split 3
operand LEA in late lower pass so that we can do common-subexpression
elimination for splited LEAs.
Updates #21735
Change-Id: Ied49139c7abab655e1a14a6fd793bdf9f987d1f1
Reviewed-on: https://go-review.googlesource.com/c/go/+/440035
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Joedian Reid <joedian@golang.org>
Assuming the two values are valid and non-comparable, Equal should panic.
x := reflect.ValueOf([]int{1, 2, 3})
x.Equal(x) // can not report false, should panic
Assuming one of them is non-comparable and the other is invalid, it should
always report false.
x := reflect.ValueOf([]int{1, 2, 3})
y := reflect.ValueOf(nil)
x.Equal(y) // should report false
For #46746.
Change-Id: Ifecd77ca0b3de3019fae2be39048f9277831676c
Reviewed-on: https://go-review.googlesource.com/c/go/+/440037
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Remove a vestigial " // BUG" comment as there is no bug in the relevant code section and comment predated other changes. Also removed a needless allocation and conformed to the "v, ok := a[x]" standard convention. Tests are passing.
Change-Id: Id28ad1baf77447052b54b341f018e573bac0c11a
GitHub-Last-Rev: 26084698bf
GitHub-Pull-Request: golang/go#56210
Reviewed-on: https://go-review.googlesource.com/c/go/+/442815
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: David Chase <drchase@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Rob Pike <r@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
The Grow method is like the proposed slices.Grow function
in that it ensures that the slice has enough capacity to append
n elements without allocating.
The implementation of Grow is a thin wrapper over runtime.growslice.
This also changes Append and AppendSlice to use growslice under the hood.
Fixes#48000
Change-Id: I992a58584a2ff1448c1c2bc0877fe76073609111
Reviewed-on: https://go-review.googlesource.com/c/go/+/389635
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
The testasan test was added back in 2013 (CL 10126044), many years
before Go added ASAN support in 2021 (CL 298611). So, in fact,
testasan does not test Go ASAN support at all, as you might expect
(misc/cgo/testsanitizers does that). It's intended to test whether the
Go memory allocator works in a mixed C/Go binary where the C code is
compiled with ASAN. The test doesn't actually use ASAN in any way; it
just simulates where ASAN of 2013 put its shadow mappings. This made
sense to test at the time because Go was picky about where its heap
landed and ASAN happened to put its mappings exactly where Go wanted
to put its heap. These days, Go is totally flexible about its heap
placement, and I wouldn't be surprised if ASAN also works differently.
Given all of this, this test adds almost no value today. Drop it.
For #37486, since it eliminates a non-go-test from dist.
Change-Id: I0292f8efbdc0e1e39650715604535c445fbaa87f
Reviewed-on: https://go-review.googlesource.com/c/go/+/443067
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Austin Clements <austin@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
We haven't used this in a while and it's going to complicate later
changes to dist, so drop support. This was primarily for supporting
slow QEMU-based builders, but an alternative and simpler way to do
that if we need to in the future is to supply a go_exec wrapper to run
tests in QEMU, like we do for other emulated platforms.
Simplification for #37486.
Change-Id: Idc0383f59c61d8546ea3b4d2eede4acdaf30d9b6
Reviewed-on: https://go-review.googlesource.com/c/go/+/431256
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Austin Clements <austin@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Austin Clements <austin@google.com>
LLVM recently introduced ZSTD compression for ELF sections. The error
message when failing to read the compressed section calls this
COMPRESS_ZLIB+2 while it should report it as COMPRESS_ZLIB+1.
Change-Id: I869aa69baf0c8c86665475b47343c790587cb39b
Reviewed-on: https://go-review.googlesource.com/c/go/+/443035
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The existing version of this test contains several races it tries to
control with sleeps. Unfortunately, it is still flaky on darwin because
writing `fg` in bash too early can apparently result in failure to
actually continue the stopped child.
Rather than continuing to get perfect timing with bash, rewrite this to
eliminate bash and instead perform the same PTY operations that bash
would do.
This test is still quite complex because psuedo-terminals are
interminably complicated, but I believe it is no longer racy.
Technically there are still two races (waiting for child to enter read()
and waiting for the darwin kernel to wake the read after TIOCSPGRP), but
loss of either of these races should only mean we fail to test the
desired darwin EINTR case, not failure.
This test is skipped on DragonflyBSD, as it tickles a Wait hang bug
(#56132).
Updates #56132.
Fixes#37329.
Change-Id: I0ceaf5aa89f6be0f1bf68b2140f47db673cedb33
Reviewed-on: https://go-review.googlesource.com/c/go/+/440220
Run-TryBot: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
This adds the function "start line number" to runtime._func and
runtime.inlinedCall objects. The "start line number" is the line number
of the func keyword or TEXT directive for assembly.
Subtracting the start line number from PC line number provides the
relative line offset of a PC from the the start of the function. This
helps with source stability by allowing code above the function to move
without invalidating samples within the function.
Encoding start line rather than relative lines directly is convenient
because the pprof format already contains a start line field.
This CL uses a straightforward encoding of explictly including a start
line field in every _func and inlinedCall. It is possible that we could
compress this further in the future. e.g., functions with a prologue
usually have <line of PC 0> == <start line>. In runtime.test, 95% of
functions have <line of PC 0> == <start line>.
According to bent, this is geomean +0.83% binary size vs master and
-0.31% binary size vs 1.19.
Note that //line directives can change the file and line numbers
arbitrarily. The encoded start line is as adjusted by //line directives.
Since this can change in the middle of a function, `line - start line`
offset calculations may not be meaningful if //line directives are in
use.
For #55022.
Change-Id: Iaabbc6dd4f85ffdda294266ef982ae838cc692f6
Reviewed-on: https://go-review.googlesource.com/c/go/+/429638
Run-TryBot: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Enable -msan flag on freebsd/amd64 and amend PIE comment in
internal/work/init.go to indicate that MSAN requires PIE on all platforms
except linux/amd64.
R=go1.20
For #53298
Change-Id: I93d94efa95d7f292c23c433fb1d3f4301d820bde
Reviewed-on: https://go-review.googlesource.com/c/go/+/411275
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
There are getting to be enough special cases in this wrapper that
the increase in clarity from having a single file is starting to be
outweighed by the complexity from chained conditionals.
Updates #50138.
Updates #13987.
Change-Id: If4f1be19c0344e249aa6092507c28363ca6c8438
Reviewed-on: https://go-review.googlesource.com/c/go/+/442575
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Write .note signature section when targeting FreeBSD, similar to NetBSD
and OpenBSD. This allows binaries to declare the ABI version they were
compiled for and opt out of ASLR when compiled with -race.
Fixes#48164
Change-Id: Ie54dd5c70697a3f42a75fd640540350fd8a4dc71
Reviewed-on: https://go-review.googlesource.com/c/go/+/412494
Reviewed-by: Meng Zhuo <mzh@golangcn.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Yuval Pavel Zholkover <paulzhol@gmail.com>
Workspaces with a single module would enter mod=vendor mode even when
in workspace mode. Fix that by explicitly checking that we're not in
workspace mode when deciding whether to enter vendor mode.
Fixes#54130
Change-Id: I03fcd9db4160c9872aa2b7957a80f24d49f787d0
Reviewed-on: https://go-review.googlesource.com/c/go/+/439415
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Michael Matloob <matloob@golang.org>
If we use the "pipetest" helper command instead of "sleep",
we can use its stdout pipe to determine when the process
is ready to handle a SIGSTOP, and we can additionally check
that sending a SIGCONT actually causes the process to continue.
This also allows us to remove the "sleep" helper command,
making the test file somewhat more concise.
Noticed while looking into #50138.
Change-Id: If4fdee4b1ddf28c6ed07ec3268c81b73c2600238
Reviewed-on: https://go-review.googlesource.com/c/go/+/442576
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
The linker's -X flag allows setting/changing a string variable's
content at link time. Currently it resets its size then write a
new string header pointing to the new content. This mostly works.
But under ASAN build the string variable can have larger size
than the usual 2 words, due to the red zone. Resetting the size
can cause the variable to "overlap" (in ASAN's view) with other
variables. Don't reset the size.
Fixes#56175.
Change-Id: Ib364208201a7a2fd7f44f9b1797834198736a405
Reviewed-on: https://go-review.googlesource.com/c/go/+/442635
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Adds a package level doc comment to the debug/dwarf, debug/elf,
debug/macho, debug/pe, and debug/plan9obj noting that these packages
are not designed to be hardened against adversarial inputs.
Change-Id: I678d01bcdc8ad01c23805f09cc59e64cec6c3f76
Reviewed-on: https://go-review.googlesource.com/c/go/+/435417
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: Julie Qiu <julieqiu@google.com>
Run-TryBot: Roland Shoemaker <roland@golang.org>
Auto-Submit: Roland Shoemaker <roland@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
LookupFieldOrMethod appears as a hotspot when benchmarking gopls'
auto-completion. In particular, instanceLookup.add was allocating in the
common case of structs with no embedding.
This is easily fixed, by using a small array in front of the map inside
of instanceLookup. Do this, and additionally add a microbenchmark.
The benchmark improvement is significant:
name old time/op new time/op delta
LookupFieldOrMethod-12 388µs ± 6% 154µs ± 3% -60.36% (p=0.000 n=10+10)
name old alloc/op new alloc/op delta
LookupFieldOrMethod-12 152kB ± 0% 2kB ± 0% -98.77% (p=0.000 n=9+10)
name old allocs/op new allocs/op delta
LookupFieldOrMethod-12 1.41k ± 0% 0.07k ± 0% -95.38% (p=0.000 n=10+10)
It should also be noted that instanceLookup is used elsewhere, in
particular by validType. In those contexts, the scope is not just the
current type but the entire package, and so the newly added buffer is
likely to simply cause extra Identical checks. Nevertheless, those
checks are cheap, and on balance the improved LookupFieldOrMethod
performance leads overall to improved type-checking performance.
Standard library benchmark results varied by package, but type checking
speed for many packages improved by ~5%, with allocations improved by
~10%. If this weren't the case we could let the caller control the
buffer size, but that optimization doesn't seem necessary at this time.
For example:
Check/http/funcbodies/noinfo-12 71.5ms ± 4% 67.3ms ± 2% -5.90% (p=0.000 n=20+20)
Check/http/funcbodies/noinfo-12 244k ± 0% 219k ± 0% -10.36% (p=0.000 n=19+19)
Updates golang/go#53992
Change-Id: I10b6deb3819ab562dbbe1913f12b977cf956dd50
Reviewed-on: https://go-review.googlesource.com/c/go/+/423935
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Since we already provide the error code, the prefix can be deduced
automatically.
Except for the changes in errors.go, the updates were made with
regex find-and-replaces:
check\.error\((.+), InvalidSyntaxTree, invalidAST\+ =>
check.error($1, InvalidSyntaxTree,
check\.errorf\((.+), InvalidSyntaxTree, invalidAST\+ =>
check.errorf($1, InvalidSyntaxTree,
Change-Id: Ia02fc56ac7a8524bdf0c404ff2696435408327e9
Reviewed-on: https://go-review.googlesource.com/c/go/+/441975
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This change adds an API to the runtime for arenas. A later CL can
potentially export it as an experimental API, but for now, just the
runtime implementation will suffice.
The purpose of arenas is to improve efficiency, primarily by allowing
for an application to manually free memory, thereby delaying garbage
collection. It comes with other potential performance benefits, such as
better locality, a better allocation strategy, and better handling of
interior pointers by the GC.
This implementation is based on one by danscales@google.com with a few
significant differences:
* The implementation lives entirely in the runtime (all layers).
* Arena chunks are the minimum of 8 MiB or the heap arena size. This
choice is made because in practice 64 MiB appears to be way too large
of an area for most real-world use-cases.
* Arena chunks are not unmapped, instead they're placed on an evacuation
list and when there are no pointers left pointing into them, they're
allowed to be reused.
* Reusing partially-used arena chunks no longer tries to find one used
by the same P first; it just takes the first one available.
* In order to ensure worst-case fragmentation is never worse than 25%,
only types and slice backing stores whose sizes are 1/4th the size of
a chunk or less may be used. Previously larger sizes, up to the size
of the chunk, were allowed.
* ASAN, MSAN, and the race detector are fully supported.
* Sets arena chunks to fault that were deferred at the end of mark
termination (a non-public patch once did this; I don't see a reason
not to continue that).
For #51317.
Change-Id: I83b1693a17302554cb36b6daa4e9249a81b1644f
Reviewed-on: https://go-review.googlesource.com/c/go/+/423359
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
This change makes (*mheap).sysAlloc take an explicit list of hints and a
boolean as to whether or not any newly-created heapArenas should be
registered in the full arena list.
This is a refactoring in preparation for arenas.
For #51317.
Change-Id: I0584a033fce3fcb60c5d0bc033d5fb8bd23b2378
Reviewed-on: https://go-review.googlesource.com/c/go/+/432078
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Dragonfly and FreeBSD both used numerical values for these constants
chosen to be the same as on Solaris. For some reason, NetBSD did not,
and happens to interpret value 0 as P_ALL instead of P_PID
(see 3323ceb782/sys/sys/idtype.h (L43-L44)).
Using the correct value for P_PID should cause wait6 to wait for the
correct process, which may help to avoid the deadlocks reported in
For #50138.
Updates #13987.
Change-Id: I0eacd1faee4a430d431fe48f9ccf837f49c42f39
Reviewed-on: https://go-review.googlesource.com/c/go/+/442478
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Benny Siegert <bsiegert@gmail.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Skip TestTransportPersistConnLeakShortBody in HTTP/2 mode;
it's flaky and was previously HTTP/1-only.
Don't run TestTransportEventTrace and TestTransportIgnores408
in parallel.
Change-Id: I76bc540fac9317185ef7d414c9deafb35bc926b6
Reviewed-on: https://go-review.googlesource.com/c/go/+/442495
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This brings go/types error reporting closer to types2.
Except for removing the error functions and one manual correction,
these changes were made by regex-replacing:
check\.invalidAST\((.*), " =>
check.errorf($1, InvalidSyntaxTree, invalidAST+"
check\.invalidOp\((.*), " =>
check.errorf($1, invalidOp+"
check\.invalidArg\((.*), " =>
check.errorf($1, invalidArg+"
A follow-up CL ensures that we use error instead of errorf where
possible.
Change-Id: Iac53dcd9c122b058f98d26d0fb307ef1dfe4e79b
Reviewed-on: https://go-review.googlesource.com/c/go/+/441955
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
When running go install --mod=readonly module@version. modfetch.GoSumFile
was not set, so the checksumOk check that's done when checking whether
we need to set the GoVersion from the go mod file was failing. Bypass
the checksumOk check when there's no GoSumFile.
For #54908
Change-Id: I56cf9d36a505b1223e6bf82a7d455746e2f09849
Reviewed-on: https://go-review.googlesource.com/c/go/+/439855
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
As of CL 438347, multiple concurrents calls to Close should be safe.
This removes some indirection and may also make some programs that use
type-assertions marginally more efficient. For example, if a program
calls (*exec.Cmd).StdinPipe to obtain a pipe and then sets that as the
Stdout of another command, that program will now allow the second
command to inherit the file descriptor directly instead of copying
everything through a goroutine.
This will also cause calls to Close after the first to return an error
wrapping os.ErrClosed instead of nil. However, it seems unlikely that
programs will depend on that error behavior: if a program is calling
Write in a loop followed by Close, then if a racing Close occurs it is
likely that the Write would have already reported an error. (The only
programs likely to notice a change are those that call Close — without
Write! — after a call to Wait.)
Updates #56043.
Updates #9307.
Updates #6270.
Change-Id: Iec734b23acefcc7e7ad0c8bc720085bc45988efb
Reviewed-on: https://go-review.googlesource.com/c/go/+/439195
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Convert SLT/SLTU with a suitably valued constant into a SLTI/SLTIU instruction.
This can reduce instructions and avoid register loads. Now that we generate
more SLTI/SLTIU instructions, absorb these into branches when it makes sense
to do so.
Removes more than 800 instructions from the Go binary on linux/riscv64.
Change-Id: I42c4e00486697acd4da7669d441b5690795f18ae
Reviewed-on: https://go-review.googlesource.com/c/go/+/428499
Reviewed-by: Wayne Zuo <wdvxdr@golangcn.org>
Run-TryBot: Joel Sing <joel@sing.id.au>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Joedian Reid <joedian@golang.org>
CL 403995 fixed static init of literal contains dynamic exprs, by
ensuring their init are ordered properly. However, we still need to walk
the generated init codes before appending to parent init. Otherwise,
codes that requires desugaring will be unhandled, causing the compiler
backend crashing.
Fixes#56105
Change-Id: Ic25fd4017473f5412c8e960a91467797a234edfd
Reviewed-on: https://go-review.googlesource.com/c/go/+/440455
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@golang.org>
Use InvalidSyntaxError where the zero error code was used before.
Fix a couple of places that didn't set an error code.
Panic in error reporting if no error code is provided.
Change-Id: I3a537d42b720deb5c351bf38871e04919325e231
Reviewed-on: https://go-review.googlesource.com/c/go/+/439566
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Type checkers should use InvalidSyntaxTree as error code
for invalid syntax tree errors instead of zero. This way
the zero value can be used to mark an unset error code.
Also, add an example for BlankPkgName (and adjust the
test harness slightly to make it work).
Change-Id: Ic15fa0e8e46be698e52352f2f0e4915b75e509d7
Reviewed-on: https://go-review.googlesource.com/c/go/+/439565
Reviewed-by: Robert Findley <rfindley@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The internal/types/errors package defines all error codes used by
the type checkers. This is the 1st step of several that factor out
the error codes from go/types and types2; the package is not yet
used.
- The file codes.go is a copy of go/types/errorcodes.go. The
only change is the updated package name (types -> errors).
- The file codes_test.go is a copy of go/types/errorcodes_test.go
with updated package name (types_test -> errors_test) and minor
changes to walkCodes so that it doesn't require the pkgFiles
helper function (the test only parses a single file).
Change-Id: Idb977b1220737b56b330de1d977f698f022daafc
Reviewed-on: https://go-review.googlesource.com/c/go/+/439560
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
If a generic type declaration is missing a constraint, syntactically
it is an array type declaration with an undefined array length.
Mention the possibility of a missing constraint in the error message
for the undefined array length.
For #56064.
For #55961.
For #51145.
Change-Id: Ic161aeda9ea44faa8aa3bf3e9d62b3b13a95d4c5
Reviewed-on: https://go-review.googlesource.com/c/go/+/439559
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Add "-O2" to all compiler/linker tests. This makes compiler/linker
feature probing better resemble actual compiling later.
Why?
----
zig c++ is a clang front-end[1] that accepts, among other things, the
target over the command line. This command:
zig c++ -target x86_64-linux-gnu main.o -o main
Will:
1. Pre-compile libc++.a.
2. Link the program with libc++.a from (1).
Currently Go only is learning about one flag from the linker, that is,
"--no-gc-sections". The resulting command that tests for the flag
support is this:
c++ -Wl,--no-gc-sections -x c - -o
This causes Zig to pre-compile libc++.a in debug mode. Then the actual
compiler+linker command from CGo adds a few more flags, including "-O2":
c++ <...> -Wl,--no-gc-sections -O2 <...>
From Zig perspective, debug-mode libc++.a is different from the
optimized one; that causes Zig to compile a new libc++.a. Specifically,
Zig adds "-fno-omit-frame-pointer" for debug builds, and
"-fomit-frame-pointer" for optimized builds.
As a result, we have to two sets of libc++.a for every arch/os tuple.
That takes CPU time and a bit of disk space.
Zig performance impact
----------------------
First compilation of a simple CGo application is faster by ~2.5 seconds
or ~60%:
$ CC="zig c++ -target x86_64-linux-gnu.2.28" hyperfine \
--warmup 3 --runs 10 \
--prepare 'rm -fr ~/.cache/zig ~/.cache/go-build /tmp/go-*' \
--parameter-list go go1.19,go1.19-O2 \
'/code/go/bin/{go} build .'
Benchmark 1: /code/go/bin/go1.19 build .
Time (mean ± σ): 6.168 s ± 0.059 s [User: 7.465 s, System: 1.578 s]
Range (min … max): 6.111 s … 6.242 s 10 runs
Benchmark 2: /code/go/bin/go1.19-O2 build .
Time (mean ± σ): 3.816 s ± 0.080 s [User: 4.730 s, System: 1.130 s]
Range (min … max): 3.657 s … 3.958 s 10 runs
Summary
'/code/go/bin/go1.19-O2 build .' ran
1.62 ± 0.04 times faster than '/code/go/bin/go1.19 build .'
If we add C++ to the mix, the difference grows to almost ~23 seconds, or
almost 90%:
$ CC="zig c++ -target x86_64-linux-gnu.2.28" hyperfine \
--warmup 1 --runs 3 \
--prepare 'rm -fr ~/.cache/zig ~/.cache/go-build /tmp/go-*' \
--parameter-list go go1.19,go1.19-O2 \
'/code/go/bin/{go} build .'
Benchmark 1: CC="zig c++ -target x86_64-linux-gnu.2.28" /code/go/bin/go1.19 build .
Time (mean ± σ): 51.137 s ± 0.183 s [User: 234.165 s, System: 15.005 s]
Range (min … max): 50.934 s … 51.288 s 3 runs
Benchmark 2: CC="zig c++ -target x86_64-linux-gnu.2.28" /code/go/bin/go1.19-O2 build .
Time (mean ± σ): 27.102 s ± 0.068 s [User: 119.816 s, System: 8.513 s]
Range (min … max): 27.038 s … 27.174 s 3 runs
Summary
'/code/go/bin/go1.19-O2 build .' ran
1.89 ± 0.01 times faster than '/code/go/bin/go1.19 build .'
The difference is just due to the fact that Zig will not be instructed
to compile libc++.a for debug builds; Go doesn't need that.
Non-Zig performance impact
--------------------------
A.k.a. does "-O2" for this check worsen performance?
No statistically significant performance differences with both clang-15
and gcc-11. Also, it affects only the first compile of a CGo progam, as
the linker tests are cached across invocations. go1.19 binary is the
go1.19 tag; go1.19-O2 is go1.19 + this patch.
$ hyperfine --warmup 3 --runs 20 \
--prepare 'rm -fr ~/.cache/go-build/ /tmp/go-*' \
--parameter-list go go1.19,go1.19-O2 \
--parameter-list cc gcc-11,clang-15 \
'CC={cc} /code/go/bin/{go} build .'
Benchmark 1: CC=gcc-11 /code/go/bin/go1.19 build .
Time (mean ± σ): 681.1 ms ± 13.7 ms [User: 501.6 ms, System: 247.1 ms]
Range (min … max): 654.1 ms … 707.2 ms 20 runs
Benchmark 2: CC=gcc-11 /code/go/bin/go1.19-O2 build .
Time (mean ± σ): 676.8 ms ± 10.2 ms [User: 500.4 ms, System: 245.6 ms]
Range (min … max): 664.4 ms … 696.4 ms 20 runs
Benchmark 3: CC=clang-15 /code/go/bin/go1.19 build .
Time (mean ± σ): 860.1 ms ± 17.1 ms [User: 530.0 ms, System: 394.9 ms]
Range (min … max): 839.4 ms … 920.0 ms 20 runs
Benchmark 4: CC=clang-15 /code/go/bin/go1.19-O2 build .
Time (mean ± σ): 864.5 ms ± 26.6 ms [User: 537.8 ms, System: 390.1 ms]
Range (min … max): 841.9 ms … 955.5 ms 20 runs
Summary
'CC=gcc-11 /code/go/bin/go1.19-O2 build .' ran
1.01 ± 0.03 times faster than 'CC=gcc-11 /code/go/bin/go1.19 build .'
1.27 ± 0.03 times faster than 'CC=clang-15 /code/go/bin/go1.19 build .'
1.28 ± 0.04 times faster than 'CC=clang-15 /code/go/bin/go1.19-O2 build .'
cgo.go
------
package main
// #define _FILE_OFFSET_BITS 64
// #include <unistd.h>
// #include <fcntl.h>
// #include <stdio.h>
// char* hello() { return "hello, world"; }
// void phello() { printf("%s, your lucky number is %p\n", hello(), fcntl); }
import "C"
func main() {
C.phello()
}
func Chello() string {
return C.GoString(C.hello())
}
Alternatives considered
-----------------------
There are a few alternatives:
1. Add "-O2" for linker-only tests. That looks like too much catering to
zig alone. If we can add it, then add for everything.
2. Add "-fomit-frame-pointer" instead of "-O2". This flag does not
universally imply debug mode, thus same argument applies as to (1).
3. Add "-O2" for this particular test (`--no-gc-sections`). This is
brittle and not future-proof: a future linker test may omit this
flag.
Hardware
--------
Tested on a 4-core (8 HT) Intel(R) Core(TM) i7-8665U CPU on Debian 11,
Linux 5.10.0-15-amd64.
[1]: https://andrewkelley.me/post/zig-cc-powerful-drop-in-replacement-gcc-clang.html
Change-Id: I5223a5cf53fc5d2b77ac94a6c5712c32c7fbdf36
GitHub-Last-Rev: 2e998b831a
GitHub-Pull-Request: golang/go#55966
Reviewed-on: https://go-review.googlesource.com/c/go/+/436884
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run "go build ./x" in this workspace:
-- go.work --
use ./y
-- x/go.mod --
module x
go 1.19
-- x/m.go --
package m
It fails with: "go: open /tmp/foo/y/go.mod: no such file or directory".
It's unclear where the name "y" comes from.
This change will emit error like: "go: cannot load module listed in
go.work file: open /tmp/foo/y/go.mod: no such file or directory"
Fixes#55952.
Change-Id: Ia45dd915e3fbd6e33340f352b3d6235c6c31190b
GitHub-Last-Rev: 410de1b4a7
GitHub-Pull-Request: golang/go#56050
Reviewed-on: https://go-review.googlesource.com/c/go/+/438147
Run-TryBot: hopehook <hopehook@golangcn.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
This permits us to safely support concurrent access to files on Plan 9.
Concurrent access was already safe on other systems.
This does introduce a change: if one goroutine calls a blocking read
on a pipe, and another goroutine closes the pipe, then before this CL
the close would occur. Now the close will be delayed until the blocking
read completes.
Also add tests that concurrent I/O and Close on a pipe are OK.
For #50436
For #56043
Change-Id: I969c869ea3b8c5c2f2ef319e441a56a3c64e7bf5
Reviewed-on: https://go-review.googlesource.com/c/go/+/438347
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: David du Colombier <0intro@gmail.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Rob Pike <r@golang.org>
An older version of go compiling a main module that references a
standard library package from a newer release (e.g. net/netip added in
go 1.18) currently produces a confusing error message. This changes adds
a new error message including go version diagnostics.
Fixes#48966
Change-Id: I1e8319dafcf1f67d1b1ca869fe84190c3b3f3c3e
Reviewed-on: https://go-review.googlesource.com/c/go/+/432075
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This change moves Linux epoll's syscalls implementation to the
"runtime/internal/syscall" package. The intention in this CL was to
minimise behavioural changes but make the code more generalised. This
also will allow adding new syscalls (like epoll_pwait2) without the
need to implement assembly stubs for each arch.
It also drops epoll_create as not all architectures provide this call.
epoll_create1 was added to the kernel in version 2.6.27 and Go requires
Linux kernel version 2.6.32 or later since Go 1.18. So it is safe to
always use epoll_create1.
This is a resubmit as the previous CL 421994 was reverted due to test
failures after the merge with the master. The issue was fixed in
CL 438615
For #53824
For #51087
Change-Id: I1bd0f23a85b4f9b80178c5dd36fd3e95ff4f9648
Reviewed-on: https://go-review.googlesource.com/c/go/+/440115
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Michael Pratt <mpratt@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
This commit was dedicated to adding an example of using B.ReportMetrics
with B.RunParallel called ExampleB_ReportMetric_parallel. In this
example, the same algorithm for ExampleB_ReportMetric was used, instead
with a concurrent for loop using PB.Next instead of a standard one.
There is also notes noting when to use the B.ReportMetric methods when
running concurrent testing.
Fixes#50756
Change-Id: I2a621b4e367af5f4ec47d38a0da1035a8d52f628
Reviewed-on: https://go-review.googlesource.com/c/go/+/437815
Reviewed-by: Carlos Amedee <carlos@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Replace the ad-hoc approach to running tests in HTTP/1 and HTTP/2
modes with a 'run' function that executes a test in various modes.
By default, these modes are HTTP/1 and HTTP/2, but tests can
opt-in to HTTPS/1 as well.
The 'run' function also takes care of post-test cleanup (running the
afterTest function).
The 'run' function runs tests in parallel by default. Tests which
can't run in parallel (generally because they use global test hooks)
pass a testNotParallel option to disable parallelism.
Update clientServerTest to use t.Cleanup to clean up after itself,
rather than leaving this up to tests to handle.
Drop an unnecessary mutex in SetReadLoopBeforeNextReadHook.
Test hooks can't be set in parallel, and we want the race detector
to notify us if two simultaneous tests try to set a hook.
Fixes#56032
Change-Id: I16be64913c426fc93d84abc6ad85dbd3bc191224
Reviewed-on: https://go-review.googlesource.com/c/go/+/438137
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Move the writev definition for solaris from package
internal/syscall/unix to package syscall. This corresponds to where
writev is defined on aix, darwin and openbsd as well and is
go:linkname'ed from internal/poll. This also allows updating the
generated wrappers more easily if needed.
Change-Id: I671ed8232d25319f8e63f549f786d77a17602148
Reviewed-on: https://go-review.googlesource.com/c/go/+/436597
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Carlos Amedee <carlos@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
This cuts the wall duration for 'go test os/exec' and
'go test -race os/exec' roughly in half on my machine,
which is an even more significant speedup with a high '-count'.
For better or for worse, it may also increase the repro rate
of #34988.
Tests that use Setenv or Chdir or check for FDs opened during the test
still cannot be parallelized, but they are only a few of those.
Change-Id: I8d284d8bff05787853f825ef144aeb7a4126847f
Reviewed-on: https://go-review.googlesource.com/c/go/+/439196
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
This test has been disabled for over nine years (since CL 12869049).
Although it still compiles, it seems likely to have rotted since then,
and if it was going to detect a real bug it also seems like that bug
would have been encountered and reported by users since then (and
would presumably have its own regression tests).
To eliminate overhead from mainining it (or skipping over it while
maintaining other tests), let's just delete it.
Fixes#5780.
Change-Id: I2a85cba20cba98a1dc6fc82336ae5e22d2242e99
Reviewed-on: https://go-review.googlesource.com/c/go/+/439197
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
For example:
movb a0, a0
srai $1, a0, a0
the assembler will expand to:
slli $56, a0, a0
srai $56, a0, a0
srai $1, a0, a0
this CL optimize to:
slli $56, a0, a0
srai $57, a0, a0
Remove 270+ instructions from Go binary on linux/riscv64.
Change-Id: I375e19f9d3bd54f2781791d8cbe5970191297dc8
Reviewed-on: https://go-review.googlesource.com/c/go/+/428496
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org>
Reviewed-by: Joel Sing <joel@sing.id.au>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Set a 128 MB limit on the amount of space used by []syntax.Inst
in the compiled form corresponding to a given regexp.
Also set a 128 MB limit on the rune storage in the *syntax.Regexp
tree itself.
Thanks to Adam Korczynski (ADA Logics) and OSS-Fuzz for reporting this issue.
Fixes CVE-2022-41715.
Fixes#55949.
Change-Id: Ia656baed81564436368cf950e1c5409752f28e1b
Reviewed-on: https://go-review.googlesource.com/c/go/+/439356
Auto-Submit: Roland Shoemaker <roland@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Roland Shoemaker <roland@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
We were saving non-go file information in the module index files,
leading in an unnecessary increase in memory usage in modules
containing many non-go files. This was a bug because this information
is never used. Don't save that information.
For #54226
Change-Id: I0644064f83f96e3a9f43b7e66ca94d69d9603376
Reviewed-on: https://go-review.googlesource.com/c/go/+/439118
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
Run-TryBot: Michael Matloob <matloob@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Use shorter more Go-like names for the new APIs being added in the
runtime/coverage package for writing coverage data under user control
from server programs. Old names were a bit too clunky/verbose.
Updates #51430.
Change-Id: Ifdd5b882a88613c7c4342b40ed93b58547483c77
Reviewed-on: https://go-review.googlesource.com/c/go/+/438503
Reviewed-by: Russ Cox <rsc@golang.org>
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The traceback code has special "jump stack" logic, to trace back
stack switches through systemstack. If we're at the entry of
systemstack, the stack switch hasn't happened, so don't jump to
user stack.
The jump stack logic is only used if we're on the g0 stack. It can
happen that we're at the entry of a recursive systemstack call on
the g0 stack. In we jump stack here, there will be two problems:
1. There are frames between entering the g0 stack and this
recursive systemstack call. Those frames will be lost.
2. Worse, we switched frame.sp but frame.fp calculation will use
the entry SP delta (0), which will be wrong, which in turn
leads wrong frame.lr and things will go off.
For now, don't jump stack if we're at entry of systemstack (SP
delta is 0).
Using a per-PC SPWRITE marker may be a better fix. If we haven't
written the SP, we haven't switched the stack so we can just
unwind like a normal function.
May fix#55851.
Change-Id: I2b624c8c086b235b34d9c7d3cebd4a37264f00f8
Reviewed-on: https://go-review.googlesource.com/c/go/+/437299
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
On Mac OS X, the default stack size for non-main threads created by cgo is
fixed at 512KB and cannot be altered by setrlimit. This stack size is too
small for some recursive scenarios. We can solve this problem by explicitly
copying the stack size of the main thread when creating a new thread.
Change-Id: I400d5b2e929a1ee261502914a991e208759f64a8
GitHub-Last-Rev: b29c74599e
GitHub-Pull-Request: golang/go#53667
Reviewed-on: https://go-review.googlesource.com/c/go/+/415915
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: hopehook <hopehook@golangcn.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Imagine that initFuncTypes is called with n=3, funcTypes will be
[nil, nil, nil, **reflect.rtype] afterward, then it's called with n=2.
The current implementation will copy funcTypes because funcTypes[2] is
nil. This is unnecessary. It should make a new slice and copy funcTypes
into it only when n >= len(funcTypes).
Updates #56011.
Change-Id: Ia093d2f550d6924a4c58bcd21325093e32b40baa
GitHub-Last-Rev: a599eae7c2
GitHub-Pull-Request: golang/go#56024
Reviewed-on: https://go-review.googlesource.com/c/go/+/438395
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
This reverts CL 437176.
Reason for revert: broke programs that plumb StdoutPipe from one command to Stdin on another and then call Wait on the former.
os/exec itself uses a type-assertion to *os.File to determine whether to copy stdin using a goroutine or just pass a file descriptor. An early Wait using a *os.File is benign (because closing the pipe doesn't close the child's inherited file descriptor), but an early Wait using a non-*os.File is not.
Updates #50436.
Change-Id: I4a2993e290982834f91696d890dfe77364c0cc50
Reviewed-on: https://go-review.googlesource.com/c/go/+/438695
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
[Roll-forward of CL 436915 by Tobias Klauser, with builtin and gen
directories dropped now that they've been handled separately.]
The minimum bootstrap version for Go ≥ 1.20 is Go 1.17. That version
supports the new style //go:build lines. Thus the old style //+build
lines can be dropped in this part of the tree as well. Leave the
//+build lines in cmd/dist which will ensure the minimum Go version
during bootstrap.
As suggested by Cherry during review of CL 430496
For #44505
Change-Id: Ifa686656c3e50bf7f92f70747b44d74a7d51bad8
Reviewed-on: https://go-review.googlesource.com/c/go/+/435473
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
These two directories are full of //go:build ignore files.
We can ignore them more easily by putting an underscore
at the start of the name. That also works around a bug
in Go 1.17 that was not fixed until Go 1.17.3.
Change-Id: Ia5389b65c79b1e6d08e4fef374d335d776d44ead
Reviewed-on: https://go-review.googlesource.com/c/go/+/435472
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Read counters using atomic ops so as to avoid problems with the race
detector if a goroutine happens to still be executing at the end of a
test run when we're writing out counter data. In theory we could guard
the atomic use on the counter mode, but it's better just to do it in
all cases, leaves us with a simpler implementation.
Fixes#56006.
Change-Id: I81c2234b5a1c3b00cff6c77daf2c2315451b7f6c
Reviewed-on: https://go-review.googlesource.com/c/go/+/438256
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
Separate out the functions from cmd/internal/sys/support.go and
migrate them to a new package internal/platform, so that functions such as
"RaceDetectorSupported" can be called from tests in std as well as in
cmd. This isn't a complete move of everything in cmd/internal/sys;
there are still many functions left.
The original version of this CL (patch set 1) called the new package
"internal/sys", but for packages that needed both "internal/sys" and
"cmd/internal/sys" the import of the former had to be done with a
different name, which was confusing and also required a hack in
cmd/dist.
Updates #56006.
Change-Id: I866d62e75adbf3a640a06e2c7386a6e9e2a18d91
Reviewed-on: https://go-review.googlesource.com/c/go/+/438475
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
Update the minimum version required for asan to be gcc9.
This will ensure that go build -asan is supported only on
systems with the required version of gcc. Update the asan
error message to include the name of the compiler (the
error message is updated to include the name of the compiler
instead of C compiler for other platforms as well).
Related to CL 425355
Change-Id: Ib864d43b2b3028f39ffcf792890a678361f507f6
Reviewed-on: https://go-review.googlesource.com/c/go/+/436740
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Archana Ravindar <aravind5@in.ibm.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
SecCreatePolicySSL returns null when called from a binary that has a
strange path. This seems to be a weirdo macos bug, but we should be
properly handling those null returns anyway. Also add handling for
SecTrustGetCertificateAtIndex.
Fixes#54590
Change-Id: I251e74f3b0bf65890a80b094b3e88718e13fd3db
Reviewed-on: https://go-review.googlesource.com/c/go/+/438135
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Roland Shoemaker <roland@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
Auto-Submit: Roland Shoemaker <roland@golang.org>
When decoding a struct, if a positive delta is large enough to overflow
when added to fieldnum, we would panic due to the resulting negative index.
Instead, catch this problem and produce an error like we do with
negative delta integers. If fieldnum ends up being negative or smaller
than state.fieldnum, the addition overflowed.
While here, remove an unnecessary break after an error call,
since those error functions cause a panic.
Fixes#55337.
Change-Id: I7a0e4f43e5c81a703e79c1597e3bb3714cc832c8
Reviewed-on: https://go-review.googlesource.com/c/go/+/432715
Reviewed-by: Rob Pike <r@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
TryBot-Result: Gopher Robot <gobot@golang.org>
CL 425314 made creating funcTypes using StructOf, and using a mutex to
protect read+write to funcTypes. However, after initializing funcTypes,
it is accessed in FuncOf without holding lock, causing a race.
Fixing it by returning the n-th Type directly from initFuncTypes, so the
accessing funcTypes will always be guarded by a mutex.
Fixes#56011
Change-Id: I1b50d1ae342943f16f368b8606f2614076dc90fb
Reviewed-on: https://go-review.googlesource.com/c/go/+/437997
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
In TestCoverageApis/emitToNonexistentDir there is a list of error
messages to match when a nonexistent directory is opened. The list
has message text only for Unix and Windows. Add the corresponding
text for Plan 9.
Fixes#55983
Change-Id: Id32130300cb02394b319e1aeb1229ee147b4afb2
Reviewed-on: https://go-review.googlesource.com/c/go/+/437557
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: David du Colombier <0intro@gmail.com>
Run-TryBot: David du Colombier <0intro@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
The flag is now removed from `go build` and `go test`.
It has been deprecated since Go 1.16, printing a warning message.
The idea was to fully delete it in Go 1.17, but that didn't happen.
First, delete the BuildI variable and its flag declarations,
as well as all the bits of docs that mentioned the flag.
Second, delete or simplify the code paths that used BuildI.
Third, adapt the tests to the removed flag.
Some of them are removed, like test_relative_import_dash_i.txt and
TestGoTestDashIDashOWritesBinary, as they centered around the flag.
The rest are modified to not cover or use the flag.
Finally, change cmd/dist to no longer use `go install -i`.
The purpose of the flag was that, when bootstrapping the toolchain,
all of its dependencies would also be installed as object files.
When removing the use of the -i flags, the checkNotStale call right
after building toolchain3 would fail as expected,
because runtime/internal/sys is now only up to date in the build cache.
Luckily, that's not a problem: we run `go install std cmd` right after,
so all standard library packages will be installed as object files.
Move the checkNotStale call after that install command.
Fixes#41696.
Change-Id: I5d8139f18aaee07da886d483e663f3f2f00d5f3a
Reviewed-on: https://go-review.googlesource.com/c/go/+/416094
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
TryBot-Result: Gopher Robot <gobot@golang.org>
It's all local to a single file and responsible for 1.7% of total
space allocated summed over compilation of the bent benchmarks.
Showing nodes accounting for 27.16GB, 27.04% of 100.44GB total
Dropped 1622 nodes (cum <= 0.50GB)
Showing top 10 nodes out of 321
flat flat% sum% cum cum%
4.87GB 4.85% 4.85% 4.87GB 4.85% cmd/compile/internal/objw.init
4.79GB 4.77% 9.62% 4.81GB 4.79% runtime.allocm
4.72GB 4.70% 14.32% 4.72GB 4.70% cmd/compile/internal/types.newType
3.10GB 3.09% 17.41% 3.17GB 3.15% cmd/compile/internal/ssagen.InitConfig
1.86GB 1.85% 19.26% 2.61GB 2.60% cmd/compile/internal/ssa.cse
1.72GB 1.71% 20.97% 2.25GB 2.24% cmd/internal/obj.(*Link).traverseFuncAux
1.66GB 1.65% 22.62% 1.66GB 1.65% runtime.malg
1.61GB 1.61% 24.23% 1.64GB 1.63% cmd/compile/internal/ssa.schedule
1.42GB 1.41% 25.64% 1.42GB 1.41% cmd/compile/internal/ir.NewNameAt
Change-Id: Id18ee3b83cb23a7042d59714a0c1ca074e7bc7a8
Reviewed-on: https://go-review.googlesource.com/c/go/+/437297
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: David Chase <drchase@google.com>
The mismatch between Unified IR and the old frontend is not about how
they number the closures, but how they name them. For nested closure,
the old frontend use the immediate function which contains the closure
as the outer function, while Unified IR uses the outer most function as
the outer for all closures.
That said, what important is matching the number of closures, not their
name prefix. So this CL relax the test to match both "main.func1.func2"
and "main.func1.2" to satisfy both Unified IR and the old frontend.
Updates #53058
Change-Id: I66ed816d1968aa68dd3089a4ea5850ba30afd75b
Reviewed-on: https://go-review.googlesource.com/c/go/+/437216
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This CL optimizes the Hypot function by putting the Abs function in
front of the IsInf check. This simplifies the judgment of IsInf.
Benchmarks:
On linux/arm64,
name old time/op new time/op delta
Hypot-160 5.26ns ± 0% 4.53ns ± 0% -13.84% (p=0.000 n=4+5)
HypotGo-160 5.19ns ± 0% 4.85ns ± 0% -6.53% (p=0.008 n=5+5)
On linux/amd64,
name old time/op new time/op delta
Hypot-44 5.99ns ± 0% 5.99ns ± 0% ~ (p=0.667 n=5+5)
HypotGo-44 7.46ns ± 0% 6.61ns ± 0% -11.37% (p=0.008 n=5+5)
On darwin/arm64,
name old time/op new time/op delta
Hypot-8 3.58ns ± 0% 2.79ns ± 0% -22.01% (p=0.008 n=5+5)
HypotGo-8 3.58ns ± 0% 2.79ns ± 0% -22.15% (p=0.008 n=5+5)
Change-Id: Id79236e01d9494b6e00bbda3ec08c72caf5ef3c1
Reviewed-on: https://go-review.googlesource.com/c/go/+/414974
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Eric Fang <eric.fang@arm.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
For #50436, I want to be able to close the pipes returned by
StdoutPipe and StderrPipe after the Context has been canceled
and the WaitDelay has subsequently expired.
However, the fact that the exec.onceCloser wrapper for StdinPipe
(added in CL 13329043) was retained in CL 65490 suggests to me that
(*os.File).Close is still not safe to call concurrently.
This may cause type assertions of these ReadClosers to *os.File that
once succeeded to no longer do so. However, the StdoutPipe and
StderrPipe methods return interfaces, not concrete *os.Files, so
callers already should not have been relying on that implementation
detail — and as far as I can tell the closeOnce wrapper does not mask
any (*os.File) methods, so assertions to any interface type that
previously succeeded will continue to do so.
This change is logically part of CL 401835, but since it may expose
fragile type-assertions in callers I want to keep it separate for
clearer bisection of any new test failures.
For #50436.
Change-Id: I58de1d48fb6fd788502f13657d8d4484516271cf
Reviewed-on: https://go-review.googlesource.com/c/go/+/437176
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This change undoes CL 401894, because on further consideration
it turns out not to be needed.
This also makes (*Cmd).closeDescriptors a free function, since it does
not actually use the receiver in any way and is not needed to satisfy
any interfaces.
For #50436.
Change-Id: I7915265b0e6398ed5a34fae0c12873ab08a14194
Reviewed-on: https://go-review.googlesource.com/c/go/+/437175
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
When the first call to Do finished, it calls c.wg.Done() to signal
others that the call was done. However, that happens without holding
a lock, so if others call to Do complete and be followed by a call to
ForgotUnshared, that then returns false.
Fixing this by moving c.wg.Done() inside the section guarded by g.mu, so
the two operations won't be interrupted.
Thanks bcmills@ for finding and suggesting fix.
Change-Id: I850f5eb3f9751a0aaa65624d4109aeeb59dee42c
Reviewed-on: https://go-review.googlesource.com/c/go/+/436437
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
All inlined functions are Go functions, and thus should be capable of
having a FuncInfo. Missing FuncInfo is likely indication of a compiler
bug that dropped the symbol too early, failing to add it to the symbol
list used for writing output. I believe all existing cases have been
fixed; this check will prevent regressions.
The exception is -linkshared mode. There symbols are loaded from the
shared library, and the FuncInfo is not available. This is a bug, as it
can result in incorrect the FuncID in inlinedCall, but it is very
involved to fix.
For #54959.
For #55954.
Change-Id: Ib0dc4f1ea62525b55f68604d6013ff33223fdcdd
Reviewed-on: https://go-review.googlesource.com/c/go/+/429637
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
The linker needs FuncInfo metadata for all inlined functions. This is
typically handled by gc.enqueueFunc calling ir.InitLSym for all function
declarations in typecheck.Target.Decls (ir.UseClosure adds all closures
to Decls).
However, non-trivial closures in Decls are ignored, and are insteaded
enqueued when walk of the calling function discovers them.
This presents a problem for direct calls to closures. Inlining will
replace the entire closure definition with its body, which hides the
closure from walk and thus suppresses symbol creation.
Explicitly create a symbol early in this edge case to ensure we keep
this metadata.
InitLSym needs to move out of ssagen to avoid a circular dependency (it
doesn't have anything to do with ssa anyway). There isn't a great place
for it, so I placed it in ir, which seemed least objectionable.
The added test triggers one of these inlined direct non-trivial closure
calls, though the test needs CL 429637 to fail, which adds a FuncInfo
assertion to the linker. Note that the test must use "run" instead of
"compile" since the assertion is in the linker, and "compiler" doesn't
run the linker.
Fixes#54959.
Change-Id: I0bd1db4f3539a78da260934cd968372b7aa92546
Reviewed-on: https://go-review.googlesource.com/c/go/+/436240
Run-TryBot: Michael Pratt <mpratt@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This change moves Linux epoll's syscalls implementation to the
"runtime/internal/syscall" package. The intention in this CL was to
minimise behavioural changes but make the code more generalised. This
also will allow adding new syscalls (like epoll_pwait2) without the
need to implement assembly stubs for each arch.
It also drops epoll_create as not all architectures provide this call.
epoll_create1 was added to the kernel in version 2.6.27 and Go requires
Linux kernel version 2.6.32 or later since Go 1.18. So it is safe to
always use epoll_create1.
For #53824
For #51087
Change-Id: I9a6a26b7f2075a38e041de1bab4691da0ecb94fc
Reviewed-on: https://go-review.googlesource.com/c/go/+/421994
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
This reverts commit 6616573982, corresponding to CL 436915.
Reason for revert: this is causing some bootstrap build problems with older versions of Go 1.17, as I understand it. Still under investigation.
Change-Id: Idb6e17ff7b47004cbf87f967af6d84f214d8abb5
Reviewed-on: https://go-review.googlesource.com/c/go/+/435471
Reviewed-by: David Chase <drchase@google.com>
The minimum bootstrap version for Go ≥ 1.20 is Go 1.17. That version
supports the new style //go:build lines. Thus the old style //+build
lines can be dropped in this part of the tree as well. Leave the
//+build lines in cmd/dist which will ensure the minimum Go version
during bootstrap.
As suggested by Cherry during review of CL 430496
For #44505
Change-Id: If53c0b02cacbfb055a33e73cfd38578dfd3aa340
Reviewed-on: https://go-review.googlesource.com/c/go/+/436915
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Than McIntosh <thanm@google.com>
An error which implements an "Unwrap() []error" method wraps all the
non-nil errors in the returned []error.
We replace the concept of the "error chain" inspected by errors.Is
and errors.As with the "error tree". Is and As perform a pre-order,
depth-first traversal of an error's tree. As returns the first
matching result, if any.
The new errors.Join function returns an error wrapping a list of errors.
The fmt.Errorf function now supports multiple instances of the %w verb.
For #53435.
Change-Id: Ib7402e70b68e28af8f201d2b66bd8e87ccfb5283
Reviewed-on: https://go-review.googlesource.com/c/go/+/432898
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Rob Pike <r@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Joseph Tsai <joetsai@digital-static.net>
When the compiler crashes, it is not uncommon to see many hundreds
of goroutines all blocked waiting their turn to be one of the nWorkers
goroutines that is allowed to run. All these goroutine stacks are not a
terribly efficient use of memory, and they also make the crash dumps
hard to read.
Introduce a manager goroutine to hand out work to at most nWorker
goroutines, maintaining pending work in a local slice, rather than
having all those blocked goroutines hanging around waiting to run.
Change-Id: I46cb4e1afd6392805f359e14554ebc17d538bcba
Reviewed-on: https://go-review.googlesource.com/c/go/+/431956
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Rework the mechanism for passing a list of output files from cmd/go to
cmd/cover when running new-style package-scope coverage
instrumentation (-pkgcfg mode). The old scheme passed a single string
with all output files joined together with os.PathListSeparator, but
this scheme is not viable on plan9, where strings containing the
separator character are not permitted when running exec.Command().
Instead, switch cmd/cover to use an arguments file (a file containing
a list of names) to specify names of instrumented output files. This
fixes the cmd/cover test failures on the plan9 builders.
Updates #51430.
Change-Id: I919f5e0a79500e28648fb9177225a9b938e4fdee
Reviewed-on: https://go-review.googlesource.com/c/go/+/436675
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
If a cycle has length 1, don't enumerate the single cycle entry;
instead just mention "refers to itself". For instance, for an
invalid recursive type T we now report:
invalid recursive type: T refers to itself
instead of:
invalid recursive type T
T refers to
T
Adjust tests to check for the different error messages.
Change-Id: I5bd46f62fac0cf167f0d0c9a55f952981d294ff4
Reviewed-on: https://go-review.googlesource.com/c/go/+/436295
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Robert Griesemer <gri@google.com>
Add hooks/apis to support writing of coverage counter data and
meta-data under user control (from within an executing "-cover"
binary), so as to provide a way to obtain coverage data from programs
that do not terminate. This patch also adds a hook for clearing the
coverage counter data for a running program, something that can be
helpful when the intent is to capture coverage info from a specific
window of program execution.
Updates #51430.
Change-Id: I34ee6cee52e5597fa3698b8b04f1b34a2a2a418f
Reviewed-on: https://go-review.googlesource.com/c/go/+/401236
Reviewed-by: David Chase <drchase@google.com>
Some of the unit tests in Go's "cmd" tree wind up building a separate
copy of the tool being tested, then exercise the freshly built tool as
a way of doing regression tests. The intent is to make sure that "go
test" is testing the current state of the source code, as opposed to
whatever happened to be current when "go install <tool>" was last run.
Doing things this way is unfriendly for coverage testing. If I run "go
test -cover cmd/mumble", and the cmd/mumble test harness builds a
fresh copy of mumble.exe, any runs of that new executable won't
generate coverage data.
This patch updates the test harnesses to use the unit test executable
as a stand-in for the tool itself, so that if "go test -cover" is in
effect, we get the effect of building the tool executable for coverage
as well. Doing this brings up the overall test coverage number for
cmd/cover quite dramatically:
before change:
$ go test -cover .
ok cmd/cover 1.100s coverage: 1.5% of statements
after change:
$ go test -cover .
ok cmd/cover 1.299s coverage: 84.2% of statements
Getting this to work requires a small change in the Go command as
well, to set GOCOVERDIR prior to executing a test binary.
Updates #51430.
Change-Id: Ifcf0ea85773b80fcda794aae3702403ec8e0b733
Reviewed-on: https://go-review.googlesource.com/c/go/+/404299
Reviewed-by: Bryan Mills <bcmills@google.com>
Use a separate channel to report the final error of the copying
goroutines, receiving a value only when all of the copying goroutines
have completed. In a followup change (CL 401835), that will allow
waiters to select on goroutine completion alongside other events (such
as Context cancellation).
Also mildly refactor the watchCtx helper method so that its structure
better matches what will be needed to implement WaitDelay.
For #50436.
Change-Id: I54b3997fb6931d204814d8382f0a388a67b520f1
Reviewed-on: https://go-review.googlesource.com/c/go/+/435995
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
This fixes a bug introduced in CL 401834 in which calling Start twice
with pipes attached to a command would spuriously close those pipes
when returning the error from the second Start call.
For #50436.
Change-Id: I3563cc99c0a0987752190fef95da3e9927a76fda
Reviewed-on: https://go-review.googlesource.com/c/go/+/436095
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Add a slightly expanded version of the Decoder type comment to the top
level package doc, which explains that this package is not designed
to be hardened against adversarial inputs.
Change-Id: I8b83433838c8235eb06ded99041fdf726c811ee5
Reviewed-on: https://go-review.googlesource.com/c/go/+/436096
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Roland Shoemaker <roland@golang.org>
Auto-Submit: Roland Shoemaker <roland@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
There were many more uses of the variable name "cause" than "reason"
to hold error message details. Consistently use "cause" throughout.
Accordingly, s/MissingMethodReason/MissingMethodCause/.
Change-Id: I171d784faabc66a4c58ba8944784204687595203
Reviewed-on: https://go-review.googlesource.com/c/go/+/435418
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
This is close to what the compiler used to say, except now we say
"as T value" rather than "as type T" which is closer to the truth
(we cannot use a value as a type, after all). Also, place the primary
error and the explanation (cause) on a single line.
Make respective (single line) adjustment to the matching "cannot
convert" error.
Adjust various tests.
For #55326.
Change-Id: Ib646cf906b11f4129b7ed0c38cf16471f9266b88
Reviewed-on: https://go-review.googlesource.com/c/go/+/436176
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This change causes the parser to record the positions of the first
and last character in the file in new ast.File fields FileStart
and FileEnd.
The behavior of the existing Pos() and End() methods,
which record the span of declarations, must remain unchanged
for compatibility.
Fixesgolang/go#53202
Change-Id: I250b19e69f41e3590292c3fe6dea1943ec98f629
Reviewed-on: https://go-review.googlesource.com/c/go/+/427955
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Alan Donovan <adonovan@google.com>
Auto-Submit: Alan Donovan <adonovan@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
This CL removes some opcode placeholders that do not correspond
to any existing instructions and hence create confusion. Some
instructions that are no longer valid like LDMX are also removed.
Any references to this instruction in ISA 3.0 are considered
as documentation errata.
Change-Id: Ib71a657099723bbe1db88873233ee573b5c42fe7
Reviewed-on: https://go-review.googlesource.com/c/go/+/429860
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Paul Murphy <murp@ibm.com>
Run-TryBot: Archana Ravindar <aravind5@in.ibm.com>
Reviewed-by: Benny Siegert <bsiegert@gmail.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Archana Ravindar <aravind5@in.ibm.com>
If GOEXPERIMENT=coverageredesign is in effect, introduce a new
top-level '-cover' option to "go build" to turn on new-style hybrid
code coverage instrumentation. Similarly, use the new instrumentation
for "go test -cover".
The main effects of "-cover" under the hood are to instrument files at
the package level using cmd/cover and to pass additional options to
the compiler when building instrumented packages.
The previous workflow for "go tool -cover mypkg" would expand to a
series of "go tool cover" commands (one per file) followed by a single
package compilation command to build the rewritten sources.
With the new workflow, the Go command will pass all of the Go files in
a package to the cover tool as a chunk (along with a config file
containing other parameters), then the cover tool will write
instrumented versions of the sources along with another "output"
config with info on coverage variable names for the the compiler. The
Go command will then kick off the compiler on the modified source
files, also passing in the config file generated by cmd/cover.
Updates #51430.
Change-Id: Id65621ff6a8c70a30168c1412c2d6f805ff3b9e7
Reviewed-on: https://go-review.googlesource.com/c/go/+/355452
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Refactor references to "a.Package" in Builder.build to be consistent,
e.g. all via the "p" helper variable. No change in functionality, this
is just intended to make the code more readable/consistent (since
prior to this point we had a random mix of "p" and "a.Package").
Change-Id: Ifea5ff9b314c39a0cf6e688511907d08cc56c603
Reviewed-on: https://go-review.googlesource.com/c/go/+/424819
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
Add a new mode of coverage instrumentation that works as a hybrid
between purely tool-based and purely compiler-based. The cmd/cover
tool still does source-to-source rewriting, but it also generates
information to be used by the compiler to do things like marking
meta-data vars as read-only.
In hybrid mode, the cmd/cover tool is invoked not on a single source
file but on all the files in a package, and is passed a config file
containing the import path of the package in question, along with
other parameters needed for the run. It writes a series of modified
files and an output config file to be passed to the compiler when
compiling the modified files.
Not completely useful by itself, still needs a corresponding set of
changes in the Go command and in the compiler.
Updates #51430.
Change-Id: I0fcbd93a9a8fc25064187b159152486a2549ea54
Reviewed-on: https://go-review.googlesource.com/c/go/+/395896
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
This patch fleshes out the runtime support for emitting coverage data
at the end of a run of an instrumented binary. Data is emitted in the
form of a pair of files, a meta-out-file and counter-data-outfile,
each written to the dir GOCOVERDIR. The meta-out-file is emitted only
if required; no need to emit again if an existing meta-data file with
the same hash and length is present.
Updates #51430.
Change-Id: I59d20a4b8c05910c933ee29527972f8e401b1685
Reviewed-on: https://go-review.googlesource.com/c/go/+/355451
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Than McIntosh <thanm@google.com>
Add a set of helper packages for reading collections of related
meta-data and counter-data files ("pods") produced by runs of
coverage-instrumented binaries, and a new tool program (cmd/covdata)
for dumping and/or manipulating coverage data files.
Currently "go tool covdata" subcommands include 'merge', 'intersect',
'percent', 'pkglist', 'subtract', 'debugdump', and 'textfmt'
(conversion to the legacy "go tool cover" format).
Updates #51430.
Change-Id: I44167c578f574b4636ab8726e726388531fd3258
Reviewed-on: https://go-review.googlesource.com/c/go/+/357609
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Matches compiler behavior and is consistent with what we do with other
binary operations.
While at it, also use parentheses rather than a colon for a couple of
errors caused by not having a core type.
For #55326.
Change-Id: I0a5cec1a31ffda98d363e5528791965a1ccb5842
Reviewed-on: https://go-review.googlesource.com/c/go/+/435618
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Fix the coversion between our sentinel salt length variables and the
BoringSSL versions in SignRSAPSS. We previously set -1 (hash length
equals salt length) when 0 was passed when we should've been setting
-2. This now matches the conversion that happens in VerifyRSAPSS. Also
adds a note documenting why we do this.
Additionally in non-Boring mode, properly handle passing of salt lengths
with a negative value which aren't one of the magic constants, returning
an error instead of panicking.
See https://commondatastorage.googleapis.com/chromium-boringssl-docs/rsa.h.html#RSA_sign_pss_mgf1
for the BoringSSL docs.
Fixes#54803
Change-Id: Id1bd14dcf0ef4733867367257830ed43e25ef882
Reviewed-on: https://go-review.googlesource.com/c/go/+/426659
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
Run-TryBot: Roland Shoemaker <roland@golang.org>
When a comparison is invalid due to mismatched types, we only know
when we see the 2nd operand; so use that operand's position for the
error message. This matches compiler behavior.
For #55326.
Change-Id: I79450756bbdd2b4bb90ed4e960a451be0197b186
Reviewed-on: https://go-review.googlesource.com/c/go/+/435555
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
An SHT_NOBITS section contains no bytes and occupies no space in the
file. This change makes it return an error on reading from this section
so that it will force the caller to check for an SHT_NNOBITS section.
We have considered another option to return "nil, nil" for the Data
method. It's abandoned because it might lead a program to simply do
the wrong thing, thinking that the section is empty.
Please note that it breaks programs which expect a byte slice with the
length described by the sh_size field. There are two reasons to
introduce this breaking change: 1. SHT_NOBITS means no data and it's
unnecessary to allocate memory for it; 2. it could result in an OOM if
the file is corrupted and has a huge sh_size.
Fixes#54967.
Change-Id: I0c3ed4e097214fe88413d726a89122105ad45d4f
GitHub-Last-Rev: 994c12d9da
GitHub-Pull-Request: golang/go#54994
Reviewed-on: https://go-review.googlesource.com/c/go/+/429601
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Compromise between old compiler error "T.m redeclared in this block"
(where the "in this block" is not particularly helpful) and the old
type-checker error "method m already declared for type T ...".
In the case where we have position information for the original
declaration, the error message is "method T.m already declared at
<position>". The new message is both shorter and more precise.
For #55326.
Change-Id: Id4a7f326fe631b11db9e8030eccb417c72d6c7db
Reviewed-on: https://go-review.googlesource.com/c/go/+/435016
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
This is a compromise of the error reported by the compiler (quotes
around field name removed) and the error reported by the type checkers
(added mention of struct type).
For #55326.
Change-Id: Iac4fb5c717f17c6713e90d327d39e68d3be40074
Reviewed-on: https://go-review.googlesource.com/c/go/+/434815
Reviewed-by: Robert Findley <rfindley@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
This matches longstanding compiler behavior.
Also, for unused packages, report:
`"pkg" imported and not used`
`"pkg" imported as X and not used`
This matches the other `X declared and not used` errors.
For #55326.
Change-Id: Ie71cf662fb5f4648449c64fc51bede298a1bdcbf
Reviewed-on: https://go-review.googlesource.com/c/go/+/432557
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Remove OpPPC64LoweredMuluhilo as this operation can be done
more efficiently with MULHDU and MULLD directly. This has the
benefit of not needing to use tuple select operations, and giving
the scheduler more freedom to place these operations.
The primary reason to avoid using tuples here is to to avoid
suboptimal scheduling when carry ops (e.x ADDC/ADDE) are used in
the same block as 64->128b multiples. CL 432275 modifies the
scheduling priorities which may cause non-flag/non-carry generating
tuple ops to interfere with carry opcodes. Thus resulting in excess
saving and restoring of the XER register.
This allows CL 432275 to adjust the scheduling priorities without
having to workaround odd tuple scheduling behavior.
Change-Id: Id04ef009ec4b86416e5436f2b44ae1474e73720e
Reviewed-on: https://go-review.googlesource.com/c/go/+/434855
Run-TryBot: Paul Murphy <murp@ibm.com>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Prior to this change, the scanner, in ScanComments mode, would emit
the implicit SEMICOLON token generated by a newline ahead of any
immediately preceding comment "tokens". For example:
foo /*a*/ /*b*/ /*c*/ \n => [IDENT SEMICOLON COMMENT COMMENT COMMENT]
Now, tokens and comments are emitted by the scanner in lexical order
of their start position. SEMICOLON tokens corresponding to newlines
always have the position of the newline (even in the special case
in which the newline is in the middle of a general comment /*\n*/).
The scanner no longer needs to look ahead, possibly across multiple
comments, for a newline, when it encounters a comment.
The scanner semicolon tests have been rewritten to be less magical.
The parser must now expect line comments before an implicit semicolon.
Line comments for an explicit semicolon still appear after.
The new assertions in the parser TestLeadAndLineComments are
not changes to behavior.
Fixesgolang/go#54941
Change-Id: Iffe97fd10e9e0b52882da8659307698ccb31c093
Reviewed-on: https://go-review.googlesource.com/c/go/+/429635
Auto-Submit: Alan Donovan <adonovan@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Alan Donovan <adonovan@google.com>
Update the linker's XCOFF writer to handle coverage counter sections
properly; they need to be treated as bss, not data. Fixes a problem
with the aix-ppc64 builder introduced in CL 401235), e.g.
"runtime.covctrs: A symbol with type XTY_SD must be in a .text or
.data section".
Updates #51430.
Change-Id: I3fc385a37e2549c46cc7cc3b4718af989a36752a
Reviewed-on: https://go-review.googlesource.com/c/go/+/435335
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: David Chase <drchase@google.com>
PPC64 ELFv2 uses the st_other field of a symbol to specify an offset
from the global entry point to its local entry point. Similarly, some
values (i.e 1) may also require additional linker support which is
missing today.
For now, generate an error if we encounter unsupported local entry
values on PPC64, and update the Localentry values to use bytes, not
32b instruction words.
Similarly, ELFv2 1.5 also updates the wording of values 2-6. They
now map to a specific number of bytes.
Change-Id: Id1b71c3b0fea982bdcfb7eac91d9f93e04ae43f9
Reviewed-on: https://go-review.googlesource.com/c/go/+/431876
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
The spec https://refspecs.linuxfoundation.org/elf/gabi4+/ch4.eheader.html
states:
1. e_shnum: If the number of sections is greater than or equal to
SHN_LORESERVE (0xff00), this member has the value zero and the actual
number of section header table entries is contained in the sh_size
field of the section header at index 0.
2. e_shstrndx: If the section name string table section index is
greater than or equal to SHN_LORESERVE (0xff00), this member has the
value SHN_XINDEX (0xffff) and the actual index of the section name
string table section is contained in the sh_link field of the section
header at index 0.
This CL makes these changes to support files with >= 0xff00 sections:
1. if shoff > 0 && shnum == 0, read sh_size from the initial section
header entry as shnum.
2. if shstrndx == SHN_XINDEX, read sh_link from the initial section
header entry as shstrndx.
It returns an error if the type of the initial section is not SHT_NULL.
A file with >= 0xff00 sections is too big to include in the repository,
so the test case constructs one on the fly, with some of the sections
zeroed out.
While here, remove the unnecessary use of reflect.DeepEqual in the test.
Fixes#55294.
Change-Id: I15ec43612c7cce6e8decfe4e81da3a5b16de47f7
GitHub-Last-Rev: 797c16480b
GitHub-Pull-Request: golang/go#55295
Reviewed-on: https://go-review.googlesource.com/c/go/+/432255
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Relocate cmd/go's search.MatchPattern helper routine to a new package
in cmd/internal from its current location, as to allow it to be used
in other tools that accept package pattern command line flags. No
change in functionality along the way.
Updates #51430.
Change-Id: I726e974ccd66a055bb5a94497b36b8d68d47cad1
Reviewed-on: https://go-review.googlesource.com/c/go/+/432757
Reviewed-by: Bryan Mills <bcmills@google.com>
Extra Ms may lead to the "no consistent ordering of events possible" error when parsing trace file with cgo enabled, since:
1. The gs in the extra Ms may be in `_Gdead` status while starting trace by invoking `runtime.StartTrace`,
2. and these gs will trigger `traceEvGoSysExit` events in `runtime.exitsyscall` when invoking go functions from c,
3. then, the events of those gs are under non-consistent ordering, due to missing the previous events.
Add two events, `traceEvGoCreate` and `traceEvGoInSyscall`, in `runtime.StartTrace`, will make the trace parser happy.
Fixes#29707
Change-Id: I2fd9d1713cda22f0ddb36efe1ab351f88da10881
GitHub-Last-Rev: 7bbfddb81b
GitHub-Pull-Request: golang/go#54974
Reviewed-on: https://go-review.googlesource.com/c/go/+/429858
Run-TryBot: Michael Pratt <mpratt@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: xie cui <523516579@qq.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Michael Pratt <mpratt@google.com>
Add an additional guard to ensure that we don't try to run the "-race"
variant of the exit hooks test when CGO is explicitly turned off via
CGO_ENABLED=0 (this fixes a failure on the no-cgo builder caused
by CL 354790).
Change-Id: I9dc7fbd71962e9a098916da69d9119a753f27116
Reviewed-on: https://go-review.googlesource.com/c/go/+/434935
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Adds a -coveragecfg=<configfile> command line option to the compiler
to help support a cooperative "tool and compiler" mode for coverage
instrumentation. In this mode the cmd/cover tool generates most of the
counter instrumentation via source-to-source rewriting, but the
compiler fixes up the result if passed the "-coveragecfg" option. The
fixups include:
- reclassifying counter variables (special storage class)
- marking meta-data variables are read-only
- adding in an init call to do registation
Updates #51430.
Change-Id: Iead72b85209725ee044542374465f118a3ee72e0
Reviewed-on: https://go-review.googlesource.com/c/go/+/395895
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Add a new "coverage counter" classification for variables to be used
for storing code coverage counter values (somewhat in the same way
that we identify fuzzer counters). Tagging such variables allows us to
aggregate them in the linker, and to treat updates specially.
Updates #51430.
Change-Id: Ib49fb05736ffece98bcc2f7a7c37e991b7f67bbb
Reviewed-on: https://go-review.googlesource.com/c/go/+/401235
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Add a new API (not public/exported) for registering a function with
the runtime that should be called when program execution terminates,
to be used in the new code coverage re-implementation. The API looks
like
func addExitHook(f func(), runOnNonZeroExit bool)
The first argument is the function to be run, second argument controls
whether the function is invoked even if there is a call to os.Exit
with a non-zero status. Exit hooks are run in reverse order of
registration, e.g. the first hook to be registered will be the last to
run. Exit hook functions are not allowed to panic or to make calls to
os.Exit.
Updates #51430.
Change-Id: I906f8c5184b7c1666f05a62cfc7833bf1a4300c4
Reviewed-on: https://go-review.googlesource.com/c/go/+/354790
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Add support to the runtime for registering coverage-instrumented
packages, using a new hook that can be called from the init function
of an instrumented package. The hook records the meta-data symbol for
the package (chaining it onto a list), and returns a package ID to be
used to identify functions in the package. This new hook is not yet
called; that will be added in a subsequent patch. The list of
registered meta-data objects will be used (again in a future patch) as
part of coverage data file writing.
Special handling is required for packages such as "runtime" or
"internal/cpu", where functions in the package execute before the
package "init" func runs. For these packages hard-code the package ID,
then record the position of the package in the overall list so that we
can fix things up later on.
Updates #51430.
Change-Id: I6ca3ddf535197442a2603c6d7a0a9798b8496f40
Reviewed-on: https://go-review.googlesource.com/c/go/+/401234
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Than McIntosh <thanm@google.com>
Add utilities for reading and writing of counter data files as part of
the new code coverage implementation.
Trivia note: currently the contents of counter data files are emitted
in little-endian form, which for the counters is somewhat painful in
that we need to visit each counter value and properly encode it. It
might be better to instead emit counters in native endianity and then
teach the tools to decode properly in the case of an endianity
mismatch.
Updates #51430.
Change-Id: I124fdcb40fc339a48b64b35264bf24c3be50ddd4
Reviewed-on: https://go-review.googlesource.com/c/go/+/359403
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
follow the TODO to use StructOf, this cl can save the used memory.
for example, old code alloc [128]*rtype for func with 65 arguments(in+out),
this cl change to alloc [65]*rtype to save memory.
Change-Id: I1494bb6b3524d0d46869c3f24a628fec88119d8b
Reviewed-on: https://go-review.googlesource.com/c/go/+/425314
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
We need a better approach eventually, but this is ok for now.
In go/types, always use _UnsupportedFeature for unavailable
version-specific features.
Change-Id: I15b47e34eda167db3133bd481aa2f55cf3662c31
Reviewed-on: https://go-review.googlesource.com/c/go/+/433195
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Add a coverage meta-data decoder, which provides APIs for reading
encoded coverage meta-data and expanding it usable form. This package
is intended to be used in the coverage tooling that reads data files
emitted from coverage runs. Along with the new decoding package is a
unit test that runs the encode/decode paths together to check to make
sure that "decode(encode(X)) == X".
Updates #51430.
Change-Id: I81d27d8da0b2fcfa5039114a6e35a4b463d19b3c
Reviewed-on: https://go-review.googlesource.com/c/go/+/353454
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Add a new package with APIs for encoding coverage meta-data. This
provides support for accumulating information about each function
during the compilation process, and then encoding and emitting a
payload for a coverage meta-data symbol. Not yet connected to the
rest of the coverage machinery (that will appear in a later patch).
Updates #51430.
Change-Id: I61054ce87f205b25fb1bfedaa740fd7425c34de4
Reviewed-on: https://go-review.googlesource.com/c/go/+/353453
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The pipes pumped by goroutines can be closed as soon as their
respective goroutines are done.
The pipes pumped by user code, however, are documented to be closed in
Wait. When we add the WaitDelay field, it isn't obvious that we should
terminate the user-pumped pipes when the WaitDelay expires, since Wait
itself isn't going to wait for those user-controlled goroutines to
complete.
(It's a bit more complicated than that because the documentation
currently states that Wait must not be called while the pipes are
being read — but it isn't obvious to me that that advice is entirely
correct.)
For #50436.
Change-Id: I97909c91d2097fb75138a360747168c08609696d
Reviewed-on: https://go-review.googlesource.com/c/go/+/401894
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Change the childFiles field to a local variable, since it was
populated during Start and (as far as I can determine) has no purpose
after Start returns.
Rename closeAfterStart and closeAfterWait to childIOFiles and
parentIOPipes respectively. That makes their contents clearer, and also
helps to clarify what should happen on error (when, for example, Wait
shouldn't be called at all).
Use a deferred call instead of individual calls to close child (and,
if necessary, pipe) FDs after Start. That helps to clarify the
invariants around when they are closed, and also makes the function a
bit more robust for future refactoring.
Also nil out the slices containing the file closers so that they can
be collected earlier.
This CL is intended as a pure refactor in preparation for #50436.
Change-Id: I05d13fa91d539b95b84b2ba923c1733f9a6203e5
Reviewed-on: https://go-review.googlesource.com/c/go/+/401834
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Fix up the package test harness to avoid errors of the form
go_test.go:NNN: internal testsuite error: path(".") with no tempdir
when the "-testwork" flag is passed when running "go test".
Fixes#55874.
Change-Id: I76cc39902f51a62cb4cd0da8794b97d620fb4873
Reviewed-on: https://go-review.googlesource.com/c/go/+/434455
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This avoids allocating an overly large slice for corrupt input.
Change the saferio.SliceCap function to take a pointer to the element type,
so that we can handle slices of interface types. This revealed that a
couple of existing calls were actually incorrect, passing the slice type
rather than the element type.
No test case because the problem can only happen for invalid data. Let
the fuzzer find cases like this.
Fixes#55338
Change-Id: I3c1724183cc275d4981379773b0b8faa01a9cbd2
Reviewed-on: https://go-review.googlesource.com/c/go/+/433296
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Daniel Martí <mvdan@mvdan.cc>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Query parameter smuggling occurs when a proxy's interpretation
of query parameters differs from that of a downstream server.
Change ReverseProxy to avoid forwarding ignored query parameters.
Remove unparsable query parameters from the outbound request
* if req.Form != nil after calling ReverseProxy.Director; and
* before calling ReverseProxy.Rewrite.
This change preserves the existing behavior of forwarding the
raw query untouched if a Director hook does not parse the query
by calling Request.ParseForm (possibly indirectly).
Fixes#54663
Fixes CVE-2022-2880
Change-Id: If1621f6b0e73a49d79059dae9e6b256e0ff18ca9
Reviewed-on: https://go-review.googlesource.com/c/go/+/432976
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
Currently, for version errors, types2 adds the helpful hint
(-lang was set to go1.xx; check go.mod)
where 1.xx is the respective language version, to the error message.
This requires that the type checker knows that it was invoked by the
compiler, which is done through the Config.CompilerErrorMessages flag.
This change looks for version errors being returned by the type checker
and then adds the hint at that point, external to the type checker.
This removes a dependency on the Config.CompilerErrorMessages. Once
we have removed all dependencies on Config.CompilerErrorMessages we
can remove it.
For #55326.
Change-Id: I1f9b2e472c49fe785a2075e26c4b3d9b8fcdbf4d
Reviewed-on: https://go-review.googlesource.com/c/go/+/432559
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
A linear search through a list of 22 strings takes ~80ns.
A quick check for 3-4 byte strings reduces this check to 2ns
for a vast majority of inputs.
In the event of a name match, the new logic is either just
as fast (for "CON") or 10x faster (for "LPT9").
Change-Id: I412fa73beebd7c81dc95f9ed12c35ca1d5d6baf0
Reviewed-on: https://go-review.googlesource.com/c/go/+/433175
Reviewed-by: Damien Neil <dneil@google.com>
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Pprof's converter from legacy text format to protobuf format
assumes that if the alloc and inuse stats are equal, then what's
really going on is that the program makes no distinction, and it
reads them as a two-column profile: objects and bytes.
Most of the time, some sampled object has been freed, and alloc != inuse.
In that case, pprof reads the profile as a four-column profile, with
alloc_objects, alloc_bytes, inuse_objects, inuse_bytes.
The 2-column form causes problems in a few ways. One is that if
you are reading the proto form and expect samples with the 4-column
names, they're not there. Another is that pprof's profile merger insists
on having the same number of columns and same names. This means
that
pprof *.memprofile
works most of the time but fails if one of the memory profiles hit
the unlikely condition that alloc == inuse, since now its converted
form differs from the others.
Most programs should simply not be using this output form at all,
but cmd/compile and cmd/link still do, because x/tools/cmd/compilebench
reads some extra values from the text form that we have not yet added
to the proto form.
For the programs still writing this form, the easiest way to avoid the
column collapse issues is to ensure that the header never reports
alloc == inuse. The actual values in the header are ignored by pprof now,
except for the equality check (they should sum to the other values in the
file, so they are technically redundant). Because the actual values are not
used except for the equality check, we could hard-code different values
like 0 and 1, but just in case, to break as little as possible, this CL only
adjusts the values when they would otherwise be equal. In that case it
adds 1 to allocBytes. For most profiles, where alloc != inuse already, there
is no effect at all.
Change-Id: Ia563e402573d0f6eb81ae496645db27c08f9fe31
Reviewed-on: https://go-review.googlesource.com/c/go/+/432758
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
For PIE binaries, the .gopclntab section doesn't have the usual
name, but .data.rel.ro.gopclntab. Try the relro version as well.
If both failed (e.g. for externally linked PIE binaries), try
runtime.pclntab symbol.
This should make cmd/objdump able to print the file/line
information for PIE binaries.
I attempted to do this a few years ago, but that wasn't enough,
because the pclntab itself contains dynamic relocations which are
not applied by the tool. As of Go 1.18 the pclntab is mostly
position independent and does not contain dynamic relocations, so
this should be possible now.
Fixes#17883.
Updates #46639.
Change-Id: I85dc3d50ffcc1a4b187a349479a6a162de1ab2b5
Reviewed-on: https://go-review.googlesource.com/c/go/+/227483
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Joel Sing <joel@sing.id.au>
TryBot-Result: Gopher Robot <gobot@golang.org>
After CL 22425, there're two optimizations for slice expr which are
never applied during walk pass:
s[i:len(s)]
s[i:j:cap(s)]
The order pass have already rewritten len/cap expression to use autotmp,
thus the same safe expression check will never fire. The code can now be
simplified by moving the only case left from reduceSlice to walkSlice,
then removing reduceSlice entirely.
Passes toolstash-check.
Change-Id: Ia8cfb15c8e96c186a214c17b42d0fee51b0d3a1c
Reviewed-on: https://go-review.googlesource.com/c/go/+/432695
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
cmd/dist can't easily hard-code -buildvcs=false because not all
versions of cmd/go supported for bootstrapping recognize that flag.
However, we don't want to stamp the bootstrap binaries: the stamping
is redundant with the VERSION file writted during bootstrapping (which
is why it is normally omitted for standard-library packages and
commands), and it may also interfere with building the Go repo from a
source tarball or zip file.
Fixes#54852.
Change-Id: If223f094af137c4c202d6bf622619bd2da397ec4
Reviewed-on: https://go-review.googlesource.com/c/go/+/432435
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
RFC 7231 permits HEAD requests to contain a body, although it does
state there are no defined semantics for payloads of HEAD requests
and that some servers may reject HEAD requests with a payload.
Accept HEAD requests with a body.
Fix a bug where a HEAD request with a chunked body would interpret
the body as the headers for the next request on the connection.
For #53960.
Change-Id: I83f7112fdedabd6d6291cd956151d718ee6942cd
Reviewed-on: https://go-review.googlesource.com/c/go/+/418614
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This commit adds an ASCII fast path to bytes/strings EqualFold that
roughly doubles performance when all characters are ASCII.
It also changes strings.EqualFold to use `for range` for the first
string since this is ~10% faster than using utf8.DecodeRuneInString for
both (see #31666).
Performance (similar results on arm64 and amd64):
name old time/op new time/op delta
EqualFold/Tests-10 238ns ± 0% 172ns ± 1% -27.91% (p=0.000 n=10+10)
EqualFold/ASCII-10 20.5ns ± 0% 9.7ns ± 0% -52.73% (p=0.000 n=10+10)
EqualFold/UnicodePrefix-10 86.5ns ± 0% 77.6ns ± 0% -10.37% (p=0.000 n=10+10)
EqualFold/UnicodeSuffix-10 86.8ns ± 2% 71.3ns ± 0% -17.88% (p=0.000 n=10+8)
Change-Id: I058f3f97a08dc04d65af895674d85420f920abe1
Reviewed-on: https://go-review.googlesource.com/c/go/+/425459
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
On Linux a signal sent using tgkill will have si_code == SI_TKILL,
not SI_USER. Treat the two cases the same. Add a Linux-specific test.
Change the test to use the C pause function rather than sleeping
for a second, as that achieves the same effect.
This is a roll forward of CL 431255 which was rolled back in CL 431715.
This new version skips flaky tests on more systems, and marks a new method
nosplit.
Change-Id: Ibf2d3e6fc43d63d0a71afa8fcca6a11fda03f291
Reviewed-on: https://go-review.googlesource.com/c/go/+/432136
Auto-Submit: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
When one has a []byte on hand, but desires to call the Parse function,
the conversion from []byte to string would allocate.
This occurs frequently through UnmarshalText and UnmarshalJSON.
This changes it such that the input string never escapes from
any of the Parse functions. Together with the compiler optimization
where the compiler stack allocates any string smaller than 32B
this makes most valid inputs for Parse(layout, string(input))
not require an allocation for the input string.
This optimization works well for most RFC3339 timestamps.
All timestamps with second resolution
(e.g., 2000-01-01T00:00:00Z or 2000-01-01T00:00:00+23:59)
or timestamps with nanosecond resolution in UTC
(e.g., 2000-01-01T00:00:00.123456789Z)
are less than 32B and benefit from this optimization.
Unfortunately, nanosecond timestamps with non-UTC timezones
(e.g., 2000-01-01T00:00:00.123456789+23:59)
do not benefit since they are 35B long.
Previously, this was not possible since the input leaked
to the error and calls to FixedZone with the zone name,
which causes the prover to give up and heap copy the []byte.
We fix this by copying the input string in both cases.
The advantage of this change is that you can now call Parse
with a []byte without allocating (most of the times).
The detriment is that the timezone and error path has an extra allocation.
Handling of timezones were already expensive (3 allocations and 160B allocated),
so the additional cost of another string allocation is relatively minor.
We should optimize for the common case, rather than the exceptional case.
Performance:
name old time/op new time/op delta
ParseRFC3339UTCBytes 54.4ns ± 1% 40.3ns ± 1% -25.91% (p=0.000 n=9+10)
Now that parsing of RFC3339 has been heavily optimized in CL 425197,
the performance gains by this optimization becomes relatively more notable.
Related to CL 345488.
Change-Id: I2a8a9cd6354b3bd46c2f57818ed2646a2e485f36
Reviewed-on: https://go-review.googlesource.com/c/go/+/429862
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
TryBot-Result: Gopher Robot <gobot@golang.org>
Replace custom append functions in the hash functions with the implementation of the encoding/binary package that do the same thing.
The binary bigendian functions are already used in other parts of the code in the crypto package.
Change-Id: I76d2dbe143fc72a3b4ac06be312caf72bd71378a
GitHub-Last-Rev: 1c6c68279e
GitHub-Pull-Request: golang/go#55085
Reviewed-on: https://go-review.googlesource.com/c/go/+/431035
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Use "%w" instead of "%v" as format verb for error value in the NextPart
method. This way it will be possible to use common go error utilities
from std library when parsing from custom io.Readers.
This issue was discovered during attempts to use
http.Request.ParseMultipartForm together with http.MaxBytesHandler.
Change-Id: Idb82510fb536b66b51ed1d943737c4828f07c2f2
GitHub-Last-Rev: 8bc49c945c
GitHub-Pull-Request: golang/go#55133
Reviewed-on: https://go-review.googlesource.com/c/go/+/431675
Run-TryBot: Ian Lance Taylor <iant@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Damien Neil <dneil@google.com>
This CL adds some optimizaion rules:
1, Converts CMP to CMN, or vice versa, when comparing with a negative
number.
2, For equal and not equal comparisons, CMP can be converted to CMN in
some cases. In theory we could do the same optimization for LT, LE, GT
and GE, but need to account for overflow, this CL doesn't handle them.
There are no noticeable performance changes.
Change-Id: Ia49266c019ab7908ebc9510c2f02e121b1607869
Reviewed-on: https://go-review.googlesource.com/c/go/+/429795
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Eric Fang <eric.fang@arm.com>
get, at least, is called from typedmemclr which must not be interruptible.
These were previously nosplit by accident before CL 424395 (the only
call they had was an intrinsic, so they were leaf functions, so they had
no prologue). After CL 424395 they contained a call (in noinline builds),
thus had a prologue, thus had a suspension point.
I have no idea how we might test this.
This is another motivating use case for having a nosplitrec directive
in the runtime.
Fixes#55156Fixes#54779Fixes#54906Fixes#54907
Change-Id: I851d733d71bda7172c4c96e027657e22b499ee00
Reviewed-on: https://go-review.googlesource.com/c/go/+/431919
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Although I can't think of any reason to do this, it is possible for
a user-defined flag to implement IsBoolFlag but return "false".
This is nuts because checking the interface is satisfied should
obviously be sufficient, but the documentation kinda implies it's
not. And if you try this, you'll discover that the usage message
ignores the return value even though the rest of the package plays
nice. Bother.
So we fix it, as the fix is trivial: call the method when creating
the usage message.
Fixes#53473
Change-Id: I1ac80a876ad5626eebfc5ef6cb972cd3007afaad
Reviewed-on: https://go-review.googlesource.com/c/go/+/431102
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
The nounified frontend currently tries to construct dictionaries that
correspond to invalid instantiations (i.e., instantiations T[X] where
X does not satisfy the constraints specified on T's type parameter).
As a consequence, we may fail to find method expressions needed by the
dictionary.
The real fix for this is to avoid creating those dictionaries in the
first place, because they should never actually be needed at runtime.
But that seems scary for a backport: we've repeatedly attempted to
backport generics fixes, which have fixed one issue but introduced
another.
This CL is a minimally invasive solution to #54225, which avoids the
ICE by instead skipping emitting the invalid dictionary. If the
dictionary ends up not being needed (which I believe will always be
the case), then the linker's reachability analysis will simply ignore
its absence.
Or worst case, if the dictionary *is* reachable somehow, we've simply
turned an ICE into a link-time missing symbol failure. That's not
great for user experience, but it seems like a small trade off to
avoid risking breaking any other currently working code.
Updates #54225.
Change-Id: Ic379696079f4729b1dd6a66994a58cca50281a84
Reviewed-on: https://go-review.googlesource.com/c/go/+/429655
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
In unified IR, local defined types are promoted to a global defined
type with a "vargen" suffix. These shouldn't actually be exposed to
go/types users, because they're only relevant within function bodies,
which go/types doesn't support importing.
Moreover, in the case of defined types that were declared within a
generic function, they can have ambient type parameters, which the
go/types importer doesn't know how to handle (because they shouldn't
be needed for that use case).
While here, prune the gcimporter_test.go skip list, because some of
the listed failures have actually been fixed and all of them are
specific to the Go 1.18 (nounified) frontend. They all work correctly
with GOEXPERIMENT=unified.
Fixes#55110.
Change-Id: I7dd8b86355d910dfed1d47edbad7695144c3f84d
Reviewed-on: https://go-review.googlesource.com/c/go/+/431495
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The type of the source and destination of a memmove call isn't
always accurate. It will always be a pointer (or an unsafe.Pointer), but
the base type might not be accurate. This comes about because multiple
copies of a pointer with different base types are coalesced into a single value.
In the failing example, the IData selector of the input argument is a
*[32]byte in one branch of the type switch, and a *[]byte in the other branch.
During the expand_calls pass both IDatas become just copies of the input
register. Those copies are deduped and an arbitrary one wins (in this case,
*[]byte is the unfortunate winner).
Generally an op v can rely on v.Type during rewrite rules. But relying
on v.Args[i].Type is discouraged.
Fixes#55122
Change-Id: I348fd9accf2058a87cd191eec01d39cda612f120
Reviewed-on: https://go-review.googlesource.com/c/go/+/431496
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
TestReadRequest_Bad's tests for leading whitespace in the first header
were also exercising the test verifying that a HEAD request has no
Content-Length. Also, the test intended to test a leading tab was
actually testing for a leading \t (literal backslash, literal t).
Change-Id: I05b46d05851b49bf75f1d1257c421b953b66ea9c
Reviewed-on: https://go-review.googlesource.com/c/go/+/428134
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Now we have 8-byte alignment types on 32-bit system, so in some rare
case, e.g, generated wrapper for embedded interface, the function
argument may need more than 4 byte alignment. We could pad somehow, but
this is a rare case which makes it hard to ensure that we've got it right.
So relaxing the check for argument and return value region of the stack.
Fixes#54991
Change-Id: I34986e17a920254392a39439ad3dcb323da2ea8d
Reviewed-on: https://go-review.googlesource.com/c/go/+/431098
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
When SLTI/SLTIU is used with ANDI/ORI, it may be possible to determine the
outcome based on the values of the immediates. Resolve these cases.
Improves code generation for various shift operations.
While here, sort tests by architecture to improve readability and ease
future maintenance.
Change-Id: I87e71e016a0e396a928e7d6389a2df61583dfd8d
Reviewed-on: https://go-review.googlesource.com/c/go/+/428217
Reviewed-by: Wayne Zuo <wdvxdr@golangcn.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Jenny Rakoczy <jenny@golang.org>
Reviewed-by: Jenny Rakoczy <jenny@golang.org>
Run-TryBot: Joel Sing <joel@sing.id.au>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Jenny Rakoczy <jenny@golang.org>
CL 427615 causes failures when Len64 can't be inlined. It's unclear to
me why this wasn't a problem before, but it is used in sensitive
contexts and therefore really should be marked. Confirmed that the
failures in question reproduce without this change, and don't reproduce
with it.
Fixes#55117.
Change-Id: Ic3aa96af1420cc0c39551908d83f954725c712f5
Reviewed-on: https://go-review.googlesource.com/c/go/+/431058
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
If one (accidentally) uses Go 1.16 as bootstrap toolchain, we
want it to print the error
found packages main (build.go) and building_Go_requires_Go_1_17_or_later (notgo117.go)
But because some files lack old style build tags, Go 1.16 instead
prints
//go:build comment without // +build comment
Add the build tags to make the error message work.
Change-Id: Iaa9b3c12e71842bb40f8687b2fda2cc4cb15b113
Reviewed-on: https://go-review.googlesource.com/c/go/+/431057
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
On Linux a signal sent using tgkill will have si_code == SI_TKILL,
not SI_USER. Treat the two cases the same. Add a Linux-specific test.
Change the test to use the C pause function rather than sleeping
for a second, as that achieves the same effect.
Change-Id: I2a36646aecabcab9ec42ed9a048b07c2ff0a3987
Reviewed-on: https://go-review.googlesource.com/c/go/+/431255
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Currently bgsweep attempts to be a low-priority background goroutine
that runs mainly when the application is mostly idle. To avoid
complicating the scheduler further, it achieves this with a simple
heuristic: call Gosched after each span swept. While this is somewhat
inefficient as there's scheduling overhead on each iteration, it's
mostly fine because it tends to just come out of idle time anyway. In a
busy system, the call to Gosched quickly puts bgsweep at the back of
scheduler queues.
However, what's problematic about this heuristic is the number of
tracing events it produces. Average span sweeping latencies have been
measured as low as 30 ns, so every 30 ns in the sweep phase, with
available idle time, there would be a few trace events emitted. This
could result in an overwhelming number, making traces much larger than
they need to be. It also pollutes other observability tools, like the
scheduling latencies runtime metric, because bgsweep stays runnable the
whole time.
This change fixes these problems with two modifications to the
heursitic:
1. Check if there are any idle Ps before yielding. If there are, don't
yield.
2. Sweep at least 10 spans before trying to yield.
(1) is doing most of the work here. This change assumes that the
presence of idle Ps means that there is available CPU time, so bgsweep
is already making use of idle time and there's no reason it should stop.
This will have the biggest impact on the aforementioned issues.
(2) is a mitigation for the case where GOMAXPROCS=1, because we won't
ever observe a zero idle P count. It does mean that bgsweep is a little
bit higher priority than before because it yields its time less often,
so it could interfere with goroutine scheduling latencies more. However,
by sweeping 10 spans before volunteering time, we directly reduce trace
event production by 90% in all cases. The impact on scheduling latencies
should be fairly minimal, as sweeping a span is already so fast, that
sweeping 10 is unlikely to make a dent in any meaningful end-to-end
latency. In fact, it may even improve application latencies overall by
freeing up spans and sweep work from goroutines allocating memory. It
may be worth considering pushing this number higher in the future.
Another reason to do (2) is to reduce contention on npidle, which will
be checked as part of (1), but this is a fairly minor concern. The main
reason is to capture the GOMAXPROCS=1 case.
Fixes#54767.
Change-Id: I4361400f17197b8ab84c01f56203f20575b29fc6
Reviewed-on: https://go-review.googlesource.com/c/go/+/429615
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
This change adds a metric to the runtime/metrics package which tracks
total mutex wait time for sync.Mutex and sync.RWMutex. The purpose of
this metric is to be able to quickly get an idea of the total mutex wait
time.
The implementation of this metric piggybacks off of the existing G
runnable tracking infrastructure, as well as the wait reason set on a G
when it goes into _Gwaiting.
Fixes#49881.
Change-Id: I4691abf64ac3574bec69b4d7d4428b1573130517
Reviewed-on: https://go-review.googlesource.com/c/go/+/427618
Reviewed-by: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently, wait reasons are set somewhat inconsistently. In a follow-up
CL, we're going to want to rely on the wait reason being there for
casgstatus, so the status quo isn't really going to work for that. Plus
this inconsistency means there are a whole bunch of cases where we could
be more specific about the G's status but aren't.
So, this change adds a new function, casGToWaiting which is like
casgstatus but also sets the wait reason. The goal is that by using this
API it'll be harder to forget to set a wait reason (or the lack thereof
will at least be explicit). This change then updates all casgstatus(gp,
..., _Gwaiting) calls to casGToWaiting(gp, ..., waitReasonX) instead.
For a number of these cases, we're missing a wait reason, and it
wouldn't hurt to add a wait reason for them, so this change also adds
those wait reasons.
For #49881.
Change-Id: Ia95e06ecb74ed17bb7bb94f1a362ebfe6bec1518
Reviewed-on: https://go-review.googlesource.com/c/go/+/427617
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This change adds 3 new waitReasons that correspond to sync.Mutex.Lock,
sync.RWMutex.RLock, and sync.RWMutex.Lock that are plumbed down into
semacquire1 by exporting new functions to the sync package from the
runtime.
Currently these three functions show up as "semacquire" in backtraces
which isn't very clear, though the stack trace itself should reveal
what's really going on. This represents a minor improvement to backtrace
readability, though blocking on an RWMutex.w.Lock will still show up as
blocking on a regular mutex (I suppose technically it is).
This is a step toward helping the runtime identify when a goroutine is
blocked on a mutex of some kind.
For #49881.
Change-Id: Ia409b4d27e117fe4bfdc25fa541e9c58d6d587b9
Reviewed-on: https://go-review.googlesource.com/c/go/+/427616
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
This changes adds a breakdown for estimated CPU usage by time. These
estimates are not based on real on-CPU counters, so each metric has a
disclaimer explaining so. They can, however, be more reasonably
compared to a total CPU time metric that this change also adds.
Fixes#47216.
Change-Id: I125006526be9f8e0d609200e193da5a78d9935be
Reviewed-on: https://go-review.googlesource.com/c/go/+/404307
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Josh MacDonald <jmacd@lightstep.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
There are lots of useless buckets with too much precision. Introduce a
minimum level of precision with a minimum bucket bit. This cuts down on
the size of a time histogram dramatically (~3x). Also, pick a smaller
sub bucket count; we don't need 6% precision.
Also, rename super-buckets to buckets to more closely line up with HDR
histogram literature.
Change-Id: I199449650e4b34f2a6dca3cf1d8edb071c6655c0
Reviewed-on: https://go-review.googlesource.com/c/go/+/427615
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Includes cases where the core type of the variadic parameter is
a slice or bytestring. Permits a client to create the signature
for various instantiations of append.
Fixes#55030.
Change-Id: I0f4983eb00c088cbe1d87954ee0b2df0ccc3bc49
Reviewed-on: https://go-review.googlesource.com/c/go/+/430455
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Apparently the new darwin linker starts to emit a warning about
-no_pie deprecation. Maybe we want to switch to PIE by default.
For now, suppress the warning. This also makes it easier for
backporting to previous releases.
For #54482.
Change-Id: I1a3b74c237a9d00ec3b030fc3a9940a31e5cd37e
Reviewed-on: https://go-review.googlesource.com/c/go/+/430937
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
When running on Go 1.19, we can further simplify some of the exec.Cmd
helpers due to API improvements. There's not much point in doing this
while the bootstrap is still 1.17, but this will queue up this
simplification in an obvious way for when we next upgrade the
bootstrap toolchain (#54265).
Updates #44505.
Change-Id: I2ebc3d5c584375ec862a1d48138ab134bd9b2366
Reviewed-on: https://go-review.googlesource.com/c/go/+/427958
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
function "ExampleFprint" will be rewritten to function "main"
when displayed on the godoc pages, so the online example is failed to
run:
Output:
panic: function not found
goroutine 1 [running]:
main.parseFunc({0x4f772e, 0xf}, {0x4f713f, 0xd})
/tmp/sandbox1264544227/prog.go:23 +0x13b
main.main()
/tmp/sandbox1264544227/prog.go:30 +0x45
See: https://pkg.go.dev/go/printer#example-Fprint
Add printSelf function to prevent the function not found when running in godoc
sandbox. Beside, deleting the dummy test function to make the example show
the entire file, as we want to show the newly added printSelf function.
Change-Id: Ia2b772937081b58a0fce9860838959c95f2d650c
GitHub-Last-Rev: bac1189173
GitHub-Pull-Request: golang/go#53141
Reviewed-on: https://go-review.googlesource.com/c/go/+/409314
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Daniel Martí <mvdan@mvdan.cc>
The object header string is meant to record the relevant toolchain
configuration, so that we don't import or link object files that are
incompatible with each other. One important part of compatibility
is the sub-architecture version (GOARM for GOARCH=arm, and so on).
Add the sub-architecture info to the object header line so that
binaries cannot be built that have inconsistent sub-architecture
configurations across the build.
This check is only important when the build system makes a mistake.
Builds using the go command don't make this kind of mistake anymore,
but we just debugged a difficult problem inside Google where a custom
build system had built part of a program with GOARM=5 and part of
a program with GOARM=7, resulting in corrupted execution when
signal-based preemption was attempted. Updating the check will avoid
this kind of problem in the future, in any custom build system, or if the
go command makes a mistake.
After this change:
% sed 3q pkg/darwin_amd64/runtime.a
!<arch>
__.PKGDEF 0 0 0 644 30525 `
go object darwin amd64 devel go1.20-102ebe10b7 Wed Aug 17 14:31:01 2022 -0400 GOAMD64=v1 X:regabiwrappers,regabiargs
%
Change-Id: I901e0758f1002dd2c58292dc65e2d06da86e4495
Reviewed-on: https://go-review.googlesource.com/c/go/+/427174
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
The concurrent model for delivering tokens was fine for pedagogy,
but has caused a few problems as the package has evolved (that is,
got more complicated). It's easy to eliminate it, simplifying or
removing some of the hacks used to work around these prolems.
The old lexer would deliver tokens over a channel to the parsing
goroutine, and continue running until EOF. In this rewrite, we
instead run the machine until a token is ready, and shut it down
until the next token is needed. The mechanism is just to return nil
as the state function, which requires a bit more threading of return
values through the state functions but is not difficult. The change
is modest.
A couple of error messages change, but otherwise the change has no
external effect. This is just an internal cleanup, long overdue.
benchmark old ns/op new ns/op delta
BenchmarkParseLarge-20 12222729 6769966 -44.61%
BenchmarkVariableString-20 73.5 73.4 -0.16%
BenchmarkListString-20 1827 1841 +0.77%
benchmark old allocs new allocs delta
BenchmarkVariableString-20 3 3 +0.00%
BenchmarkListString-20 31 31 +0.00%
benchmark old bytes new bytes delta
BenchmarkVariableString-20 72 72 +0.00%
BenchmarkListString-20 1473 1473 +0.00%
Fixes#53261
Change-Id: I4133bed2f8df16d398b707fb9509230325765c57
Reviewed-on: https://go-review.googlesource.com/c/go/+/421883
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
Use ppc64map (from x/arch) to generate ISA 3.1 support for the
assembler. A new file asm9_gtables.go is added which contains
generated code to encode ISA 3.1 instructions, a function to assist
filling out the oprange structure, a lookup table for the fixed
bits of each instructions, and a slice of string name. Generated
functions are shared if their bitwise encoding match, and the
translation from an obj.Prog structure matches.
The generated file is entirely self-contained, and does not require
regenerating any other files for changes within it. If opcodes in
a.out.go are reordered or changed, anames.go must be updated in
the same way as before.
Future improvements could shrink the generated opcode table
to 32 bit entries as there is much less variation of the
encoding of the prefix word, but it is not always identical
for instructions which share a similar encoding of arguments
(e.g PLWA and PLWZ).
Updates #44549
Change-Id: Ie83fa02497c9ad2280678d68391043d3aae63175
Reviewed-on: https://go-review.googlesource.com/c/go/+/419535
Run-TryBot: Paul Murphy <murp@ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Jenny Rakoczy <jenny@golang.org>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Jenny Rakoczy <jenny@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Auto-Submit: Jenny Rakoczy <jenny@golang.org>
This is the initial trivial implemenation. Further improvements can be
made for local calls.
A test is added, but the -fno-plt option is ignored by gcc if binutils
does not support inline plt relocations, so the test is effectively
skipped on such hosts.
Fixes#53345
Change-Id: Ibf31c26b1a8551c942b21019df8782c00b7a563e
Reviewed-on: https://go-review.googlesource.com/c/go/+/412714
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Jenny Rakoczy <jenny@golang.org>
Auto-Submit: Jenny Rakoczy <jenny@golang.org>
Run-TryBot: Jenny Rakoczy <jenny@golang.org>
Merge the CRC32 Update and Write functions using an unexported function, to avoid duplication of code and make it more readable.
The only difference between them is the check of the initialization of the IEEE table, and a boolean value specifies that.
Throughout the crc32.go file, in the switches the default value is inserted inside the switch statement, this change uniforms the style of the MakeTable function, making it like the other pieces of code.
Change-Id: I3889f6c6671210c82f0d7250cea67907bccf3ce7
GitHub-Last-Rev: b8777ee213
GitHub-Pull-Request: golang/go#55044
Reviewed-on: https://go-review.googlesource.com/c/go/+/430456
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Jenny Rakoczy <jenny@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Jenny Rakoczy <jenny@golang.org>
Auto-Submit: Jenny Rakoczy <jenny@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Use fcntl(oldfd, F_DUP2FD_CLOEXEC, newfd) to duplicate the file
descriptor and mark is as close-on-exec instead of dup2 & fcntl.
Note that the value for F_DUP2FD_CLOEXEC is different on Solaris and
Illumos and thus the definition is moved from zerrors_solaris_amd64.go
to solaris/illumos specific files.
Change-Id: I9a52801d1a01471ec3f065520575e3fafee92855
Reviewed-on: https://go-review.googlesource.com/c/go/+/428375
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
Auto-Submit: Jenny Rakoczy <jenny@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Jenny Rakoczy <jenny@golang.org>
Run-TryBot: Jenny Rakoczy <jenny@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Calling close is no longer necessary.
It was was originally necessary to cleanup goroutines
spawned to decompress the stream.
This has not been the case since CL 4548079.
Update the documentation to mention how it handles trailing data
after the end of the DEFLATE stream.
Change-Id: Ieacba264230560713b7b8d604665223fd096f4ec
Reviewed-on: https://go-review.googlesource.com/c/go/+/430377
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
Auto-Submit: Joseph Tsai <joetsai@digital-static.net>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
We added -pagezero_size in CL 72730, where it was intented for iOS.
The current code passes it only on macOS/AMD64 instead. It is not
really necessary there. Also, the new darwin linker starts to emit
a warning about deprecation of the flag. Stop passing it.
For #54482.
Change-Id: If9db7a1645c37d4284e48f075856912df8d8c1a0
Reviewed-on: https://go-review.googlesource.com/c/go/+/430936
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Allow us to select a race .syso file based on subarch values.
Note that this doesn't actually change the syso used. This CL
just moves things around in preparation for adding v3-specific
versions in future CLs.
Change-Id: I14e3c273a7c6f07b13b22193b7a851ea94c765cb
Reviewed-on: https://go-review.googlesource.com/c/go/+/424034
Reviewed-by: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Noticed in a manual audit from a customer codebase that the pattern
w.WriteString(fmt.Sprint*(args...))
was less efficient and in most cases we can just invoke:
fmt.Fprint*(w, args...)
and from the simple benchmarks we can see quick wins in all dimensions:
$ benchstat before.txt after.txt
name old time/op new time/op delta
DetailString-8 5.48µs ±23% 4.40µs ±11% -19.79% (p=0.000 n=20+17)
name old alloc/op new alloc/op delta
DetailString-8 2.63kB ± 0% 2.11kB ± 0% -19.76% (p=0.000 n=20+20)
name old allocs/op new allocs/op delta
DetailString-8 63.0 ± 0% 50.0 ± 0% -20.63% (p=0.000 n=20+20)
Change-Id: I47a2827cd34d6b92644900b1bd5f4c0a3287bdb1
Reviewed-on: https://go-review.googlesource.com/c/go/+/429861
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: hopehook <hopehook@golangcn.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Emmanuel Odeke <emmanuel@orijtech.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Daniel Martí <mvdan@mvdan.cc>
The existing implementation allocates a new string even when the
count is 1, where we know the output is the same as the input.
While we wouldn't expect a count of 1 for hardcoded values of the
parameter, it is expected when the parameter is computed based on
a different value (e.g., the length of a input slice).
name old time/op new time/op delta
Repeat/5x0-10 2.03ns ± 0% 2.02ns ± 0% ~ (p=1.000 n=1+1)
Repeat/5x1-10 13.7ns ± 0% 2.0ns ± 0% ~ (p=1.000 n=1+1)
Repeat/5x2-10 18.2ns ± 0% 18.1ns ± 0% ~ (p=1.000 n=1+1)
Repeat/5x6-10 27.0ns ± 0% 27.0ns ± 0% ~ (p=1.000 n=1+1)
Repeat/10x0-10 2.02ns ± 0% 2.02ns ± 0% ~ (p=1.000 n=1+1)
Repeat/10x1-10 16.1ns ± 0% 2.0ns ± 0% ~ (p=1.000 n=1+1)
Repeat/10x2-10 20.8ns ± 0% 20.9ns ± 0% ~ (p=1.000 n=1+1)
Repeat/10x6-10 29.2ns ± 0% 29.4ns ± 0% ~ (p=1.000 n=1+1)
Change-Id: I48e08e08f8f6d6914d62b3d6a61d563d637bec59
GitHub-Last-Rev: 068f58e08b
GitHub-Pull-Request: golang/go#53321
Reviewed-on: https://go-review.googlesource.com/c/go/+/411477
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Keith Randall <khr@google.com>
The changes improve the documentation for DefaultTransport, making
the style with how the HTTP proxy environment variables are being
referred to, consistent with the rest of the project's
documentation.
Also mention HTTPS_PROXY environment variables, as suggested in #32649.
Change-Id: I4e6b49881d7b30b5a0d4699531fa7c2929fc49f7
GitHub-Last-Rev: 2fc751937b
GitHub-Pull-Request: golang/go#54996
Reviewed-on: https://go-review.googlesource.com/c/go/+/430135
Reviewed-by: Damien Neil <dneil@google.com>
Auto-Submit: Jenny Rakoczy <jenny@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Jenny Rakoczy <jenny@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
objabi.HeadType is typically used as a non-pointer type, however the String function
is declared on a pointer receiver. This means that in most cases its integer value
is printed, rather than the value from the String function.
Change-Id: I3d28d9680e88a714bc1152ed5e1df4ac43d7a33f
Reviewed-on: https://go-review.googlesource.com/c/go/+/430556
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Jenny Rakoczy <jenny@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Jenny Rakoczy <jenny@golang.org>
If the entry symbol cannot be found (for example, a new port is being brought
up and no rt0 code has been provided), the linker will currently panic. Rather
than panicing, generate an error that aids in debugging:
missing entry symbol "_rt0_arm64_openbsd"
Change-Id: I9cc38eaab48f730d596ca7fa9e9e3d68250ae4d3
Reviewed-on: https://go-review.googlesource.com/c/go/+/430555
Auto-Submit: Jenny Rakoczy <jenny@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Jenny Rakoczy <jenny@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
RFC 3339 is the most common time representation,
being used in an overwhelming 57.3% of all specified formats,
while the next competitor only holds 7.5% usage.
Specially optimize parsing to handle the RFC 3339 format.
To reduce the complexity of error checking,
parseRFC3339 simply returns a bool indicating parsing success.
It leaves error handling to the general parse path.
To assist in fuzzing, the internal parse function was left unmodified
so that we could test that parseRFC3339 and parse agree with each other.
Performance:
name old time/op new time/op delta
ParseRFC3339UTC 112ns ± 1% 37ns ± 1% -67.37% (p=0.000 n=9+9)
ParseRFC3339TZ 259ns ± 2% 67ns ± 1% -73.92% (p=0.000 n=10+9)
Credit goes to Amarjeet Anand for a prior CL attemping to optimize this.
See CL 425014.
Fixes#54093
Change-Id: I14f4e8c52b092d44ceef6863f261842ed7e83f4c
Reviewed-on: https://go-review.googlesource.com/c/go/+/425197
Reviewed-by: Rob Pike <r@golang.org>
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Jenny Rakoczy <jenny@golang.org>
Add a new flag 'paramNames' to typeWriter struct to control whether
function parameter names are written or not (set by default). Unset
it when we want the function signature w/o parameter names, e.g. when
showing two signatures that are not identical. This makes is much
easier to see the typw differences in the error message.
To avoid needing to provide yet another (rarely used) boolean parameter
to typeString, remove that function in favor of setting the paramNames
flag explicitly. Adjust the code in errors.go that used typeString; the
resulting code is also more efficient (fewer bytes.Buffer allocations).
While at it, rename the typeWriter 'debug' field to 'tpSubscripts'
because that is what it controls.
Add test case and adjusted existing expected output for existing tests.
Fixes#54942.
Change-Id: I625eae30c403c39ce89951b8ea6214d783c92c75
Reviewed-on: https://go-review.googlesource.com/c/go/+/430416
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
This CL adds a suitable error code to every error reporting,
matching what go/types already does.
For now, the error codes are not progagated through the API,
but eventually they will be available to clients.
Also, for now the errorcodes.go file is a 1:1 copy (with
adjusted package name) of go/types/errorcodes.go.
A subsequent CL will factor out this file.
In contrast to go/types, for errors due to incorrect Go version,
the error code is always _UnsupportedFeature. In go/types, the
error sometimes is related to the specific operation.
Change-Id: I18771bc3d00bbdbf6d705bf25f2aea3c3d977b1c
Reviewed-on: https://go-review.googlesource.com/c/go/+/429355
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
The "&x.v" cost us 2 inline cost budget, causing extra inline cost for
the caller. By taking the fact that "v" is laid out in memory as the
first field of all atomic types, we can accessing it without addressing.
Discovering why attempting to convert sync.RWMutex.readerCount to atomic
type. RWMutex.RUnlock have the inline cost 75, with extra 7 inline cost
from Int32.Add causing it not inlinable anymore.
With this change, Int32.Add only has 5 inline cost budget, RWMutex can
use it while still be inlinable.
Change-Id: Iabe1d1bf53389b0b8b5f56b4611231b732fd9df5
Reviewed-on: https://go-review.googlesource.com/c/go/+/429766
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Implement CLONE_INTO_CGROUP feature, allowing to put a child in a
specified cgroup in a clean and simple way. Note that the feature only
works for cgroup v2, and requires Linux kernel 5.7 or newer.
Using the feature requires a new syscall, clone3. Currently this is the
only reason to use clone3, but the code is structured in a way so that
other cases may be easily added in the future.
Add a test case.
While at it, try to simplify the syscall calling code in
forkAndExecInChild1, which became complicated over time because:
1. It was using either rawVforkSyscall or RawSyscall6 depending on
whether CLONE_NEWUSER was set.
2. On Linux/s390, the first two arguments to clone(2) system call are
swapped (which deserved a mention in Linux ABI hall of shame). It
was worked around in rawVforkSyscall on s390, but had to be
implemented via a switch/case when using RawSyscall6, making the code
less clear.
Let's
- modify rawVforkSyscall to have two arguments (which is also required
for clone3);
- remove the arguments workaround from s390 asm, instead implementing
arguments swap in the caller (which still looks ugly but at least
it's done once and is clearly documented now);
- use rawVforkSyscall for all cases (since it is essentially similar to
RawSyscall6, except for having less parameters, not returning r2, and
saving/restoring the return address before/after syscall on 386 and
amd64).
Updates #51246.
Change-Id: Ifcd418ebead9257177338ffbcccd0bdecb94474e
Reviewed-on: https://go-review.googlesource.com/c/go/+/417695
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Run-TryBot: Kirill Kolyshkin <kolyshkin@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
First, we know that Go source files almost always weigh at least a few
kilobytes, so we can kickstart the output buffer to be a reasonable size
and reduce the initial number of incremental allocations and copies when
appending bytes or strings to output.
Second, in nodeSize we use a nested printer, but we don't actually need
its printed bytes - we only need to know how many bytes it prints.
For that reason, use a throwaway buffer: the part of our output buffer
between length and capacity, as we haven't used it yet.
Third, use a sync.Pool to reuse allocated printers.
The current API doesn't allow reusing printers,
and some programs like gofmt will print many files in sequence.
Those changes combined result in a modest reduction in allocations and
CPU usage. The benchmark uses testdata/parser.go, which has just over
two thousand lines of code, which is pretty standard size-wise.
We also split the Print benchmark to cover both a medium-sized ast.File
as well as a pretty small ast.Decl node. The latter is a somewhat common
scenario in gopls, which has code actions which alter small bits of the
AST and print them back out to rewrite only a few lines in a file.
name old time/op new time/op delta
PrintFile-16 5.43ms ± 1% 4.85ms ± 3% -10.68% (p=0.000 n=9+10)
PrintDecl-16 19.1µs ± 0% 18.5µs ± 1% -3.04% (p=0.000 n=10+10)
name old speed new speed delta
PrintFile-16 9.56MB/s ± 1% 10.69MB/s ± 3% +11.81% (p=0.000 n=8+10)
PrintDecl-16 1.67MB/s ± 0% 1.73MB/s ± 1% +3.05% (p=0.000 n=10+10)
name old alloc/op new alloc/op delta
PrintFile-16 332kB ± 0% 107kB ± 2% -67.87% (p=0.000 n=10+10)
PrintDecl-16 3.92kB ± 0% 3.28kB ± 0% -16.38% (p=0.000 n=10+10)
name old allocs/op new allocs/op delta
PrintFile-16 3.45k ± 0% 2.42k ± 0% -29.90% (p=0.000 n=10+10)
PrintDecl-16 56.0 ± 0% 46.0 ± 0% -17.86% (p=0.000 n=10+10)
Change-Id: I475a3babca77532b2d51888f49710f74763d81d2
Reviewed-on: https://go-review.googlesource.com/c/go/+/424924
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
After we deprecated reflect.{SliceHeader, StringHeader}, it is recommended
to use unsafe.{Slice, String} to replace its work. However, the compiler
and linker cannot be migrated for the time being.
As a temporary strategy, using the "internal/unsafeheader" package like
other code is the most suitable choice at present.
For #53003.
Change-Id: I69d0ef72e2d95caabd0706bbb247a719d225c758
Reviewed-on: https://go-review.googlesource.com/c/go/+/429755
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: hopehook <hopehook@golangcn.org>
As discussed in CL 401434 there are substantial misuses of these in the
wild, and they are a potential source of unsafety even for code that
does not use them directly.
Since proposal #53003 has already been implemented, now is the right
time to deprecate reflect.{SliceHeader, StringHeader}.
For #53003.
Change-Id: I724cf46d4b22d2ed3cbf2b948e6aac5ee4bf0f6e
Reviewed-on: https://go-review.googlesource.com/c/go/+/428757
Run-TryBot: hopehook <hopehook@golangcn.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
When the type assertion fails, the test mistakenly prints the expected
(rather than the actual) type.
When the error string doesn't match, the text mistakenly prints the
original (rather than the converted) error (although there might not be
any difference in the output, the code looks wrong).
Fix both issues.
Change-Id: Ia7dd0632fc677f458fec25d899c46268a12f76e7
Reviewed-on: https://go-review.googlesource.com/c/go/+/428916
Run-TryBot: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
FixedZone is transitively called by Time.UnmarshalJSON or Time.UnmarshalText
for any RFC 3339 timestamp that is not in UTC.
This function is relatively slow since it allocates 3 times.
Given that RFC 3339 never has a zone name and most offsets are by the hour,
we can cache unnamed zones on hour offsets.
Caching a Location should be safe since it has no exported fields or methods
that can mutate the Location. It is functionally immutable.
The only way to observe that the Location was cached is either
by pointer comparison or by shallow copying the struct value.
Neither operation seems sensible to do with a *time.Location.
Performance:
name old time/op new time/op delta
UnmarshalText 268ns ± 2% 182ns ± 1% -32.01% (p=0.000 n=10+10)
Change-Id: Iab5432f34bdbb485512bb8b5464e076c03fd106f
Reviewed-on: https://go-review.googlesource.com/c/go/+/425116
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
Auto-Submit: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Ian Lance Taylor <iant@google.com>
This change fixes an old TODO that made it a uint64 because it would
make alignment within mheap more complicated. Now that we don't have to
worry about it since we're using atomic types as much as possible,
switch to using a Uintptr. This likely will improve performance a tiny
bit on 32-bit platforms, but really it's mostly cleanup.
Change-Id: Ie705799a111ccad977fc1f43de8b50cf611be303
Reviewed-on: https://go-review.googlesource.com/c/go/+/429221
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
pageAlloc.chunks used to require an atomic store when growing the heap
because the scavenger would look at the list without locking the heap
lock. However, the scavenger doesn't do that anymore, and it looks like
nothing really does at all.
This change updates the comment and makes the store non-atomic.
Change-Id: Ib452d147861060f9f6e74e2d98ee111cf89ce8f5
Reviewed-on: https://go-review.googlesource.com/c/go/+/429219
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Right now, span sets use a lot of unsafe.Pointer and naked atomics
operations. This change modifies it to use atomic types everywhere and
wraps any atomic.UnsafePointer in a type to improve type safety.
This change should functionally be a no-op.
Change-Id: I32e6c460faaf6ec41ab1163158f6da7938eef3de
Reviewed-on: https://go-review.googlesource.com/c/go/+/429218
Reviewed-by: Keith Randall <khr@golang.org>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
These can be simplified with the knowledge of how arguments are
assigned to obj.Prog objects on PPC64. If the argument is not
a register type, the Reg argument (a2 in optab) of obj.Prog is
not used, and those arguments are placed into RestArgs (a3, a4, a5
in optab).
This relaxes the special case handling enforced by IsPPC64RLD and
IsPPC64ISEL. Instead, arguments are assigned as noted above, and
incorrect usage of such opcodes is checked by optab rules, not by
the assembler front-end.
Likewise, add support for handling 6 argument opcodes, these do
not exist today, but will be added with ISA 3.1 (Power10).
Finally, to maintain backwards compatibility, some 4-arg opcodes
whose middle arguments are a register and immediate, could swap
these arguments and generate identical machine code. This likely
wasn't intended, but is possible. These are explicitly fixed up
in the backend, and the asm tests are extended to check these.
Change-Id: I5f8190212427dfe8e6f062185bfefb5fa4fd0e75
Reviewed-on: https://go-review.googlesource.com/c/go/+/427516
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
This CL adds shiftIsBounded checks for the Lsh* and Rsh* rules in arm64.
There is no need to check the shift value again with CMP + CSEL when the
shift value is valid.
Change-Id: I54620de64f02a1b5a11089add237248ae2de01b4
Reviewed-on: https://go-review.googlesource.com/c/go/+/417714
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Apparently, some testing environments do not allow root to mount tmpfs
(due to e.g. AppArmor profile disallowing mount(2) syscall).
Always skip the test if the mount has failed.
Fixes the test issue introduced by CL 414824.
Change-Id: Ic565d2e6f277f2926d85a351be7df2498ffba656
Reviewed-on: https://go-review.googlesource.com/c/go/+/429175
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
When naming case variables, the unified frontend was using
typecheck.Lookup, which uses the current package, rather than
localIdent, which uses the package the variable was originally
declared in. When inlining across package boundaries, this could cause
the case variables to be associated with the wrong package.
In practice, I don't believe this has any negative consequences, but
it's inconsistent and triggered an ICE in typecheck.ClosureType, which
expected all captured variables to be declared in the same package.
Easy fix is to ensure case variables are declared in the correct
package by using localIdent.
Fixes#54912.
Change-Id: I7a429c708ad95723f46a67872cb0cf0c53a6a0d6
Reviewed-on: https://go-review.googlesource.com/c/go/+/428918
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Benny Siegert <bsiegert@gmail.com>
If test would fail, the error message will have wrong error and its
type, because e is used after the failed type assertion.
To fix, use the original err.
While at it,
- combine the checks for error type and value into one statement;
- use the standard "got ..., want ..." format.
Fixes: 212d2f82e0 ("os: add ErrClosed, return for use of closed File")
Change-Id: I862a96607b461ab89cce6bed2443b28aa2c16468
Reviewed-on: https://go-review.googlesource.com/c/go/+/428915
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Implement Slicemask the same way every other architecture does - negate
then arithmetic right shift. This sets or clears the sign bit, before
extending it to the entire register.
Removes around 2,500 instructions from the Go binary on linux/riscv64.
Change-Id: I4d675b826e7eb23fe2b1e6e46b95dcd49ab49733
Reviewed-on: https://go-review.googlesource.com/c/go/+/426354
Reviewed-by: Meng Zhuo <mzh@golangcn.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Joel Sing <joel@sing.id.au>
Reviewed-by: Wayne Zuo <wdvxdr@golangcn.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Having an executable bit set for a binary is not enough for it to be
executable -- there might be more checks in the kernel. For example,
binaries on a filesystem mounted with "noexec" flag couldn't be
executed. There might be other scenarios involving ACLs, SELinux,
file capabilities, and so on.
As a result, LookPath might either find a non-executable (while going
over $PATH elements), or return a false positive that the argument
provided is an executable.
One possible fix would be to perform the check by using access(2)
syscall with X_OK flag.
Now, since access(2) uses real (rather than effective) uid and gid,
when used by a setuid or setgid binary, it checks permissions of the
(real) user who started the binary, rather than the actual effective
permissions. Therefore, using access with X_OK won't work as expected
for setuid/setgid binaries.
To fix this, modern platforms added ways to check against effective uid
and gid, with the most common being the faccessat(2) call with the
AT_EACCESS flag, as described by POSIX.1-2008 (in Linux, only
faccessat2(2) supports flags such as AT_EACCESS). Let's use it, and fall
back to checking permission bits if faccessat is not available.
Wrap the logic into unix.Eaccess, which is currently only implemented on
Linux. While many other OSes (Free/Net/OpenBSD, AIX, Solaris/Illumos, and
Darwin) do implement faccessat(2) with AT_EACCESS, it is not wired in
syscall package (except for AIX), so these platforms are left out for now.
In the future, eaccess can be implemented for these OSes, too.
Alas, a call to unix.Eaccess is not enough since we have to filter out
directories, so use both stat and Eaccess.
One minor change introduced by this commit is that LookPath and Command
now returns "is a directory" error when the argument contains a slash
and is a directory. This is similar to what e.g. bash does on Linux:
$ bash -c /etc
bash: line 1: /etc: Is a directory
Add a test case, which, unfortunately, requires root, is specific to
Linux, and needs a relatively new kernel (supporting faccessat2). Other
platforms either have different semantics for tmpfs with noexec, or have
different ways to set up a binary which has x bit set but nevertheless
could not be executed.
Change-Id: If49b6ef6bf4dd23b2c32bebec8832d83e511a4bb
Reviewed-on: https://go-review.googlesource.com/c/go/+/414824
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
The parent, file, and line fields are no longer used now that we have
parentPc to find the parent and NOPs in the parent to attach file/line
pcdata to.
Removing these fields reduces the binary size of cmd/go on linux-amd64
by 1.1%.
Fixes#54849.
Change-Id: If58f08622736b2b322288608776f8bedf0c3fd17
Reviewed-on: https://go-review.googlesource.com/c/go/+/427960
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
Reviewed-by: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
For defer/go calls, the function/method value are evaluated immediately.
So after devirtualizing, it may trigger a panic when implicitly deref
a nil pointer receiver, causing the program behaves unexpectedly.
It's safer to not devirtualizing defer/go calls at all.
Fixes#52072
Change-Id: I562c2860e3e577b36387dc0a12ae5077bc0766bf
Reviewed-on: https://go-review.googlesource.com/c/go/+/428495
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
The writev syscall is available since at least Solaris 11.3.
Reuse the existing illumos writev wrapper on solaris to implement
internal/poll.writev for net.(*netFD).writeBuffers.
Change-Id: I23adc3bb4637740c72bfb61bfa9697b432dfe3db
Reviewed-on: https://go-review.googlesource.com/c/go/+/427714
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
The triv.go example serves the entire contents of $HOME by default.
That seems bad, let's not do that.
Also change it to listen on localhost only.
Change-Id: I8f1b7bd6b7d737852273e2ba82deabc4a2d11f6b
Reviewed-on: https://go-review.googlesource.com/c/go/+/428237
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Tatiana Bradley <tatiana@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
CL 424257 modified gentraceback to switch gp when jumping from a
system stack to a user stack to simplify reasoning through the rest of
the function. This has the unintended side-effect of also switching
all references to gp.m. The vast majority of the time, g0.m and curg.m
are the same across a stack switch, making this a no-op, but there's
at least one case where this isn't true: if a profiling signal happens
in execute between setting mp.curg and setting gp.m. In this case,
mp.curg.m is briefly nil, which can cause gentraceback to crash with a
nil pointer dereference. We see this failure (surprisingly
frequently!) in profiling tests in the morestack=mayMoreStackPreempt
testing mode (#48297).
Fix this by making only jumping stacks if doing so will not switch Ms.
This restores the original property that gp.m doesn't change across
the stack jump, and makes gentraceback a little more conservative
about jumping stacks.
Fixes#54885.
Change-Id: Ib1524c41c748eeff35896e0f3abf9a7efbe5969f
Reviewed-on: https://go-review.googlesource.com/c/go/+/428656
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Austin Clements <austin@google.com>
This reverts https://go.dev/cl/425881.
Reason for revert: broke make.bash on linux/amd64 with Linux 5.19.6.
[...]
Building Go toolchain2 using go_bootstrap and Go toolchain1.
go install internal/unsafeheader: copying /tmp/go-build4206185186/b007/_pkg_.a to /home/mvdan/tip/pkg/linux_amd64/internal/unsafeheader.a: write /home/mvdan/tip/pkg/linux_amd64/internal/unsafeheader.a: copy_file_range: invalid cross-device link
go install internal/goarch: copying /tmp/go-build4206185186/b006/_pkg_.a to /home/mvdan/tip/pkg/linux_amd64/internal/goarch.a: write /home/mvdan/tip/pkg/linux_amd64/internal/goarch.a: copy_file_range: invalid cross-device link
[...]
Change-Id: I793856935d4315a870c2d31da46be00cc342b5f8
Reviewed-on: https://go-review.googlesource.com/c/go/+/428396
Reviewed-by: Wayne Zuo <wdvxdr@golangcn.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
Reviewed-by: hopehook <hopehook@golangcn.org>
Reviewed-by: Andy Pan <panjf2000@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Use an atomic.Uint32 to represent the state of finalizer goroutine.
fingStatus will only be changed to fingWake in non fingWait state,
so it is safe to set fingRunningFinalizer status in runfinq.
name old time/op new time/op delta
Finalizer-8 592µs ± 4% 561µs ± 1% -5.22% (p=0.000 n=10+10)
FinalizerRun-8 694ns ± 6% 675ns ± 7% ~ (p=0.059 n=9+8)
Change-Id: I7e4da30cec98ce99f7d8cf4c97f933a8a2d1cae1
Reviewed-on: https://go-review.googlesource.com/c/go/+/400134
Reviewed-by: Joedian Reid <joedian@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
Reviewed-by: Michael Pratt <mpratt@google.com>
The initial CL 229101 didn't limit the kernel version, but relies on error checking to
ensure the kernel version >= 4.5 or >= 5.3 when it's calling copy_file_range(2) to copy data across file systems.
Since we have now put the kernel version checking at the beginning of the function, introduced by CL 268338,
which returns early instead of going forward to the code behind when the kernel verion is older than 5.3,
therefore, those subsequent related error checks are no longer needed.
Change-Id: Ifc4a530723e21f0bde91d6420cde9cb676081922
Reviewed-on: https://go-review.googlesource.com/c/go/+/425881
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: hopehook <hopehook@golangcn.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Daniel Martí <mvdan@mvdan.cc>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Consistently wait for idle connections to become readable before
starting the ReadHeaderTimeout timer. Previously, connections with no
idle timeout skipped directly to reading headers, so the
ReadHeaderTimeout also included time spent idle.
Fixes#54784
Change-Id: Iff1a876f00311d03dfa0fbef5b577506c62f7c41
GitHub-Last-Rev: 09332743ad
GitHub-Pull-Request: golang/go#54785
Reviewed-on: https://go-review.googlesource.com/c/go/+/426895
Run-TryBot: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: Benny Siegert <bsiegert@gmail.com>
As net package has one of the biggest init time in standard library, I have tried to improve performance by doing two things in net/addrselect.go:
1. Precompute slice with RFC rules. Currently the rules are computed and sorted in init() function. We could save the time and allocations by using prepopulated values in sorted manner. The rules haven't changed since 2015. To be extra safe we could move order validation as test case. It should slightly speed up startup of each binary with "net" package and go dns resolver. It also saves 38 allocations, ~50% of allocations in init phase of `net` module.
2. Replace internal net.IP usage with netip.Addr in `sortByRFC6724` function. It results in ~40% performance improvement on samples from tests.
The only risk is the difference between net.IP and netip.Addr behaviour.
Init benchmark:
Init-8 1.89µs ± 2% 0.12µs ± 3% -93.79% (p=0.000 n=5+5)
name old alloc/op new alloc/op delta
Init-8 1.05kB ± 0% 0.38kB ± 0% ~ (zero variance)
name old allocs/op new allocs/op delta
Init-8 39.0 ± 0% 1.0 ± 0% ~ (zero variance)
Whole sortByRFC6724 function benchmark:
name old time/op new time/op delta
SortByRFC6724/0-8 463ns ± 3% 303ns ± 4% -34.72% (p=0.000 n=5+5)
SortByRFC6724/1-8 481ns ± 8% 306ns ± 1% -36.46% (p=0.000 n=5+5)
SortByRFC6724/2-8 470ns ± 4% 307ns ± 4% -34.77% (p=0.000 n=5+5)
SortByRFC6724/3-8 567ns ± 3% 367ns ± 3% -35.28% (p=0.000 n=5+5)
SortByRFC6724/4-8 918ns ± 3% 560ns ± 2% -38.93% (p=0.000 n=5+5)
Updates #54032
Change-Id: Ic18df1ea73805cb184c6ceb73470ca7f0b922032
Reviewed-on: https://go-review.googlesource.com/c/go/+/419356
Reviewed-by: Heschi Kreinick <heschi@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Daniel Martí <mvdan@mvdan.cc>
Currently, there are 3 functions returning Linux kernel version numbers.
Two of them are identical:
- in net, initially added by commit 0a9dd47dd817904e;
- in internal/poll, initially added by commit 1c7650aa93bd53;
(both were later fixed by commit 66c0264506).
The third one is a more complex, regexp-based implementation in
runtime/pprof, which is only used for a test.
Instead of adding one more, let's consolidate existing ones.
Remove the complex implementation, and move the simple one into
internal/syscall/unix. Use it from all the three places mentioned above.
Change-Id: I4a34d9ca47257743c16def30e4dd634e36056091
Reviewed-on: https://go-review.googlesource.com/c/go/+/424896
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Kirill Kolyshkin <kolyshkin@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Also, consistently use declaration: var buf strings.Builder.
We don't change exported signatures to match go/types (where we
can't change the exported signatures for backward-compatibility).
Change-Id: I75350886aa231889ae2fd5c4008dd4be9ed6e09f
Reviewed-on: https://go-review.googlesource.com/c/go/+/428094
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
The kernel knob /proc/sys/kernel/unprivileged_userns_clone is
only available in Debian (and Ubuntu) kernels, so if the tests
are run on e.g. Fedora, skipUnprivilegedUserClone() skips a lot
of tests.
Modify it to treat ENOENT as "it should work".
Change-Id: I959201ede139ede989cc8ab646c9bf51e0539ada
Reviewed-on: https://go-review.googlesource.com/c/go/+/417694
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Kirill Kolyshkin <kolyshkin@gmail.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Previously when a printer had a large buffer we dropped both
the buffer and the printer. There is no need to drop the printer
in this case, as a printer with a nil buffer is valid. So we
just drop the buffer and recycle the printer anyway.
This saves one allocation in case the buffer is over the limit.
Also tighten some of the tests for other unrelated cases.
Change-Id: Iba1b6a71ca4691464b8c68ab0b6ab0d4d5d6168c
Reviewed-on: https://go-review.googlesource.com/c/go/+/427395
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Rob Pike <r@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Reviewed-by: Rob Pike <r@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently, the itabs section for runtime dictionaries includes its own
redundant *runtime._type pointers for typ and iface, which were
sometimes necessary. This simplified the initial implementation, but
is a little wasteful of space when the same type or interface appeared
across multiple (typ, iface) pairs.
This CL instead reuses the pointers from the rtypes section.
Change-Id: I48448515c319c0403c1a8e7706794d443176f0a4
Reviewed-on: https://go-review.googlesource.com/c/go/+/427754
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Convert subtraction from const to a negated ADDI with negative const
value, where possible. At worst this avoids a register load and uses
the same number of instructions. At best, this allows for further
optimisation to occur, particularly where equality is involved.
For example, this sequence:
li t0,-1
sub t1,t0,a0
snez t1,t1
Becomes:
addi t0,a0,1
snez t0,t0
Removes more than 2000 instructions from the Go binary on linux/riscv64.
Change-Id: I68f3be897bc645d4a8fa3ab3cef165a00a74df19
Reviewed-on: https://go-review.googlesource.com/c/go/+/426263
Reviewed-by: Meng Zhuo <mzh@golangcn.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Joel Sing <joel@sing.id.au>
The FNES and FNED instructions are pseudo-instructions, which the
assembler expands to FEQS/NEG or FEQD/NEG - if we're comparing the
result via a branch instruction, we can avoid an instruction by
negating both the branch comparision and the floating point
comparision.
This only removes a handful of instructions from the Go binary,
however, it will provide benefit to floating point intensive code.
Change-Id: I4e3124440b7659acc4d9bc9948b755a4900a422f
Reviewed-on: https://go-review.googlesource.com/c/go/+/426261
Reviewed-by: Meng Zhuo <mzh@golangcn.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Joel Sing <joel@sing.id.au>
Run-TryBot: Meng Zhuo <mzh@golangcn.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
The stkframe struct and its methods are strewn across different source
files. Since they actually have a pretty coherent theme at this point,
migrate it all into a new file, stkframe.go. There are no code changes
in this CL.
For #54466, albeit rather indirectly.
Change-Id: Ibe53fc4b1106d131005e1c9d491be838a8f14211
Reviewed-on: https://go-review.googlesource.com/c/go/+/424516
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Austin Clements <austin@google.com>
Auto-Submit: Austin Clements <austin@google.com>
Currently, stkframe.arglen and stkframe.argmap are populated by
gentraceback under a particular set of circumstances. But because they
can be constructed from other fields in stkframe, they don't need to
be computed eagerly at all. They're also rather misleading, as they're
only part of computing the actual argument map and most callers should
be using getStackMap, which does the rest of the work.
This CL drops these fields from stkframe. It shifts the functions that
used to compute them, getArgInfoFast and getArgInfo, into
corresponding methods stkframe.argBytes and stkframe.argMapInternal.
argBytes is expected to be used by callers that need to know only the
argument frame size, while argMapInternal is used only by argBytes and
getStackMap.
We also move some of the logic from getStackMap into argMapInternal
because the previous split of responsibilities didn't make much sense.
This lets us return just a bitvector from argMapInternal, rather than
both a bitvector, which carries a size, and an "actually use this
size".
The getArgInfoFast function was inlined before (and inl_test checked
this). We drop that requirement from stkframe.argBytes because the
uses of this have shifted and now it's only called from heap dumping
(which never happens) and conservative stack frame scanning (which
very, very rarely happens).
There will be a few follow-up clean-up CLs.
For #54466. This is a nice clean-up on its own, but it also serves to
remove pointers from the traceback state that would eventually become
troublesome write barriers once we stack-rip gentraceback.
Change-Id: I107f98ed8e7b00185c081de425bbf24af02a4163
Reviewed-on: https://go-review.googlesource.com/c/go/+/424514
Run-TryBot: Austin Clements <austin@google.com>
Auto-Submit: Austin Clements <austin@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently, when traceback jumps from the system stack to a user stack
(e.g., during profiling tracebacks), it leaves gp pointing at the g0.
This is currently harmless since it's only used during profiling, so
the code paths in gentraceback that care about gp aren't used, but
it's really confusing and would certainly break if _TraceJumpStack
were ever used in a context other than profiling.
Fix this by updating gp to point to the user g when we switch stacks.
For #54466.
Change-Id: I1541e004667a52e37671803ce45c91d8c5308830
Reviewed-on: https://go-review.googlesource.com/c/go/+/424257
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Austin Clements <austin@google.com>
The f funcInfo argument is always the same as frame.fn, so we don't
need to pass it. I suspect that was there to make the signatures of
getArgInfoFast and getArgInfo more similar, but it's not necessary.
For #54466.
Change-Id: Idc717f4df09e97cad49d52c5b7edf28090908cba
Reviewed-on: https://go-review.googlesource.com/c/go/+/424255
Run-TryBot: Austin Clements <austin@google.com>
Auto-Submit: Austin Clements <austin@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently, gentraceback tracks the closure context of the outermost
frame. This used to be important for "unstarted" calls to reflect
function stubs, where "unstarted" calls are either deferred functions
or the entry-point of a goroutine that hasn't run. Because reflect
function stubs have a dynamic argument map, we have to reach into
their closure context to fetch to map, and how to do this differs
depending on whether the function has started. This was discovered in
issue #25897.
However, as part of the register ABI, "go" and "defer" were made much
simpler, and any "go" or "defer" of a function that takes arguments or
returns results gets wrapped in a closure that provides those
arguments (and/or discards the results). Hence, we'll see that closure
instead of a direct call to a reflect stub, and can get its static
argument map without any trouble.
The one case where we may still see an unstarted reflect stub is if
the function takes no arguments and has no results, in which case the
compiler can optimize away the wrapper closure. But in this case we
know the argument map is empty: the compiler can apply this
optimization precisely because the target function has no argument
frame.
As a result, we no longer need to track the closure context during
traceback, so this CL drops all of that mechanism.
We still have to be careful about the unstarted case because we can't
reach into the function's locals frame to pull out its context
(because it has no locals frame). We double-check that in this case
we're at the function entry.
I would prefer to do this with some in-code PCDATA annotations of
where to find the dynamic argument map, but that's a lot of mechanism
to introduce for just this. It might make sense to consider this along
with #53609.
Finally, we beef up the test for this so it more reliably forces the
runtime down this path. It's fundamentally probabilistic, but this
tweak makes it better. Scheduler testing hooks (#54475) would make it
possible to write a reliable test for this.
For #54466, but it's a nice clean-up all on its own.
Change-Id: I16e4f2364ba2ea4b1fec1e27f971b06756e7b09f
Reviewed-on: https://go-review.googlesource.com/c/go/+/424254
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Auto-Submit: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
In go.dev/cl/421821, I included a hack to force OCONVNOP back to
OCONVIFACE for conversions involving shape types and non-empty
interfaces. The comment correctly noted that this was only needed for
conversions between non-identical types, but the code was conservative
and applied to even conversions between identical types.
This CL adds an extra bool to record whether the conversion is between
identical types, so we can keep OCONVNOP instead of forcing back to
OCONVIFACE. This has a small improvement to generated code, because we
no longer need a convI2I call (as demonstrated by codegen/ifaces.go).
But more usefully, this is relevant to pruning unnecessary itab slots
in runtime dictionaries (next CL).
Change-Id: I94f89e961cd26629b925037fea58d283140766ff
Reviewed-on: https://go-review.googlesource.com/c/go/+/427678
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This CL changes the heuristic used to determine whether we can inline a
struct equality check or if we must generate a function and call that
function for equality.
The old method was to count struct fields, but this can lead to poor
in lining decisions. We should really be determining the cost of the
equality check and use that to determine if we should inline or generate
a function.
The new benchmark provided in this CL returns the following when compared
against tip:
```
name old time/op new time/op delta
EqStruct-32 2.46ns ± 4% 0.25ns ±10% -89.72% (p=0.000 n=39+39)
```
Fixes#38494
Change-Id: Ie06b80a2b2a03a3fd0978bcaf7715f9afb66e0ab
GitHub-Last-Rev: e9a18d9389
GitHub-Pull-Request: golang/go#53326
Reviewed-on: https://go-review.googlesource.com/c/go/+/411674
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
Auto-Submit: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
This CL optimizes RotateLeft8/16 on arm64.
For 16 bits, we form a 32 bits register by duplicating two 16 bits
registers, then use RORW instruction to do the rotate shift.
For 8 bits, we just use LSR and LSL instead of RORW because the code is
simpler.
Benchmark Old ThisCL delta
RotateLeft8-46 2.16 ns/op 1.73 ns/op -19.70%
RotateLeft16-46 2.16 ns/op 1.54 ns/op -28.53%
Change-Id: I09cde4383d12e31876a57f8cdfd3bb4f324fadb0
Reviewed-on: https://go-review.googlesource.com/c/go/+/420976
Reviewed-by: Keith Randall <khr@google.com>
Auto-Submit: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
For the following code case:
var x uint64
x >> (shift & 63)
We can directly genereta `x >> shift` on arm64, since the hardware will
only use the bottom 6 bits of the shift amount.
Benchmark old time/op new time/op delta
ShiftArithmeticRight-8 0.40ns 0.31ns -21.7%
Change-Id: Id58c8a5b2f6dd5c30c3876f4a36e11b4d81e2dc9
Reviewed-on: https://go-review.googlesource.com/c/go/+/425294
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
This seems more sensible than the func keyword. With this change,
go/types uses the same error position as types2 and we can narrow
the error tolerance a bit.
(The types2 change doesn't change its position, but it makes the
code clearer and symmetric to go/types.)
Change-Id: Iedea7c80caa7239a4343c8748cb779ec545e84d5
Reviewed-on: https://go-review.googlesource.com/c/go/+/427775
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
On AIX when external linking, for some symbols we need to add
dummy references to prevent the external linker from discarding
them. Currently we add the reference unconditionally. But if the
symbol doesn't exist, the linking fails in a later stage for
generating external relocation of a nonexistent symbol. The
symbols are special symbols that almost always exist, except that
go:buildid may not exist if the linker is invoked without the
-buildid flag. The go command invokes the linker with the flag, so
this can only happen with manual linker invocation. Specifically,
test/run.go does this in some cases.
Fix this by checking the symbol existence before adding the
reference. Re-enable tests on AIX.
Perhaps the linker should always emit a dummy buildid even if the
flag is not set...
Fixes#54814.
Change-Id: I43d81587151595309e189e38960cbda9a1c5ca32
Reviewed-on: https://go-review.googlesource.com/c/go/+/427620
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
So it won't be visible outside of runtime package. There are changes to
make tests happy:
- For test/directive*.go files, using "go:noinline" for testing misplaced
directives instead.
- Restrict test/fixedbugs/bug515.go for gccgo only.
- For test/notinheap{2,3}.go, using runtime/cgo.Incomplete for marking
the type as not-in-heap. Though it's somewhat clumsy, it's the easiest
way to keep the test errors for not-in-heap types until we can cleanup
further.
- test/typeparam/mdempsky/11.go is about defined type in user code marked
as go:notinheap, which can't happen after this CL, though.
Fixes#46731
Change-Id: I869f5b2230c8a2a363feeec042e7723bbc416e8e
Reviewed-on: https://go-review.googlesource.com/c/go/+/421882
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Joedian Reid <joedian@golang.org>
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
This CL moves the directories check, examples, fixedbugs, and spec
from inside go/types/testdata to internal/types/testdata. Except
for the directory adjustments to check_test.go files, this is a
pure file move.
With this CL, both type checkers now share identical tests in an
independent location.
Fixes#54511.
Change-Id: Ib335692d927e93867a158b338f105c2b87e74dbe
Reviewed-on: https://go-review.googlesource.com/c/go/+/427674
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
CL 425735 consolidated the testdata/check/shifts.go files between
go/types and types2. Because some shifts don't work correctly with
types2, the corresponding tests were disabled in the shared file.
Make sure we keep testing those shifts for go/types by adding a
local test file.
For #52080.
For #54511.
Change-Id: I53507e535bf83b204eaf18fc6c2efefcebf5ebf7
Reviewed-on: https://go-review.googlesource.com/c/go/+/426661
Reviewed-by: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Since the fixedbugs tests are now identical between the two type checkers,
remove the local copy of the fixedbugs tests and (for now) use the tests
in go/types/testdata/fixedbugs instead. Eventually we may decide to move
all tests out of the type checker directories and place them in a
shared space (e.g. internal/types/testdata).
For #54511.
Change-Id: I451c20c784710c36fa50b1d24ac97af554c572af
Reviewed-on: https://go-review.googlesource.com/c/go/+/426658
Auto-Submit: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Use the go/types version of testdata/fixedbugs tests where diffs
are only in the error positions (the types2 test harness allows
for some position tolerance). Consolidate files where there are
other minor differences.
Add files to respective directories if they only existed for
one of the type checkers.
Move types2-only test issue47996.go out of testdata/fixedbugs
into testdata. Making it work for both type checkers requires
some more work.
With this CL, the testdata/fixedbugs files are identical between
the two type checkers.
For #54511.
Change-Id: I0d67f0db75ad1743c62da9181a6d0032c8bdb728
Reviewed-on: https://go-review.googlesource.com/c/go/+/427236
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Report a syntax error if the first element of a type instance is
not actually a type (but some other expression), rather then relying
on the type checker error in this case. This matches the behavior of
go/parser. Adjust the corresponding types2 test case.
For #54511.
Change-Id: Ia82b3a7d444738c56955ce6c15609470b3431fd1
Reviewed-on: https://go-review.googlesource.com/c/go/+/426657
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Rename .../issue46404.go1 to .../issue46404.go so that it is
not skipped anymore when running tests, and copy for types2.
Disable the code for now due to a difference in error
reporting due to the slightly different handling of index
expressions. This allows us to make progress with test
consolidation.
For #54511.
Change-Id: Ib5c9ffa49b1b24ec680ddb5001bc3dcb1df7eb1f
Reviewed-on: https://go-review.googlesource.com/c/go/+/426656
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
This is a port of CL 427156 from the syntax package's parser
to go/parser.
While at it, remove an unused token.Pos parameter from
parseSpecFunction and dependent declarations.
Also, consolidate the respective test file.
For #54511.
Change-Id: Id6a28eb3d23a46fa5fa1d85d2c4e634b7015513c
Reviewed-on: https://go-review.googlesource.com/c/go/+/427157
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Instead of checking at parse-time that the LHS of a short variable
declaration contains only identifiers, leave the check to the the
type checker which tests this already.
This removes a duplicate error and matches the behavior of the
syntax package.
For #54511.
Change-Id: I4c68f2bd8a0e015133685f9308beb98e714a83fc
Reviewed-on: https://go-review.googlesource.com/c/go/+/426476
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
The syntax for go and defer specifies an arbitrary expression, not
a call; the call requirement is spelled out in prose. Don't to the
call check in the parser; instead move it to the type checker. This
is simpler and also allows the type checker to check expressions that
are not calls, and avoid "not used" errors due to such expressions.
We would like to make the same change in go/parser and go/types
but the change requires Go/DeferStmt nodes to hold an ast.Expr
rather than an *ast.CallExpr. We cannot change that for backward-
compatibility reasons. Since we don't test this behavior for the
type checkers alone (only for the compiler), we get away with it
for now.
Follow-up on CL 425675 which introduced the extra errors in the
first place.
Change-Id: I90890b3079d249bdeeb76d5673246ba44bec1a7b
Reviewed-on: https://go-review.googlesource.com/c/go/+/425794
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
Since the check tests are now identical between the two type checkers,
remove the local copy of the check tests and (for now) use the tests
in go/types/testdata/check instead. Eventually we may decide to move
all tests out of the type checker directories and place them in a
shared space (e.g. internal/types/testdata).
For #54511.
Change-Id: Id3a97593f6c705c5eda4566089ddc7aeb7b47337
Reviewed-on: https://go-review.googlesource.com/c/go/+/425736
Reviewed-by: Alan Donovan <adonovan@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Use the go/types version of testdata/check tests where the diffs
are only in the error positions (the types2 test harness allows
for some position tolerance). Consolidate files where there are
other minor differences.
Comment out a couple of tests that are different between the two
type checkers.
With this CL, the testdata/check files are identical between the
two type checkers.
For #54511.
Change-Id: Ibdff2ca3ec9bdaca3aa84029a7883bb83d2d2060
Reviewed-on: https://go-review.googlesource.com/c/go/+/425735
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
If the go/defer syntax is bad, using a fake CallExpr may produce
a follow-on error in the type checker. Instead store a BadExpr
in the syntax tree (since an error has already been reported).
Adjust various tests.
For #54511.
Change-Id: Ib2d25f8eab7d5745275188d83d11620cad6ef47c
Reviewed-on: https://go-review.googlesource.com/c/go/+/425675
Reviewed-by: Alan Donovan <adonovan@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Remove the code that verifies that an expression is a type or non-type
expression. For one, it cannot be done perfectly accurate
(e.g., consider *p which could be an indirection or a pointer type),
it also unnecessarily slows down parsing. It's simpler to leave the
verification to the type checker which has all the information needed.
Remove short compiler tests that tested the expression/type property.
Adjust a couple of go/types tests which now trigger because the parser
doesn't complain anymore.
Change file for benchmark from "parser.go" to "../printer/nodes.go"
to avoid a moving target when benchmarking.
The parser may be marginally faster when tested on nodes.go:
name old time/op new time/op delta
ParseOnly-12 1.35ms ± 0% 1.31ms ± 0% ~ (p=0.100 n=3+3)
name old speed new speed delta
ParseOnly-12 39.9MB/s ± 0% 41.0MB/s ± 0% ~ (p=0.100 n=3+3)
For #54511.
Change-Id: I9a32c24c2c6e843c3d1af4587651c352f378b490
Reviewed-on: https://go-review.googlesource.com/c/go/+/425716
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Alan Donovan <adonovan@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Amend the "mkall.sh -syscalls" implementation to
- prepend ./ before mksyscalls.pl;
- accept the optional file list argument.
This is a preparation for CL 416115.
Change-Id: Ib4dc2b4aa0d2dd22a256414864e92f2d2fd957a5
Reviewed-on: https://go-review.googlesource.com/c/go/+/423676
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Tobias Klauser <tobias.klauser@gmail.com>
Instead of passing the original length and the new length, pass
the new length and the length increment. Also use the new length
in all the post-growslice calculations so that the original length
is dead and does not need to be spilled/restored around the growslice.
old: growslice(typ, oldPtr, oldLen, oldCap, newLen) (newPtr, newLen, newCap)
new: growslice(oldPtr, newLen, oldCap, inc, typ) (newPtr, newLen, newCap)
where inc = # of elements added = newLen-oldLen
Also move the element type to the end of the call. This makes register
allocation more efficient, as oldPtr and newPtr can often be in the
same register (e.g. AX on amd64) and thus the phi takes no instructions.
Makes the go binary 0.3% smaller.
Change-Id: I7295a60227dbbeecec2bf039eeef2950a72df760
Reviewed-on: https://go-review.googlesource.com/c/go/+/418554
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
We have supported passing lists of arguments to the compiler and linker
for some time, since https://go.dev/issue/18468 was fixed.
The reason behind it is that some systems like Windows have relatively
small limits for commands, and some Go packages contain many source files.
This wasn't done for other Go toolchain programs like cgo and asm,
as there wasn't an initial need for it. A TODO was left for them.
The need has now arisen in the form of a bug report for a build of a
large Go package involving cgo.
Do asm as well, which could be triggered by lots of asm files.
I rebuilt Go itself with some basic logging to tell if any other
commands were being run with moderately large command lengths.
I only found one other: gcc being invoked with 300-500 bytes.
I didn't spot any length close to 1KiB, and we can't safely assume that
a user's CC compiler supports these "response files", so leave that as
another TODO for the future. Just like cgo and asm, we can revisit this
if any user reports a bug on the issue tracker.
Fixes#47235.
Change-Id: Ifcc099d7c0dfac3ed2c4e9e7a2d6e3d69b0ccb63
Reviewed-on: https://go-review.googlesource.com/c/go/+/427015
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
Reviewed-by: Keith Randall <khr@golang.org>
Auto-Submit: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Since when go/types,types2 do not know about build constraints, and
runtime/cgo.Incomplete is only available on platforms that support cgo.
These tests are also failing on aix with failure from linker, so disable
them on aix to make builder green. The fix for aix is tracked in #54814
Updates #46731
Updates #54814
Change-Id: I5d6f6e29a8196efc6c457ea64525350fc6b20309
Reviewed-on: https://go-review.googlesource.com/c/go/+/427394
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
v.SetIterXXX(i) is semantically identical to v.Set(i.XXX()).
If the latter panics for unexported values, so should the former.
This change may breaking some programs, but the change is justified
under the "Go 1 and the Future of Go Programs" document because
the "library has a bug that violates the specification".
In this case, the "reflect" package does not accurately match
the behavior of the Go language specification.
Also, this API was recently released, so the number of users
who could be depending on this behavior is hopefully lower.
Fixes#54628
Change-Id: If86ede51f286e38093f6697944c089f616525115
Reviewed-on: https://go-review.googlesource.com/c/go/+/425184
Auto-Submit: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: David Chase <drchase@google.com>
In go.dev/cl/419674 I added a mechanism to the inliner to allow
inlining to fail gracefully when a function body is missing, but I
missed we already have a mechanism for that: typecheck.HaveInlineBody.
This CL makes it overridable so that unified IR can plug in its
appropriate logic, like it does with the logic for building the
ir.InlinedCallExpr node.
While here, rename inline.NewInline to inline.InlineCall, because the
name "NewInline" is now a misnomer since we initialize it to oldInline
(now named oldInlineCall).
Change-Id: I4e65618d3725919f69e6f43cf409699d20fb797c
Reviewed-on: https://go-review.googlesource.com/c/go/+/427234
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
The current mknode has a few problems:
1) It tends not to run successfully if the tree is in a broken state.
2) It requires that it be run by the go tool in the tree (somewhat related to 1)
3) It requires setting GOROOT
4) It imports code outside the tree (x/packages)
This makes mknode.go very fragile. In particular, I've spent lots of
time fighting mknode when adding or removing code, related to 1.
Rewrite to just use go/ast and friends. No typechecking, no importing,
etc. It can run with any go version, it doesn't need to be the one
corresponding to the code in which it is run. (e.g. you can use go
1.16 to run mknode). It will work as long as the ir package is parseable.
When run, it generates identical output to the old mknode.
Fixes#53959
Change-Id: I5ce0b55572ebcd2fcd11af57a5f29bbf9fa4ed33
Reviewed-on: https://go-review.googlesource.com/c/go/+/418375
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Currently we use a full cmpstring to do the comparison for each
split in the binary search for a string switch.
Instead, split by comparing a single byte of the input string with a
constant. That will give us a much faster split (although it might be
not quite as good a split).
Fixes#53333
R=go1.20
Change-Id: I28c7209342314f367071e4aa1f2beb6ec9ff7123
Reviewed-on: https://go-review.googlesource.com/c/go/+/414894
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
for i := 0; i < 9; i += 3
Currently we compute bounds of [0,8]. Really we know that it is [0,6].
CL 415874 computed the better bound as part of overflow detection.
This CL just incorporates that better info to the prove pass.
R=go1.20
Change-Id: Ife82cc415321f6652c2b5d132a40ec23e3385766
Reviewed-on: https://go-review.googlesource.com/c/go/+/415937
Reviewed-by: Heschi Kreinick <heschi@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
A condition check was added to parse.go in CL 405542 to prevent
usage of scaled operands on ppc64. However while trying to improve
the error notification message, an if-condition was left out by
oversight. This CL corrects that.
Change-Id: I8cef3dd194c75343354ffe888b5e639e694badde
Reviewed-on: https://go-review.googlesource.com/c/go/+/426994
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Archana Ravindar <aravind5@in.ibm.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Previously the first operand of MSR could be $0, which would be
converted to the ZR register. This is prohibited by CL 404316,
this CL restores this instruction format.
Change-Id: I5b5be59e76aa58423a0fb96942d1b2a9de62e311
Reviewed-on: https://go-review.googlesource.com/c/go/+/426198
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Run-TryBot: Eric Fang <eric.fang@arm.com>
If x+delta cannot overflow/underflow, we can derive:
x+delta < x if delta<0 (this CL included)
x+delta > x if delta>0 (this CL not included due to
a recursive stack overflow)
Remove 95 bounds checks during ./make.bat
Fixes#51622
Change-Id: I60d9bd84c5d7e81bbf808508afd09be596644f09
Reviewed-on: https://go-review.googlesource.com/c/go/+/406175
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
CL 327871 changes methodWrapper to always perform inlining after global
escape analysis. However, inlining the method may reveal closures, which
require walking all function bodies to decide whether to capture free
variables by value or by ref.
To fix it, just not doing inline if the method contains any closures.
Fixes#53702
Change-Id: I4b0255b86257cc6fe7e5fafbc545cc5cff9113e1
Reviewed-on: https://go-review.googlesource.com/c/go/+/426334
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Shape-based stenciling in unified IR is done by converting type argument
to its underlying type. So it agressively check that type argument is
not a TFORW. However, for recursive instantiated type argument, it may
still be a TFORW when shapifying happens. Thus the assertion failed,
causing the compiler crashing.
To fix it, just allow fully instantiated type when shapifying.
Fixes#54512Fixes#54722
Change-Id: I527e3fd696388c8a37454e738f0324f0c2ec16cb
Reviewed-on: https://go-review.googlesource.com/c/go/+/426335
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
When a base+displacement kind of operand is given in an index-mode
instruction, the assembler does not flag it as an invalid instruction
causing the user to get an incorrect encoding of that instruction
leading to incorrect execution of the program.
Enable assembler to recognize valid and invalid operands used in index
mode instructions by classifying SOREG type into two further types
XOREG (used uniquely in index addressing mode instructions) and SOREG
for instructions working on base+displacement operands.
Also cleaned up usage of obj.Addr.Scale on PPC64.
Change-Id: Ib4d84343ae57477c6c074f44c4c2749496e11b91
Reviewed-on: https://go-review.googlesource.com/c/go/+/405542
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Archana Ravindar <aravind5@in.ibm.com>
When spilling arg to stack or register, if it's a newly created value,
the arg position should be preserved. Otherwise, we may end up using
position information from deadcode lines.
This fixes the minimized test case in #54625 by mdempsky@, and make
building std successfully. However, the inline trees for these tests
still be corrupted:
- fixedbugs/issue53982.go
- typeparam/issue47775.go
- typeparam/issue47775b.go
- typeparam/issue49432.go
We probably still mess up the inline position somewhere else.
Updates #54625
Change-Id: I0d87e26b9ab451b85b6e79787da74a2b79a16209
Reviewed-on: https://go-review.googlesource.com/c/go/+/425785
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
CL 426079 started checking errors from RecentTag.
Unfortunately, we forgot to run "-longtest" SlowBots, and it turns out
to have broken non-short tests for non-git VCS implementations,
because those don't implement the RecentTag method.
Updates #53935.
Change-Id: I5935f2f4b3f684515e99e8bf70a840154c36249f
Reviewed-on: https://go-review.googlesource.com/c/go/+/426495
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The existing implementation creates more branches with more states: -1, 0, 1,
which makes it not very intuitive to understand, let's use sync.Once and boolean
instead to make it more straightforward.
Change-Id: I05766e5fdf7dba37d6565f84d3db4373f9342fe5
Reviewed-on: https://go-review.googlesource.com/c/go/+/425880
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Previously the first operand of FMOVD and FMOVS could be $0, which
would be converted to the ZR register. This is prohibited by CL 404316,
also it broken the encoding of "FMOVD/FMOVS ZR, Rn", this CL restores
this instruction format and fixes the encoding issue.
Fixes#54655.
Fixes#54729.
Change-Id: I9c42cd41296bed7ffd601609bd8ecaa27d11e659
Reviewed-on: https://go-review.googlesource.com/c/go/+/425188
Run-TryBot: Eric Fang <eric.fang@arm.com>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Correctly set this flag while parsing the syscall result.
The FlagUp flag can not distinguish the following situations:
1. interface is plugged, automatically up, and in running(UP) state
2. interface is not plugged, administratively or manually set to up,
but in DOWN state
So, We can't distinguish the state of a NIC by the FlagUp flag alone.
Fixes#53482
Change-Id: I43796bea1a7f72d1fddfef914efe603c81995e1b
GitHub-Last-Rev: 686b5d888e
GitHub-Pull-Request: golang/go#53484
Reviewed-on: https://go-review.googlesource.com/c/go/+/413454
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ryan Schuster <shuey19831@gmail.com>
Reviewed-by: Jianwei Mao <maojianwei2020@gmail.com>
Reviewed-by: Damien Neil <dneil@google.com>
A recent change was made for ppc64x to treat ANDCCconst as
a tuple, allowing ANDconst to be removed from the list
of ops. Included in that change were some improvements to the
rules to avoid some extra code, mainly the elimination of a
cmp 0 following an andi. and in some cases the following
isel. While those changes worked for most cases, in a few
cases some extra unnecessary code was generated.
Currently the snippet appears in archive/zip.(*FileHeader).Mode:
ANDCC R4,$1,R5 // andi. r5,r4,1
ANDCC R4,$16,R5 // andi. r5,r4,16
CMPW R5,R0 // cmpw r5,r0
ADDIS $0,$-32768,R5 // lis r5,-32768
OR R5,$511,R5 // ori r5,r5,511
MOVD $438,R6 // li r6,438
ISEL $2,R6,R5,R5 // isel r5,r6,r5,eq
MOVD $-147,R6 // li r6,-147
AND R6,R5,R6 // and r6,r5,r6
ANDCC R4,$1,R4 // andi. r4,r4,1
ISEL $2,R5,R6,R4 // isel r4,r5,r6,eq
The first ANDCC is never used and should not be there.
From the ssa.html file, the scheduler is not putting the Select1
close to the ISEL, which results in the flag being clobbered
before it can be used. By changing the score for a Select0 or Select1
with type Flags, the extra ANDCC does not occur.
Change-Id: I82f4bc7c02afb1c2b1c048dc6995e0b3f9363fb3
Reviewed-on: https://go-review.googlesource.com/c/go/+/424294
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Paul Murphy <murp@ibm.com>
Both FileServer and NewFileTransport can try to seek a file, specifically
when MIME type sniffing is performed. This can be somewhat surprising to an
implementer of an fs.FS, as their filesystem will appear to work until a
user tries to access a file with an unrecognized extension (which requires
type sniffing and therefore seeking). With FileServer, this results in a
"seeker can't seek" message, which is not very clear for the developer.
The issue arises because fs.FS does not require Seek, while http.FileSystem
does. Therefore, this change adds a line to the documentation of net/http's
adapter function mentioning the requirement.
Change-Id: Ieb955b7a7f34e2be39dd696cb712513c70100b3a
GitHub-Last-Rev: fddccdae36
GitHub-Pull-Request: golang/go#48781
Reviewed-on: https://go-review.googlesource.com/c/go/+/353874
Reviewed-by: Damien Neil <dneil@google.com>
Auto-Submit: Damien Neil <dneil@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
To make it easier to extract the HTTP/2 error code (if any) from
net/http errors, implement an As method on the vendored copy of
golang.org/x/net/http2.StreamError. The new As method lets users work
with the vendored error type as though it were the x/net/http2
StreamError.
Fixes#53896.
Change-Id: Ib18eb428adc05a3c0e19a946ece936e2378e1c7c
Reviewed-on: https://go-review.googlesource.com/c/go/+/425104
Run-TryBot: Damien Neil <dneil@google.com>
Auto-Submit: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: David Chase <drchase@google.com>
In typebits.Set we check that the offset is a multiple of the
alignment, which makes perfect sense. But for values like
atomic.Int64, which has 8-byte alignment even on 32-bit platforms
(i.e. the alignment is larger than PtrSize), if it is on stack it
may be under-aligned, as the stack frame is only PtrSize aligned.
Normally we would prevent such values on stack, as the escape
analysis force values with higher alignment to heap. But for a
composite literal assignment like x = AlignedType{...}, the
compiler creates an autotmp for the RHS then copies it to the LHS.
The autotmp is on stack and may be under-aligned. Currently this
may cause an ICE in the typebits.Set check.
This CL makes it align the _offset_ of the autotmp to 8 bytes,
which satisfies the check. Note that this is actually lying: the
actual address at run time may not necessarily be 8-byte
aligned as we only align SP to 4 bytes.
The under-alignment is probably okay. The only purpose for the
autotmp is to copy the value to the LHS, and the copying code we
generate (at least currently) doesn't care the alignment beyond
stack alignment.
Fixes#54638.
Change-Id: I13c16afde2eea017479ff11dfc24092bcb8aba6a
Reviewed-on: https://go-review.googlesource.com/c/go/+/425256
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
There are only two tests that rely on the "ERROR HERE" markers;
yet those tests are trivialy adjustable (by adding an explicit
semicolon) such that they can just use the "ERROR" markers.
For #54511.
Change-Id: Idbb96ca8d35ae2584d195a4ac7c92640b8b492c5
Reviewed-on: https://go-review.googlesource.com/c/go/+/425674
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Alan Donovan <adonovan@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
This change updates the gctrace docs to include stacks and globals in
the format line, and prints lastStackScan for "# MB stacks" instead of
maxStackScan, which is more accurate.
Fixes#54649.
Change-Id: Ibff2c390c9c9bf2b24b5b4e98ca346cc98d7cb2e
Reviewed-on: https://go-review.googlesource.com/c/go/+/425366
Run-TryBot: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
MarshalPKIXPublicKey, CreateCertificate, CreateCertificateRequest,
MarshalECPrivateKey, and MarshalPKCS8PrivateKey started raising a panic
when encoding an invalid ECDSA key in Go 1.19. Since they have an error
return value, they should return an error instead.
Fixes#54288
Change-Id: Iba132cd2f890ece36bb7d0396eb9a9a77bdb81df
Reviewed-on: https://go-review.googlesource.com/c/go/+/422298
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
When inlining function calls, we rewrite the position information on
all of the nodes to keep track of the inlining context. This is
necessary so that at runtime, we can synthesize additional stack
frames so that the inlining is transparent to the user.
However, for function literals, we *don't* want to apply this
rewriting to the underlying function. Because within the function
literal (when it's not itself inlined), the inlining context (if any)
will have already be available at the caller PC instead.
Unified IR was already getting this right in the case of user-written
statements within the function literal, which is what the unit test
for #46234 tested. However, it was still using inline-adjusted
positions for the function declaration and its parameters, which
occasionally end up getting used for generated code (e.g., loading
captured values from the closure record).
I've manually verified that this fixes the hang in
https://go.dev/play/p/avQ0qgRzOgt, and spot-checked the
-d=pctab=pctoinline output for kube-apiserver and kubelet and they
seem better.
However, I'm still working on a more robust test for this (hence
"Updates" not "Fixes") and internal assertions to verify that we're
emitting correct inline trees. In particular, there are still other
cases (even in the non-unified frontend) where we're producing
corrupt (but at least acyclic) inline trees.
Updates #54625.
Change-Id: Iacfd2e1eb06ae8dc299c0679f377461d3d46c15a
Reviewed-on: https://go-review.googlesource.com/c/go/+/425395
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Given a composite literal type S, rather than always printing
(S literal) for a composite literals, print S{} if the literal
has no elements, and print S{…} as a short form (suitable for
error messages) if there are elements. This matches types2 and
also Go1.17 compiler behavior (except that the original compiler
would print ... rather than …). Using … rather than ... makes
it clearer that we don't have real Go syntax, and it's also more
compact.
For #54511.
Change-Id: I5991e8060232f16ecbf4a1fe4ae091598fc76b68
Reviewed-on: https://go-review.googlesource.com/c/go/+/425006
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
On LR architectures, morestack (and morestack_noctxt) are called
with a special calling convention, where the caller doesn't save
LR on stack but passes it as a register, which morestack will save
to g.sched.lr. The stack unwinder currently doesn't understand it,
and would fail to unwind from it. morestack already writes SP (as
it switches stack), but morestack_noctxt (which tailcalls
morestack) doesn't. If a profiling signal lands right in
morestack_noctxt, the unwinder will try to unwind the stack and
go off, and possibly crash.
Marking morestack_noctxt SPWRITE stops the unwinding.
Ideally we could teach the unwinder about the special calling
convention, or change the calling convention to be less special
(so the unwinder doesn't need to fetch a register from the signal
context). This is a stop-gap solution, to stop the unwinder from
crashing.
Fixes#54332.
Change-Id: I75295f2e27ddcf05f1ea0b541aedcb9000ae7576
Reviewed-on: https://go-review.googlesource.com/c/go/+/425396
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Some tests in misc/cgo/testsanitizers had been disabled on ppc64le
until recently, due to an intermittent error in the tsan tests,
with the goal of trying to understand the failure.
After further investigation, I found that the code for tsan within
gcc does not work consistently when ASLR is enabled on ppc64le. A
fix for that problem was integrated in gcc 9.
This adds a check to testsanitizers to determine the gcc compiler
version on ppc64le and skip the test if the version is too old.
A similar check is needed for asan too.
Updates #54645
Change-Id: I70717d1aa9e967cf1e871566e72b3862b91fea3f
Reviewed-on: https://go-review.googlesource.com/c/go/+/425355
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Archana Ravindar <aravind5@in.ibm.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
Also:
- fine-tune the implementation for some of the new builtin functions
- make sure the go/types code is an exact as possible copy of the
types2 code
- fix the description and examples for errorcodes.go
Follow-up on CL 423754.
For #53003.
Change-Id: I5c70b74e90c724cf6c842cedc6f8ace26fde372b
Reviewed-on: https://go-review.googlesource.com/c/go/+/425454
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
For "type T interface{ M() }", go/types users expect T's underlying
interface type to specify T as the receiver parameter type (#49906).
The unified importer handles this by cloning the interface to rewrite
the receiver parameters before calling SetUnderlying.
I missed in CL 425360 that these interfaces would need to have
Complete called too.
Manually tested to confirm that this actually fixes "go test -race
golang.org/x/tools/go/analysis/internal/checker" now (when both CLs
are ported to the x/tools importer).
Updates #54653.
Change-Id: I51e6db925db56947cd39dbe880230f14734ca01c
Reviewed-on: https://go-review.googlesource.com/c/go/+/425365
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
To support concurrent use of the go/types API, importers need to call
Interface.Complete on constructed interfaces before returning.
There's an issue that the interfaces may contain embedded defined
types, whose underlying type isn't known yet. This issue will
eventually go away once CL 424876 lands, but that CL needs to wait for
CL 424854 to re-land, which needs to wait for CL 421879 to land...
In the mean time, this CL implements the same solution used by the
indexed importer: maintaining a list of constructed interfaces, and
calling Interface.Complete on them after the SetUnderlying loop and
just before returning the imported package.
Updates #54653.
Change-Id: I0f42c915a4b7d28c628bbab7ac2eab2415c7858f
Reviewed-on: https://go-review.googlesource.com/c/go/+/425360
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
This is a follow up of CL 425101 on RISCV64.
According to RISCV Volume 1, Unprivileged Spec v. 20191213 Chapter 7.1:
If both the high and low bits of the same product are required, then the
recommended code sequence is: MULH[[S]U] rdh, rs1, rs2; MUL rdl, rs1, rs2
(source register specifiers must be in same order and rdh cannot be the
same as rs1 or rs2). Microarchitectures can then fuse these into a single
multiply operation instead of performing two separate multiplies.
So we should not split Muluhilo to separate instructions.
Updates #54607
Change-Id: If47461f3aaaf00e27cd583a9990e144fb8bcdb17
Reviewed-on: https://go-review.googlesource.com/c/go/+/425203
Auto-Submit: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Ever since 'go build' was added (in CL 5483069), it has used an atexit
handler to clean up working directories. At some point (prior to CL
95900044), Init was called multiple times per builder, registering
potentially many atexit handlers that execute asynchronously and make
debugging more difficult.
The use of an AtExit handler also makes the Builder (and anything that
uses it) prone to races: the base.AtExit API is not designed for
concurrent use, but cmd/go is becoming increasingly concurrent over
time. The AtExit handler also makes the Builder inappropriate to use
within a unit-test, since the handlers do not run during the test
function and accumulate over time.
This change makes NewBuilder safe for concurrent use by registering
the AtExit handler only once (during BuildInit, which was already not
safe for concurrent use), and using a sync.Map to store the set of
builders that need cleanup in case of an unclean exit. In addition, it
causes the test variant of cmd/go to fail if any Builder instance
leaks from a clean exit, helping to ensure that functions that create
Builders do not leak them indefinitely, especially in tests.
Updates #54423.
Change-Id: Ia227b15b8fa53c33177c71271d756ac0858feebe
Reviewed-on: https://go-review.googlesource.com/c/go/+/425254
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Ever since 'go build' was added (in CL 5483069), it has used an atexit
handler to clean up working directories.
CL 154109 introduced 'cc' command to the script test framework that
called Init on a builder once per invocation. Unfortunately, since
base.AtExit is unsynchronized, the Init added there caused any script
that invokes that command to be unsafe for concurrent use.
This change fixes the race by having the 'cc' command pass in its
working directory instead of allowing the Builder to allocate one.
Following modern Go best practices, it also replaces the in-place Init
method (which is prone to typestate and aliasing bugs) with a
NewBuilder constructor function.
Fixes#54423.
Change-Id: I8fc2127a7d877bb39a1174e398736bb51d03d4d2
Reviewed-on: https://go-review.googlesource.com/c/go/+/425205
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
This CL changes the inliner to process transitive inlining iteratively
after the AST has actually been edited, rather than recursively and
immediately. This is important for handling indirect function calls
correctly, because ir.reassigned walks the function body looking for
reassignments; whereas previously the inlined reassignments might not
have been actually added to the AST yet.
Fixes#54632.
Change-Id: I0dd69813c8a70b965174e0072335bc00afedf286
Reviewed-on: https://go-review.googlesource.com/c/go/+/425257
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: David Chase <drchase@google.com>
If a struct or array is comparable, then we can leverage rtype.equal,
which is almost always faster than what Go reflection can achieve.
As a secondary optimization, pre-compute Value.Len and Value.NumField
outside of the loop conditional.
Performance:
name old time/op new time/op delta
IsZero/ArrayComparable 136ns ± 4% 16ns ± 1% -88.28% (p=0.008 n=5+5)
IsZero/ArrayIncomparable 197ns ±10% 123ns ± 1% -37.74% (p=0.008 n=5+5)
IsZero/StructComparable 26.4ns ± 0% 9.6ns ± 1% -63.68% (p=0.016 n=4+5)
IsZero/StructIncomparable 43.5ns ± 1% 27.8ns ± 1% -36.21% (p=0.008 n=5+5)
The incomparable types gain a performance boost since
they are generally constructed from nested comparable types.
Change-Id: If2c1929f8bb1b5b19306ef0c69f3c95a27d4b60d
Reviewed-on: https://go-review.googlesource.com/c/go/+/411478
Reviewed-by: Dan Kortschak <dan@kortschak.io>
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This causes a problem in the test sometimes. With a mapping like:
00400000-00411000 r--p 00000000 fe:01 4459044 /tmp/go-build1710804385/b001/pprof.test
00411000-00645000 r-xp 00011000 fe:01 4459044 /tmp/go-build1710804385/b001/pprof.test
The removed code would make the first mapping 0x400000-0x645000. Tests
then grab the first few addresses to use as PCs, thinking they are in
an executable range. But those addresses are really not in an
executable range, causing the tests to fail.
Change-Id: I5a69d0259d1fd70ff9745df1cbad4d54c5898e7b
Reviewed-on: https://go-review.googlesource.com/c/go/+/424295
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
From the append docs in the builtin package:
As a special case, it is legal to append a string to a byte slice, like this:
slice = append([]byte("hello "), "world"...)
Change-Id: Ib14039a7476873b12a3aefccd8863e8d628b9249
Reviewed-on: https://go-review.googlesource.com/c/go/+/425102
Reviewed-by: hopehook <hopehook@qq.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
If narch is very large we would allocate a lot of memory for seenArches.
In practice we aren't going to see many different architectures so
don't bother to specify a size for the seenArches map.
No debug/macho test case because the problem can only happen for
invalid data. Let the fuzzer find cases like this.
For #47653
For #52523
Change-Id: I5a3b0e3aa6172ddffd6f44d9ae513c39a00d8764
Reviewed-on: https://go-review.googlesource.com/c/go/+/425114
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Joseph Tsai <joetsai@digital-static.net>
When one has a []byte on hand, but desires to call the Parse functions,
the conversion from []byte to string would allocate.
var b []byte = ...
v, err := strconv.ParseXXX(string(b), ...)
This changes it such that the input string never escapes from
any of the Parse functions. Together with the compiler optimization
where the compiler stack allocates any string smaller than 32B
this makes most valid inputs for strconv.ParseXXX(string(b), ...)
not require an allocation for the input string.
For example, the longest int64 or uint64 encoded in decimal is 20B.
Also, the longest decimal formatting of a float64 in appendix B
of RFC 8785 is 25B.
Previously, this was not possible since the input leaked to the error,
which causes the prover to give up and instead heap copy the []byte.
We fix this by copying the input string in the error case.
The advantage of this change is that you can now call strconv.ParseXXX
with a []byte without allocations (most times) in the non-error case.
The detriment is that the error-case now has an extra allocation.
We should optimize for the non-error path, rather than the error path.
The effects of this change is transitively seen through packages
that must use strconv.ParseXXX on a []byte such as "encoding/json":
name old time/op new time/op delta
UnmarshalFloat64 186ns 157ns -15.89% (p=0.000 n=10+10)
name old alloc/op new alloc/op delta
UnmarshalFloat64 148B 144B -2.70% (p=0.000 n=10+10)
name old allocs/op new allocs/op delta
UnmarshalFloat64 2.00 1.00 -50.00% (p=0.000 n=10+10)
In order for "encoding/json" to benefit, there needs to be a
small change made to how "encoding/json" calls strconv.ParseXXX.
That will be a future change.
Credit goes to Jeff Wendling for a similar patch.
Fixes#42429
Change-Id: I512d6927f965f82e95bd7ec14a28a587f23b7203
Reviewed-on: https://go-review.googlesource.com/c/go/+/345488
Reviewed-by: Martin Möhrmann <martin@golang.org>
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Auto-Submit: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Robert Griesemer <gri@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Some code paths in the runtime (cgo, heapdump) request heap bits
without first checking that the span is !noscan. Instead of trying
to find and work around all those cases, just set the pointer bits
of noscan spans correctly. It's somewhat safer than ensuring we
caught all the possible cases.
Fixes#54557Fixes#54558
Change-Id: Ibd476e6cdea77c962e4d15aad26f29df66fd94e8
Reviewed-on: https://go-review.googlesource.com/c/go/+/425194
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Normally, when moving Go values of type T from one location to another,
we don't need to worry about partial overlaps. The two Ts must either be
in disjoint (nonoverlapping) memory or in exactly the same location.
There are 2 cases where this isn't true:
1) Using unsafe you can arrange partial overlaps.
2) Since Go 1.17, you can use a cast from a slice to a ptr-to-array.
https://go.dev/ref/spec#Conversions_from_slice_to_array_pointer
This feature can be used to construct partial overlaps of array types.
var a [3]int
p := (*[2]int)(a[:])
q := (*[2]int)(a[1:])
*p = *q
We don't care about solving 1. Or at least, we haven't historically
and no one has complained.
For 2, we need to ensure that if there might be partial overlap,
then we can't use OpMove; we must use memmove instead.
(memmove handles partial overlap by copying in the correct
direction. OpMove does not.)
Note that we have to be careful here not to introduce a call when
we're marshaling arguments to a call or unmarshaling results from a call.
Fixes#54467
Change-Id: I1ca6aba8041576849c1d85f1fa33ae61b80a373d
Reviewed-on: https://go-review.googlesource.com/c/go/+/425076
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
None of cgo, "go test", nor srcimporter make use of go/ast's object
resolution via go/ast.Object. As such, we can skip that work during
parse time, which should save some CPU time.
We don't have any benchmark numbers, as none of the three packages have
any usable benchmarks, but we measured gofmt to be about 5% faster
thanks to this tweak in https://go.dev/cl/401454.
These three packages are quite different to gofmt, but one can expect
similar speed-ups in the 1-5% range.
Two notable exceptions, which do make use of go/ast.Object, are cmd/fix
and cmd/doc - we do not modify those here.
See #46485.
Change-Id: Ie3e65600d4790641c4e4d6f1c379be477fa02cee
Reviewed-on: https://go-review.googlesource.com/c/go/+/401455
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Flush can not check for unclosed elements, as more data might be encoded
after Flush is called. Close implicitly calls Flush and also checks that
all opened elements are closed as well.
Fixes#53346
Change-Id: I889b9f5ae54e5dfabb9e6948d96c5ed7bc1110f9
Reviewed-on: https://go-review.googlesource.com/c/go/+/424777
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: David Chase <drchase@google.com>
In CL 424734, I implemented pointer shaping for unified IR. Evidently
though, we didn't have any test cases that check that uses of
pointer-shaped expressions were handled correctly.
In the reported test case, the struct field "children items[*node[T]]"
gets shaped to "children items[go.shape.*uint8]" (underlying type
"[]go.shape.*uint8"); and so the expression "n.children[i]" has type
"go.shape.*uint8" and the ".items" field selection expression fails.
The fix implemented in this CL is that any expression of derived type
now gets an explicit "reshape" operation applied to it, to ensure it
has the appropriate type for its context. E.g., the "n.children[i]"
OINDEX expression above gets "reshaped" from "go.shape.*uint8" to
"*node[go.shape.int]", allowing the field selection to succeed.
This CL also adds a "-d=reshape" compiler debugging flag, because I
anticipate debugging reshaping operations will be something to come up
again in the future.
Fixes#54535.
Change-Id: Id847bd8f51300d2491d679505ee4d2e974ca972a
Reviewed-on: https://go-review.googlesource.com/c/go/+/424936
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: hopehook <hopehook@qq.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
During walk, we sometimes desugar OEQ nodes into multiple "untyped
bool" expressions, and then use typecheck.Conv to convert back to the
original OEQ node's type.
However, typecheck.Conv had a short-circuit path that if the type is
already identical to the target type according to types.Identical,
then we skipped the conversion. This short-circuit is normally fine;
but with generic code and shape types, it considers "untyped bool" and
"go.shape.bool" to be identical types. And we could end up leaving an
expression of "untyped bool", which then fails an internal consistency
check later.
The simple fix is to change Conv to use types.IdenticalStrict, so that
we ensure "untyped bool" gets converted to "go.shape.bool". And for
good measure, make the same change to ConvNop.
This issue was discovered and reported against unified IR, but the
issue was latent within the non-unified frontend too.
Fixes#54537.
Change-Id: I7559a346b063349b35749e8a2da704be18e51654
Reviewed-on: https://go-review.googlesource.com/c/go/+/424937
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: David Chase <drchase@google.com>
To disambiguate local types, we append a "·N" suffix to their name and
then trim it off again when producing their runtime type descriptors.
However, if a local type is generic, then we were further appending
the type arguments after this suffix, and the code in types/fmt.go
responsible for trimming didn't know to handle this.
We could extend the types/fmt.go code to look for the "·N" suffix
elsewhere in the type name, but this is risky because it could
legitimately (albeit unlikely) appear in struct field tags.
Instead, the most robust solution is to just change the mangling logic
to keep the "·N" suffix at the end, where types/fmt.go can easily and
reliably trim it.
Note: the "·N" suffix is still visible within the type arguments
list (e.g., the "·3" suffixes in nested.out), because we currently use
the link strings in the type arguments list.
Fixes#54456.
Change-Id: Ie9beaf7e5330982f539bff57b8d48868a3674a37
Reviewed-on: https://go-review.googlesource.com/c/go/+/424901
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Than McIntosh <thanm@google.com>
When handling a type declaration like:
```
type B A
```
unified IR has been writing out that B's underlying type is A, rather
than the underlying type of A.
This is a bit awkward to implement and adds complexity to importers,
who need to handle resolving the underlying type themselves. But it
was necessary to handle when A was declared like:
```
//go:notinheap
type A int
```
Because we expected A's not-in-heap'ness to be conferred to B, which
required knowing that A was on the path from B to its actual
underlying type int.
However, since #46731 was accepted, we no longer need to support this
case. Instead we can write out B's actual underlying type.
One stumbling point though is the existing code for exporting
interfaces doesn't work for the underlying type of `comparable`, which
is now needed to implement `type C comparable`. As a bit of a hack, we
we instead export its underlying type as `interface{ comparable }`.
Fixes#54512.
Change-Id: I0fb892068d656f1e87bb8ef97da27756051126d5
Reviewed-on: https://go-review.googlesource.com/c/go/+/424854
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Previously we convert $0 to the ZR register for some reasons, which causes
two problems:
1. Confusion, the special case of the ZR register needs to be considered
when dealing with constants. For encoding, some places we encode ZR, and
some places we encode $0, although we have converted $0 to ZR.
2. Unexpected instruction format. All instructions that support ZR register
operands can be replaced by $0.
This patch removes this conversion. Note that this patch may cause previously
unintendedly supported instruction formats to no longer be supported.
Change-Id: I3d8d2c06711b7614a38191397da7776417f1861c
Reviewed-on: https://go-review.googlesource.com/c/go/+/404316
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Eric Fang <eric.fang@arm.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
On ARM64 we use two separate instructions to compute the hi and lo
results of a 64x64->128 multiplication. Lower to two separate ops
so if only one result is needed we can deadcode the other.
Fixes#54607.
Change-Id: Ib023e77eb2b2b0bcf467b45471cb8a294bce6f90
Reviewed-on: https://go-review.googlesource.com/c/go/+/425101
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Use OpVarDef only when the variable being defined has pointers in it.
VarDef markers are only used for liveness analysis, and that only
runs on pointer-ful variables.
Fixes#53810
Change-Id: I09b0ef7ed31e72528916fe79325f80bbe69ff9b4
Reviewed-on: https://go-review.googlesource.com/c/go/+/419320
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Joedian Reid <joedian@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
os/exec.Cmd.Wait closes the read end of os/exec.Cmd.StdoutPipe, meaning
that io.ReadAll can return fs.ErrClosed if the child exits too early,
allowing Wait to complete. The StdoutPipe docs already note this sharp
edge.
Move cmd.Wait until after we finish waiting on stdout. If the child
crashes for some reason, the write end of the pipe will implicitly close
causing io.ReadAll to return as well, so we won't get stuck.
Fixes#52725.
Change-Id: Ifae5745d86206879af2f1523a664236972e07252
Reviewed-on: https://go-review.googlesource.com/c/go/+/420597
Run-TryBot: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
I was profiling the cpu usage of go/printer's only benchmark,
and found that token.File.Unpack was one of the top offenders.
It was mainly the deferred unlock that took a big chunk of time,
and to my surprise, reoving the use of defer helped significantly:
name old time/op new time/op delta
Print-16 5.61ms ± 2% 5.38ms ± 1% -4.04% (p=0.000 n=10+8)
name old speed new speed delta
Print-16 9.27MB/s ± 2% 9.64MB/s ± 1% +4.03% (p=0.000 n=9+8)
name old alloc/op new alloc/op delta
Print-16 332kB ± 0% 332kB ± 0% ~ (p=0.363 n=10+10)
name old allocs/op new allocs/op delta
Print-16 3.45k ± 0% 3.45k ± 0% ~ (all equal)
It seems like #38471 is to blame, as the defer prevents Unlock from
being inlined. Add a TODO as a reminder to come back here once the
compiler issue is fixed.
Change-Id: I5a1c6d36a8e8357435a305a1bc0970ee0358b08a
Reviewed-on: https://go-review.googlesource.com/c/go/+/424920
Reviewed-by: Alan Donovan <adonovan@google.com>
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
To avoid false positives from the reflectvaluecompare checker #43993
Use v.IsValid() instead of
var zero reflect.Value
v != zero
Also avoid comparing directly with the singleton reflect.Value
representing a missing value. Detect the missing value by type instead.
Change-Id: I3a00d63cf61c077e7c7ae816474aa1f032be325b
Reviewed-on: https://go-review.googlesource.com/c/go/+/308769
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Daniel Martí <mvdan@mvdan.cc>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Occasionally the signal will be sent to a Go thread, which will cause
the program to exit with SIGQUIT rather than SIGSEGV.
Add TestSignalForwardingGo to test the case where the signal is
expected to be delivered to a Go thread.
This is a roll forward of CL 419014 which was rolled back in CL 424954.
This CL differs from 419014 in that it skips TestSignalForwardingGo
on darwin-amd64.
Fixes#53907
Change-Id: I5df3fd610c068df3bd48d9b3d7a9379248b97999
Reviewed-on: https://go-review.googlesource.com/c/go/+/425002
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
This follows on CL 407176 which added this function (in both
packages). This CL makes it consistent with the Cut function,
which uses “before” and “after” in return variable names.
Change-Id: Id4345d2fe0f50bf301a880803e87bf356986b518
Reviewed-on: https://go-review.googlesource.com/c/go/+/424922
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Use "method has no receiver" and "method has multiple receivers"
in error messages for invalid receiver counts, matching the
corresponding types2 errors.
For #54511.
Change-Id: I96fc99440d6206c74e9416069db052234baa8248
Reviewed-on: https://go-review.googlesource.com/c/go/+/424934
Reviewed-by: Alan Donovan <adonovan@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Use "middle" and "final" rather than "2nd" and "3rd" in error messages
for invalid slice expressions. This is the original compiler error
message and many tests check for this specific message.
For #54511.
Change-Id: I86eb739aa7218b7f393fab1ab402732cb9e9a1f6
Reviewed-on: https://go-review.googlesource.com/c/go/+/424906
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Use "invalid" rather than "incomparable" in error message for map key
types that are not comparable. This is the original compiler error
message and many tests check for this specific message. The type
checker does provide an additional explanation if the reason for
the error is not obvious (e.g. for type parameters).
For #54511.
Change-Id: Idb76c48b4dfbfd66a7deac728a552e07f14e06d7
Reviewed-on: https://go-review.googlesource.com/c/go/+/424905
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Use same approach to parsing const and var declarations
as the syntax package. Specifically, don't complain if
the first const specification in a const declaration
doesn't have a type and initialization expression. This
removes some duplicate errors when combined with the
type checker.
Adjust corresponding type checker tests accordingly.
For #54511.
Change-Id: I96702eba51dda6b581dad44577a7f93e4c02c857
Reviewed-on: https://go-review.googlesource.com/c/go/+/424904
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
All code may be generic. Remove machinery to select parsing
mode (generic vs non-generic) since the parser doesn't support
this anymore. Adjust tests to more closely match corresponding
types2 code.
Change-Id: Id2398afe64d58714974ec96656fdf67c02ff5d04
Reviewed-on: https://go-review.googlesource.com/c/go/+/424900
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Generics are part of the language now; there's no need anymore
to switch back to a syntax without generics. Remove the associated
machinery and adjust short tests accordingly.
Change-Id: I6b16c5c75fd9354ee87e3b9bee110f49f514565a
Reviewed-on: https://go-review.googlesource.com/c/go/+/424857
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
The validity of an import path string is checked by the type checker
(and possibly other tools); it doesn't need to be done by the parser.
Remove the respective code and tests.
Also, adjust a corresponding go/types test which resolves a TODO.
For #54511.
Change-Id: Id1fc80df4e3e83be3ef123da3946ccb8f759779f
Reviewed-on: https://go-review.googlesource.com/c/go/+/424855
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Since the examples tests are now identical between the two type checkers,
remove the local copy of the examples tests and (for now) use the tests
in go/types/testdata/examples instead. Eventually we may decide to move
all tests out of the type checker directories and place them in a
shared space (e.g. internal/types/testdata).
For #54511.
Change-Id: Ibd8599d09e781b2219a23114b4b2049757971181
Reviewed-on: https://go-review.googlesource.com/c/go/+/424695
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
This was never permitted in Go but the flexibility to do so
was introduced through the generics prototype code where we
experimented with parentheses to enclose type parameters.
Restore original (pre-generics) behavior.
Fixes#51655.
Change-Id: Ia7a4b2e393e0214a70e840c8663cf4474c5c754b
Reviewed-on: https://go-review.googlesource.com/c/go/+/424694
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Offsetwriter refers to the design of SectionReader and removes
the section parameter n.
Since the size of the written data is determined by the user,
we cannot know where the end offset of the original data is.
The offset of SeekEnd is not valid in Seek method.
Fixes#45899.
Change-Id: I9d9445aecfa0dd4fc5168f2f65e1e3055c201b45
Reviewed-on: https://go-review.googlesource.com/c/go/+/406776
Run-TryBot: Ian Lance Taylor <iant@google.com>
Reviewed-by: Joedian Reid <joedian@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
AppendRune appends the UTF-16 encoding of a rune to a []uint16.
BenchmarkEncodeValidASCII-12 24.61ns 16B 1allocs
BenchmarkEncodeValidJapaneseChars-12 18.79ns 8B 1allocs
BenchmarkAppendRuneValidASCII-12 6.826ns 0B 0allocs
BenchmarkAppendRuneValidJapaneseChars-12 3.547ns 0B 0allocs
The ASCII case is written to be inlineable.
Fixes#51896
Change-Id: I593b1029f603297ef6e80e036f2fee2a0938d38d
Reviewed-on: https://go-review.googlesource.com/c/go/+/409054
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Joedian Reid <joedian@golang.org>
Though it increases the execution time, the function is already quite
fast for most users, the allocation is much more important.
name old time/op new time/op delta
ByteReplacerWriteString-8 1.23µs ± 0% 2.16µs ± 1% +75.31% (p=0.000 n=10+10)
name old alloc/op new alloc/op delta
ByteReplacerWriteString-8 2.69kB ± 0% 0.00kB -100.00% (p=0.000 n=10+10)
name old allocs/op new allocs/op delta
ByteReplacerWriteString-8 1.00 ± 0% 0.00 -100.00% (p=0.000 n=10+10)
Change-Id: I6a36df5fcb8e11ef27e6c7b252aa88e869592f3c
Reviewed-on: https://go-review.googlesource.com/c/go/+/424136
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Joedian Reid <joedian@golang.org>
Support for Linux kernel versions requiring the fallback to CloseOnExec
was dropped from recent Go versions. The minimum Linux kernel version is
2.6.32 as of Go 1.18. The SOCK_CLOEXEC flag for the socket syscall is
supported since kernel version 2.6.27.
Follows a similar change for net.sysSocket in CL 403634.
For #45964
Change-Id: I8b6311f07c4ed7900a9af3ecb2e146c49db08665
Reviewed-on: https://go-review.googlesource.com/c/go/+/422374
Reviewed-by: Joedian Reid <joedian@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
The constants for these were auto-generated from the C includes
into zerrors_linux* files quite some time ago. The generator is
currently broken, but some new flags need to be added nevertheless.
As the flags won't change and the values are the same for all
architectures, we can just define them statically (as it's already
done in the runtime package):
- remove the CLONE_* constants from zerrors_linux_*.go;
- patch mkerrors.sh to not generate CLONE_ constants
(in case it will be fixed and used in the future);
- add the constants and some comments about them to exec_linux.go,
using Linux v5.17 include/uapi/sched.h as the ultimate source.
This adds the following new flags:
- CLONE_CLEAR_SIGHAND
- CLONE_INTO_CGROUP
- CLONE_NEWCGROUP
- CLONE_NEWTIME
- CLONE_PIDFD
For #51246.
Change-Id: I0c635723926218bd403d37e113ee4d62194463a8
Reviewed-on: https://go-review.googlesource.com/c/go/+/407574
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Joedian Reid <joedian@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
In order to prevent false sharing of cache lines, structs are
padded with some number of bytes. These bytes are unused, serving
only to make the size of the struct a multiple of the size of the
cache line.
The current calculation of how much to pad is an overestimation,
when the struct size is already a multiple of the cache line size
without padding. For these cases, no padding is necessary, and
the size of the inner pad field should be 0. The bug is that the
pad field is sized to a whole 'nother cache line, wasting space.
Here is the current formula that can never return 0:
cpu.CacheLinePadSize - unsafe.Sizeof(myStruct{})%cpu.CacheLinePadSize
This change simply mods that calculation by cpu.CacheLinePadSize,
so that 0 will be returned instead of cpu.CacheLinePadSize.
Change-Id: I26a2b287171bf47a3b9121873b2722f728381b5e
Reviewed-on: https://go-review.googlesource.com/c/go/+/414214
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Joedian Reid <joedian@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Occasionally the signal will be sent to a Go thread, which will cause
the program to exit with SIGQUIT rather than SIGSEGV.
Add TestSignalForwardingGo to test the case where the signal is
expected to be delivered to a Go thread.
Fixes#53907
Change-Id: Iaefb964c2be4a815c11c507fa89648f8a7740ba9
Reviewed-on: https://go-review.googlesource.com/c/go/+/419014
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Call only initHPETTimecounter on the system stack.
Use O_CLOEXEC flag when opening the HPET device.
FreeBSD 12.3-RELEASE-p2, AMD FX-8300
paulzhol@relic:~/go/src/time % ~/gocode/bin/benchcmp old_hpet.txt new_hpet.txt
benchcmp is deprecated in favor of benchstat: https://pkg.go.dev/golang.org/x/perf/cmd/benchstat
benchmark old ns/op new ns/op delta
BenchmarkNow-8 1420 1397 -1.62%
BenchmarkNowUnixNano-8 1421 1404 -1.20%
BenchmarkNowUnixMilli-8 1423 1405 -1.26%
BenchmarkNowUnixMicro-8 1423 1404 -1.34%
Update #50947
Change-Id: I553b5427fb0b86d7e070af4516b36326bc0aaf00
Reviewed-on: https://go-review.googlesource.com/c/go/+/391856
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Since the spec tests are now identical between the two type checkers,
remove the local copy of the spec tests and (for now) use the tests
in go/types/testdata/spec instead. Eventually we may decide to move
all tests out of the type checker directories and place them in a
shared space (e.g. internal/types/testdata).
For #54511.
Change-Id: Ibc62230553a5ff215160dcc97696b333ae9cf685
Reviewed-on: https://go-review.googlesource.com/c/go/+/424676
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Run-TryBot: Robert Griesemer <gri@google.com>
Besides applying gofmt (which doesn't damage the tests in this case),
this removes a TODO for a set of nil-related tests.
The test files are now identical.
For #54511.
Change-Id: I0a286ad607f317f43972c1f5ee741b4f5bc9576d
Reviewed-on: https://go-review.googlesource.com/c/go/+/424675
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Besides applying gofmt (which doesn't damage the tests in this case),
the ERROR comments in the types2 files now match the go/types files.
But because types2 still reports some errors are different positions,
the checking code now allows for some position discrepancy (similar
to what we do for other tests).
Except for an outstanding TODO in go/types/testdata/examples/types.go
the test files are now identical.
For #54511.
Change-Id: I5748e0f678d11c5c0bdf4fdf28bd04f0b11b5b23
Reviewed-on: https://go-review.googlesource.com/c/go/+/424674
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Instead of simply reporting an error but otherwise dropping the
index expression from the parse tree when an index is missing
(as in: x[]), create an index expression with a "bad expression"
as index. This matches the behavior of go/parser and permits the
use of the same test case for both parsers.
(It would be simpler to adjust the go/parser to match the syntax
parser's behavior, but that would break backward-compatibility
of the go/parser.)
Adjust the affected test files.
For #54511.
Change-Id: If7668973794604593e869a24b560da92e100b812
Reviewed-on: https://go-review.googlesource.com/c/go/+/424654
Run-TryBot: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Update the version of BoringCrypto to boringssl tag
fips-20210429, for which FIPS approval is "in process".
Add GOARCH=arm64 BoringCrypto support.
Shuffle build to run as distinct steps in Docker so that
we can rerun later parts like build-goboring.sh without
rerunning all the setup.
Strip unnecessary parts of the syso using --strip-unneeded,
which cuts the amd64 syso from 10MB to 2.5MB.
The new arm64 syso is 2MB, so even though we added a new
architecture, we have half as much syso to distribute.
Change-Id: I0f7327389a3a485b82577abea9153d006418298f
Reviewed-on: https://go-review.googlesource.com/c/go/+/423362
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
With the introduction of stack objects, VARKILL information is
no longer needed.
With stack objects, an object is dead when there are no more static
references to it, and the stack scanner can't find any live pointers
to it. VARKILL information isn't used to establish live ranges for
address-taken variables any more. In effect, the last static reference
*is* the VARKILL, and there's an additional dynamic liveness check
during stack scanning.
Next CL will actually rip out the VARKILL opcodes.
Change-Id: I030a2ab867445cf4e0e69397911f8a2e2f0ed07b
Reviewed-on: https://go-review.googlesource.com/c/go/+/419234
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
We don't need this special loop construct anymore now that we do
conservative GC scanning of the top of stack. Rewrite instead to a simple
pointer increment on every iteration. This leads to having a potential
past-the-end pointer at the end of the last iteration, but that value
immediately goes dead after the loop condition fails, and the past-the-end
pointer is never live across any call.
This simplifies and speeds up loops.
R=go1.20
TODO: actually delete all support for OFORUNTIL. It is now never generated,
but code to handle it (e.g. in ssagen) is still around.
TODO: in "for _, x := range" loops, we could get rid of the index
altogether and use a "pointer to the last element" reference to determine
when the loop is complete.
Fixes#53409
Change-Id: Ifc141600ff898a8bc6a75f793e575f8862679ba1
Reviewed-on: https://go-review.googlesource.com/c/go/+/414876
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The non-unified frontend had repeated issues with inlining and
generics (#49309, #51909, #52907), which led us to substantially
restrict inlining when shape types were present.
However, these issues are evidently not present in unified IR's
inliner, and the safety restrictions added for the non-unified
frontend can simply be disabled in unified mode.
Fixes#54497.
Change-Id: I8e6ac9f3393c588bfaf14c6452891b9640a9d1bd
Reviewed-on: https://go-review.googlesource.com/c/go/+/424775
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
As a consistency check in devirtualization, when we determine `i` (of
interface type `I`) always has dynamic type `T`, we insert a type
assertion `i.(T)`. This emits an itab check for `go:itab.T,I`, but
it's always true (and so SSA optimizes it away).
However, if `I` is instead the generic interface type `I[T]`, then
`go:itab.T,I[int]` and `go:itab.T,I[go.shape.int]` are equivalent but
distinct itabs. And notably, we'll have originally created the
interface value using the former; but the (non-dynamic) TypeAssertExpr
created by devirtualization would ultimately emit a comparison against
the latter. This comparison would then evaluate false, leading to a
spurious type assertion panic at runtime.
The comparison is just meant as an extra safety check, so it should be
safe to just disable. But for now, it's simpler/safer to just punt on
devirtualization in this case. (The non-unified frontend doesn't
devirtualize this either.)
Change-Id: I6a8809bcfebc9571f32e289fa4bc6a8b0d21ca46
Reviewed-on: https://go-review.googlesource.com/c/go/+/424774
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
For local variables of derived type, Delve relies on ir.Name.DictIndex
being set to the type's rtype index within the function's dictionary.
This CL implements that functionality within unified IR.
Manually double checked that Delve behaves correctly, at least as far
as I can tell from casual use. Specifically, I confirmed that running
the test program from TestDictIndex, stepping into testfn, and then
running `print mapvar` prints `map[int]main.CustomInt []`, which
matches the behavior under GOEXPERIMENT=nounified. (Also compare that
when ir.Name.DictIndex is *not* set by unified IR, `print mapvar`
instead prints `map[int]go.shape.int []`.)
Fixes#54514.
Change-Id: I90d443945895abfba04dc018f15e00217930091c
Reviewed-on: https://go-review.googlesource.com/c/go/+/424735
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This CL implements pointer shaping in unified IR, corresponding to the
existing pointer shaping implemented in the non-unified frontend.
For example, if `func F[T any]` is instantiated as both `F[*int]` and
`F[*string]`, we'll now generate a single `F[go.shape.*uint8]` shaped
function that can be used by both.
Fixes#54513.
Change-Id: I2cef5ae411919e6dc5bcb3cac912abecb4cd5218
Reviewed-on: https://go-review.googlesource.com/c/go/+/424734
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Even though the -race option works for ppc64le, some of the
testsanitizer test have failed in the past on our builders. These
same failures can't be reproduced on other systems.
This is an experiment to re-enable this test on ppc64le to see if
it still fails on the builders.
Updates #45040
Change-Id: I0729bec5864e6c4cc752968485e89179df027063
Reviewed-on: https://go-review.googlesource.com/c/go/+/424534
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Archana Ravindar <aravind5@in.ibm.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
printer.print is an overloaded method for multiple purposes.
When fed a position, it updates the current position.
When fed a string, it prints the string.
When fed a token, it prints the token. And so on.
However, this overloading comes at a significant cost.
Because the parameters are a list of the `any` interface type,
any type which is not of pointer or interface kind will allocate when
passed as an argument, as interfaces can only contain pointers.
A large portion of the arguments passed to the print method are of type
token.Pos, whose underlying type is int - so it allocates.
Removing those allocations has a significant benefit,
at the cost of some verbosity in the code:
name old time/op new time/op delta
Print-16 6.10ms ± 2% 5.39ms ± 2% -11.72% (p=0.000 n=10+10)
name old speed new speed delta
Print-16 8.50MB/s ± 2% 9.63MB/s ± 2% +13.28% (p=0.000 n=10+10)
name old alloc/op new alloc/op delta
Print-16 443kB ± 0% 332kB ± 0% -25.10% (p=0.000 n=10+9)
name old allocs/op new allocs/op delta
Print-16 17.3k ± 0% 3.5k ± 0% -80.10% (p=0.000 n=10+10)
There should be more significant speed-ups left, particularly for the
token.Token, string, and whiteSpace types fed to the same method.
They are left for a future CL, in case this kind of optimization is not
a path we want to take.
Change-Id: I3ff8387242c5a935bb003e60e0813b7b9c65402e
Reviewed-on: https://go-review.googlesource.com/c/go/+/412557
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
Reviewed-by: hopehook <hopehook@qq.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The golang.org/x/sys/unix package is already imported for Utsname and
Uname. Use ByteSliceToString from that package as well to replace the
locally defined utsString helper which serves the same purpose and
matches ByteSliceToString's implementation.
Change-Id: I5d9de186a5aeb1feed1387beedefbcd260fe22ff
Reviewed-on: https://go-review.googlesource.com/c/go/+/415654
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
The documentation for LoadLocation contains an enumerated list,
but does not render as such because it's missing leading spaces.
Output verified with the go doc command and godoc server.
Change-Id: I88b61d34048b7d01ee5cd77c32849af266e2f4c5
Reviewed-on: https://go-review.googlesource.com/c/go/+/423297
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Support for operating system versions requiring the fallback to
CloseOnExec/SetNonblock was dropped from recent Go versions. The minimum
Linux kernel version is 2.6.32 as of Go 1.18. FreeBSD 10 is no longer
supported as of Go 1.13.
Follows a similar change for net.sysSocket in CL 403634 and
syscall.Socket in CL 422374.
For #45964
Change-Id: I60848415742a1d8204e1fda585462ff35ad6722f
Reviewed-on: https://go-review.googlesource.com/c/go/+/422375
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Most newer architectures (e.g. arm64, riscv64, loong64) don't provide the
epoll_create syscall. Some systems (e.g. Android) block it even if it
were available. In the kernel, the epoll_create syscall is implemented
[1] the same way EpollCreate is implemented in this package for
platforms without the epoll_create syscall. The epoll_create1 syscall is
available since Linux kernel 2.6.27 and the minimum required kernel
version is 2.6.32 since Go 1.18 (see #45964). Thus, avoid the separate
wrapper and consistently implement EpollCreate using EpollCreate1.
[1] https://elixir.bootlin.com/linux/v5.15-rc1/source/fs/eventpoll.c#L2006
The same change was already done in CL 349809 for golang.org/x/sys/unix.
For #45964
Change-Id: I5463b208aa7ae236fa2c175d6d3ec6568f1840b0
Reviewed-on: https://go-review.googlesource.com/c/go/+/411594
Reviewed-by: Bryan Mills <bcmills@google.com>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
The previous implementation of isCaseSensitive called t.Fatalf in the
wrong place, causing tests after the first to proceed past an error
determining case-sensitivity. That could lead to confusing errors.
(Moreover, I would like to try to disentangle the script engine from
testing.T so that I can also use it to generate serving contents in
the replacement for vcs-test.golang.org.)
The implementation of goVersion called ts.fatalf, which is probably
fine but prevents the script environment from being computed outside
of a test, as we might want to do for debugging and other scripting.
For #27494.
Change-Id: Ibfee0704523fdcd6174b544ff84267216435025b
Reviewed-on: https://go-review.googlesource.com/c/go/+/419874
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
When the GOOS or GOARCH of the cmd/go test binary does not match the
GOOS or GOARCH of the installed 'go' binary itself, the test currently
attempts to trick 'go test' into thinking that there were no test
functions to run.
That makes it very difficult to discover how to actually run the
tests, which in turn makes it difficult to diagnose and fix
regressions in, say, the linux-386-longtest builders. (We have had a
few of those lately, and they shouldn't be as much of an ordeal to fix
as they currently are.)
There are three underlying problems:
1. cmd/go uses its own GOOS and GOARCH to figure out which variant of
other tools to use, and the cache keys for all installed tools and
libraries include the IDs of the tools used to build them. So when
cmd/go's GOARCH changes, all installed tools and binaries appear
stale *even if* they were just installed by invoking the native
cmd/go with the appropriate GOARCH value set.
2. The "go/build" library used by cmd/go toggles its default
CGO_ENABLED behavior depending on whether the GOOS and GOARCH being
imported match runtime.GOOS and runtime.GOARCH.
3. A handful of cmd/go tests explicitly use gccgo, but the user's
installed gccgo binary cannot necessarily cross-compile to the same
platforms as cmd/go.
To address the cache-invalidation problem, we modify the test variant
of cmd/go to use the host's native toolchain (as indicated by the new
TESTGO_GOHOSTOS and TESTGO_GOHOSTARCH environment variables) instead
of the toolchain matching the test binary itself. That allows a test
cmd/go binary compiled with GOARCH=386 to use libraries and tools
cross-compiled by the native toolchain, so that
$ GOARCH=386 go install std cmd
suffices to make the packages in std and cmd non-stale in the
tests.
To address the CGO_ENABLED mismatch, we set CGO_ENABLED explicitly in
the test's environment whenever it may differ from the default. Since
script tests that use cgo are already expected to use a [cgo]
condition, setting the environment to match that condition fixes the
cgo-specific tests.
To address the gccgo-specific cross-compilation failures, we add a new
script condition, [cross], which evaluates to true whenever the
platform of the test binary differs from that of the native toolchain.
We can then use that condition to explicitly skip the handful of gccgo
tests that fail under cross-compilation.
Fixes#53936.
Change-Id: I8633944f674eb5941ccc95df928991660e7e8137
Reviewed-on: https://go-review.googlesource.com/c/go/+/356611
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
This CL switches unified IR to use shape-based stenciling with runtime
dictionaries, like the existing non-unified frontend. Specifically,
when instantiating generic functions and types `X[T]`, we now also
instantiated shaped variants `X[shapify(T)]` that can be shared by
`T`'s with common underlying types.
For example, for generic function `F`, `F[int](args...)` will be
rewritten to `F[go.shape.int](&.dict.F[int], args...)`.
For generic type `T` with method `M` and value `t` of type `T[int]`,
`t.M(args...)` will be rewritten to `T[go.shape.int].M(t,
&.dict.T[int], args...)`.
Two notable distinctions from the non-unified frontend:
1. For simplicity, currently shaping is limited to simply converting
type arguments to their underlying type. Subsequent CLs will implement
more aggressive shaping.
2. For generic types, a single dictionary is generated to be shared by
all methods, rather than separate dictionaries for each method. I
originally went with this design because I have an idea of changing
interface calls to pass the itab pointer via the closure
register (which should have zero overhead), and then the interface
wrappers for generic methods could use the *runtime.itab to find the
runtime dictionary that corresponds to the dynamic type. This would
allow emitting fewer method wrappers.
However, this choice does have the consequence that currently even if
a method is unused and its code is pruned by the linker, it may have
produced runtime dictionary entries that need to be kept alive anyway.
I'm open to changing this to generate per-method dictionaries, though
this would require changing the unified IR export data format; so it
would be best to make this decision before Go 1.20.
The other option is making the linker smarter about pruning unneeded
dictionary entries, like how it already prunes itab entries. For
example, the runtime dictionary for `T[int]` could have a `R_DICTTYPE`
meta-relocation against symbol `.dicttype.T[go.shape.int]` that
declares it's a dictionary associated with that type; and then each
method on `T[go.shape.T]` could have `R_DICTUSE` meta-relocations
against `.dicttype.T[go.shape.T]+offset` indicating which fields
within dictionaries of that type need to be preserved.
Change-Id: I369580b1d93d19640a4b5ecada4f6231adcce3fd
Reviewed-on: https://go-review.googlesource.com/c/go/+/421821
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The splice syscall is buggy prior to Linux 2.6.29. CL 113999 added a
workaround to detect buggy versions and disable use of splice for these.
As of Go 1.18 the minumum Linux version is 2.6.32. Thus, a non-buggy
implementation of the splice syscall can be assumed.
For #45964Fixes#54505
Change-Id: Ied3a3334da7a3f7fa1280b7c5b1dfb9030219336
Reviewed-on: https://go-review.googlesource.com/c/go/+/422979
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Either ones where kind == kindNonBlock or those we've successfully called syscall.SetNonblock() on.
Restore blocking behavior if we detect an error registering with the netpoller and our flow was
successful in setting the inital syscall.SetNonblock().
Update #54100
Change-Id: I08934e4107c7fb36c15a7ca23ac880490b4df235
Reviewed-on: https://go-review.googlesource.com/c/go/+/420334
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Goutnik <dgoutnik@gmail.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Yuval Pavel Zholkover <paulzhol@gmail.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Auto-Submit: Ian Lance Taylor <iant@golang.org>
Plan 9 a.out was not implemented for debug/buildinfo, which
was causing test failures on Plan 9. This adds an implementation,
and causes the tests to pass.
Fixes#53949
Change-Id: I90a307ef9babf8cf381f8746d731cac2206b234a
Reviewed-on: https://go-review.googlesource.com/c/go/+/418014
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Benny Siegert <bsiegert@gmail.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Generics lets us write Cache[K, V] instead of using unsafe.Pointer,
which lets us remove all the uses of package unsafe around the
uses of the cache.
I tried to do Cache[*K, *V] instead of Cache[K, V] but that was not possible.
Change-Id: If3b54cf4c8d2a44879a5f343fd91ecff096537e9
Reviewed-on: https://go-review.googlesource.com/c/go/+/423357
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Russ Cox <rsc@golang.org>
'go tool' sets environment variables, including the GOAMD64 value
from the user's go.env file.
'go tool dist test' then rebuilds and reinstalls the toolchain and
standard library based on those variables. It should not; instead, it
should test exactly the configuration installed by the make scripts.
Fixes#54084.
Change-Id: I7cc8a21cc1d8331e06d7b7c55b14d170f8e2faab
Reviewed-on: https://go-review.googlesource.com/c/go/+/420055
Run-TryBot: Bryan Mills <bcmills@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
We have a very complex process to make VDSO calls on ARM. Create a
wrapper helper function which reduces duplication and allows for
additional calls from other packages.
vdsoCall has a few differences from the original code in
walltime/nanotime:
* It does not use R0-R3, as they are passed through as arguments to fn.
* It does not save g if g.m.gsignal.stack.lo is zero. This may occur if
it called at startup on g0 between assigning g0.m.gsignal and setting
its stack.
For #49182
Change-Id: I51aca514b4835b71142011341d2f09125334d30f
Reviewed-on: https://go-review.googlesource.com/c/go/+/362795
Run-TryBot: Michael Pratt <mpratt@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Export cipher.xorBytes as subtle.XORBytes, for proposal #53021,
to provide fast XOR to cryptography libraries outside crypto/cipher.
Along with the move, implement the alignment check TODO
in xor_generic.go, so that systems with neither unaligned
accesses nor custom assembly can still XOR a word at a time
in word-based algorithms like GCM. This removes the need
for the separate cipher.xorWords.
Fixes#53021.
Change-Id: I58f80a922f1cff671b5ebc6168eb046e702b5a4c
Reviewed-on: https://go-review.googlesource.com/c/go/+/421435
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Alan Donovan <adonovan@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Request.Cookie(name string) will return the first cookie
when cookie name is "". Since readCookies in
file net/http/cookie.go at line 247 return all cookies
when second parameter is a empty string.
To fix it, Return ErrNoCookie from Request.Cookie(""),
instead of the first cookie in the request.
Fixes#53181
Change-Id: Ie623ca4c53da64ef7623a7863292a2d771f76832
GitHub-Last-Rev: 01098cd5db
GitHub-Pull-Request: golang/go#53183
Reviewed-on: https://go-review.googlesource.com/c/go/+/409754
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Damien Neil <dneil@google.com>
GoTool was added in CL 20967, and revised in CL 21292, for #14901.
I don't fully understand what problem the GoTool function was added to
solve: the discussion on that issue was pretty sparse, but it seems
like when we run tests of GOROOT packages they always know their own
location relative to GOROOT (and thus always know where to find the
'go' tool).
Lacking that understanding, I don't want to change its behavior, but I
do at least want to verify that it resolves to the real 'go' tool in
the common case (running 'go test' on a package in GOROOT/src).
For #50892
For #50893
Updates #14901
Change-Id: I06d831e6765be631dfc4854d7fddc3d27fc1de34
Reviewed-on: https://go-review.googlesource.com/c/go/+/381834
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Optimize Time.GoString by avoiding multiple calls to absDate.
name old time/op new time/op delta
GoString-8 313ns ± 2% 197ns ± 1% -37.08% (p=0.008 n=5+5)
name old alloc/op new alloc/op delta
GoString-8 80.0B ± 0% 80.0B ± 0% ~ (all equal)
name old allocs/op new allocs/op delta
GoString-8 1.00 ± 0% 1.00 ± 0% ~ (all equal)
Fixes#54436
Change-Id: I8e6f8e7bbb9857b4bc0cdf6ed29a6b2415775db7
Reviewed-on: https://go-review.googlesource.com/c/go/+/423634
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Joseph Tsai <joetsai@digital-static.net>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
The existing implementation has an execution time higher in the benchmark than this one.
This is an optimized implementation using the copy() function and a constant 256 bytes string with the values to be copied.
```
name old time/op new time/op delta
NewEncoding-4 329ns ± 1% 231ns ± 1% -29.72% (p=0.008 n=5+5)
name old speed new speed delta
NewEncoding-4 778MB/s ± 1% 1108MB/s ± 1% +42.29% (p=0.008 n=5+5)
```
Fixes#53211
Change-Id: I80fe62aa40623125ef81ae9164a8405eed30b71b
GitHub-Last-Rev: 55dce6f636
GitHub-Pull-Request: golang/go#53212
Reviewed-on: https://go-review.googlesource.com/c/go/+/410194
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Auto-Submit: Keith Randall <khr@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Don't allocate slices that are too large; choose a smaller capacity
and build the slice using append. Use this in debug/macho to avoid
over-allocating if a fat header is incorrect.
No debug/macho test case because the problem can only happen for
invalid data. Let the fuzzer find cases like this.
For #47653Fixes#52523
Change-Id: I372c9cdbdda8626a3225e79d713650beb350ebc7
Reviewed-on: https://go-review.googlesource.com/c/go/+/413874
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Tobias Klauser <tobias.klauser@gmail.com>
[This is a retry of CL 407036 + its revert CL 422394. The only
content change is the 1-line change in cmd/internal/obj/objfile.go.]
Read the bitmaps one uintptr at a time instead of one byte at a time.
Performance so far:
Allocation heavy, no retention: ~30% faster in heapBitsSetType
Scan heavy, ~no allocation: ~even in scanobject
Change-Id: I04d899e1dbd23e989e9f552cdc1880318779c14c
Reviewed-on: https://go-review.googlesource.com/c/go/+/422635
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
[this is a retry of CL 407035 + its revert CL 422395. The content is unchanged]
Use just 1 bit per word to record the ptr/nonptr bitmap.
Use word-sized operations to manipulate the bitmap, so we can operate
on up to 64 ptr/nonptr bits at a time.
Use a separate bitmap, one bit per word of the ptr/nonptr bitmap,
to encode a no-more-pointers signal. Since we can check 64 ptr/nonptr
bits at once, knowing the exact last pointer location is not necessary.
As a followon CL, we should make the gcdata bitmap an array of
uintptr instead of an array of byte, so we can load 64 bits of it at once.
Similarly for the processing of gc programs.
Change-Id: Ica5eb622f5b87e647be64f471d67b02732ef8be6
Reviewed-on: https://go-review.googlesource.com/c/go/+/422634
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Add a new Rewrite hook to ReverseProxy, superseding the Director hook.
Director does not distinguish between the inbound and outbound request,
which makes it possible for headers added by Director to be inadvertently
removed before forwarding if they are listed in the inbound request's
Connection header. Rewrite accepts a value containing the inbound
and outbound requests, with hop-by-hop headers already removed from
the outbound request, avoiding this problem.
ReverseProxy's appends the client IP to the inbound X-Forwarded-For
header by default. Users must manually delete untrusted X-Forwarded-For
values. When used with a Rewrite hook, ReverseProxy now strips
X-Forwarded-* headers by default.
NewSingleHostReverseProxy creates a proxy that does not rewrite the
Host header of inbound requests. Changing this behavior is
cumbersome, as it requires wrapping the Director function created
by NewSingleHostReverseProxy. The Rewrite hook's ProxyRequest
parameter provides a SetURL method that provides equivalent
functionality to NewSingleHostReverseProxy, rewrites the Host
header by default, and can be more easily extended with additional
customizations.
Fixes#28168.
Fixes#50580.
Fixes#53002.
Change-Id: Ib84e2fdd1d52c610e3887af66f517d4a74e594d0
Reviewed-on: https://go-review.googlesource.com/c/go/+/407214
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
The design of FileSet encourages it to be used as a global variable.
Each call to AddFile consumes about 3KB, that is never returned,
even after an application no longer cares about the File.
This change adds a RemoveFile method that a long-running application
can use to release a File that is no longer needed, saving memory.
Fixesgolang/go#53200
Change-Id: Ifd34d650fe0d18b1395f922a4cd02a535afbe560
Reviewed-on: https://go-review.googlesource.com/c/go/+/410114
Auto-Submit: Alan Donovan <adonovan@google.com>
Run-TryBot: Alan Donovan <adonovan@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
This test was apparently mistakenly removed without a replacement in
CL 213223, but its testdata was left in the tree. This change removes
the orphaned testdata subdirectory, and restores the test that
previously used that data as a self-contained script.
For #27494.
Change-Id: Ice81895a44c558aaab198b8ef7ec046d92f5d58f
Reviewed-on: https://go-review.googlesource.com/c/go/+/417658
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
As far as I can determine, this is the only existing non-script test
that relies on vcs-test.golang.org. I am attempting to eliminate the
vcs-test server, and to keep configuration straightforward I would
like to only set up its replacement for script tests.
For #27494.
Change-Id: Ib978228b9c3f403af5f1c445e08942cbe7915b62
Reviewed-on: https://go-review.googlesource.com/c/go/+/418103
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Bryan Mills <bcmills@google.com>
Run-TryBot: Bryan Mills <bcmills@google.com>
$ cd $GOROOT/src/testing
$ go test
root
root
root
root
PASS
$
The root prints have been happening since Go 1.14.
There is a test in sub_test.go that calls b.Run directly
with a benchmark named "root", which triggers the print.
Silence them.
Change-Id: I2f0c186f04c6139bc24fab0e91975fcf0a8e80fc
Reviewed-on: https://go-review.googlesource.com/c/go/+/421437
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
WithContext makes a shallow copy of a Request, and Clone makes a
deep copy. Both set the context of the new request. The distinction
between the two is clear, and it doesn't seem useful or necessary
to say that "it's rare to need WithContext".
Also update a couple locations that mention WithContext to mention
Clone as well.
Fixes#53413.
Change-Id: I89e6ddebd7d5ca6573e522fe48cd7f50cc645cdd
Reviewed-on: https://go-review.googlesource.com/c/go/+/412778
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
When accessing (*Checker).implements from types.AssignableTo or
types.ComparableTo, we don't need to build error strings -- they won't
be used.
This string manipulation showed up as a hot spot in gopls completion,
which checks a lot of type predicates when searching for candidate
completions.
This CL yields the following results for gopls' completion benchmarks:
StructCompletion-8 24.7ms ±34% 26.0ms ±17% ~ (p=0.447 n=10+9)
ImportCompletion-8 1.41ms ± 2% 1.45ms ± 4% +2.42% (p=0.027 n=8+9)
SliceCompletion-8 27.0ms ±18% 25.2ms ± 3% -6.67% (p=0.008 n=9+8)
FuncDeepCompletion-8 57.6ms ± 4% 22.4ms ± 4% -61.18% (p=0.000 n=8+9)
CompletionFollowingEdit-8 157ms ±13% 103ms ±15% -34.70% (p=0.000 n=10+10)
Notably, deep completion (which searches many candidates) is almost 3x
faster after this change.
Fixes#54172
Change-Id: If8303a411aed3a20bd91f7b61e346d703084166c
Reviewed-on: https://go-review.googlesource.com/c/go/+/423360
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
The value of Expires is only checked when a value is set.
This fixes the bug that a cookie with a zero-valued Expire
was considered invalid, even though Expires is an optional
field.
Fixes#52989
Change-Id: I206c50e9b6ea2744a92c74673d589ce2aaa62670
Reviewed-on: https://go-review.googlesource.com/c/go/+/407654
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
CL 414836 limited the check for implicit dot for method call enabled by
a type bound. However, the checking condition for ODOTMETH only is not
right. For example, for promoted method, we have a OXDOT node instead,
and we still have to check for implicit dot in this case.
However, if the base type and embedded types have the same method name,
e.g in issue #53419, typecheck.AddImplicitDots will be confused and
result in an ambigus selector.
To fix this, we ensure methods for the base type are computed, then only
do the implicit dot check if we can find a matched method.
Fixes#54348
Change-Id: Iefe84ff330830afe35c5daffd499824db108da23
Reviewed-on: https://go-review.googlesource.com/c/go/+/422274
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
This CL make math.sqrt an intrinsic function, math.Sqrt is not affected
since compiler can inline it. With this change, we can remove all assembly
code for math.Sqrt that aims to speed up indirect call. The go compiler can
generate same or faster code (with regabi) for indirect call.
Benchmark on amd64:
name old time/op new time/op delta
SqrtIndirect 2.60ns ± 3% 1.03ns ± 4% -60.24% (p=0.000 n=10+10)
SqrtLatency 3.40ns ± 1% 3.32ns ± 1% -2.26% (p=0.000 n=10+8)
SqrtIndirectLatency 6.09ns ± 0% 3.31ns ± 0% -45.67% (p=0.000 n=10+10)
SqrtGoLatency 36.1ns ± 6% 34.6ns ± 1% ~ (p=0.101 n=10+10)
SqrtPrime 2.53µs ± 2% 2.55µs ± 6% ~ (p=0.398 n=9+9)
Change-Id: If4be0f242c1d9d4feca7d269fc9cd6e6066f163d
Reviewed-on: https://go-review.googlesource.com/c/go/+/421074
Run-TryBot: Wayne Zuo <wdvxdr@golangcn.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Single quotes don't account for whitespace in this context, which causes
output to look like this:
$ ./make.bat
'C:\Program' is not recognized as an internal or external command,
operable program or batch file.
Building Go cmd/dist using C:\Program Files\Go. (go version =)
When it should look like this:
Building Go cmd/dist using C:\Program Files\Go. (go1.19 windows/amd64)
For #44505.
Change-Id: I71328add5c74bd2829c0e23224cfa6252395ff2c
GitHub-Last-Rev: a01fda6b52
GitHub-Pull-Request: golang/go#54270
Reviewed-on: https://go-review.googlesource.com/c/go/+/421356
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Run-TryBot: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Alex Brainman <alex.brainman@gmail.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
The existing value for M1 is 64, which is the same as other arm64 cpus.
But the correct cacheLineSize for M1 should be 128, which can be
verified using the following command:
$ sysctl -a hw | grep cachelinesize
hw.cachelinesize: 128
Fixes#53075
Change-Id: Iaa8330010a4499b9b357c70743d55aed6ddb8588
GitHub-Last-Rev: df87eb9c50
GitHub-Pull-Request: golang/go#53076
Reviewed-on: https://go-review.googlesource.com/c/go/+/408576
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Meng Zhuo <mzh@golangcn.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Martin Möhrmann <martin@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Auto-Submit: Keith Randall <khr@google.com>
Extra Ms may lead to the "no consistent ordering of events possible" error when parsing trace file with cgo enabled, since:
1. The gs in the extra Ms may be in `_Gdead` status while starting trace by invoking `runtime.StartTrace`,
2. and these gs will trigger `traceEvGoSysExit` events in `runtime.exitsyscall` when invoking go functions from c,
3. then, the events of those gs are under non-consistent ordering, due to missing the previous events.
Add two events, `traceEvGoCreate` and `traceEvGoInSyscall`, in `runtime.StartTrace`, will make the trace parser happy.
Fixes#29707
Change-Id: I7cc4b80822d2c46591304a59c9da2c9fc470f1d0
GitHub-Last-Rev: 445de8eaf3
GitHub-Pull-Request: golang/go#53284
Reviewed-on: https://go-review.googlesource.com/c/go/+/411034
Run-TryBot: Michael Pratt <mpratt@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
The HTTP/1 server deletes multipart form tempfiles after ServeHTTP
returns, but the HTTP/2 server does not. Add a test to verify
cleanup happens in both cases, temporarily disabled for the
HTTP/2 path.
For #20253
Updates #25965
Change-Id: Ib753f2761fe73b29321d9d4337dbb5090fd193c2
Reviewed-on: https://go-review.googlesource.com/c/go/+/423194
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Two packages construct atomic booleans from atomic integers.
Replace these implementations with the new atomic.Bool type.
Indeed, these packages were the impetus for the new atomic.Bool
type, having demonstrated a need to access boolean values
atomically.
Change-Id: I6a0314f8e7d660984a6daf36a62ed05a0eb74b2f
Reviewed-on: https://go-review.googlesource.com/c/go/+/411400
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
When ReverseProxy forwards a request with no User-Agent header, leave
the header in the forwarded request blank rather than inserting the
default Go HTTP clent User-Agent.
We already did this for NewSingleHostReverseProxy; generalize it to
every ReverseProxy.
Change-Id: Id81a230cb8d384acdfae190b78a4265d80720388
Reviewed-on: https://go-review.googlesource.com/c/go/+/407375
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
X-Forwarded-Host contains the original request's host.
X-Forwarded-Proto contains "http" or "https", depending on whether the
original request was made on a TLS-secured connection.
Setting either header to nil in Director disables adding the header,
same as for X-Forwarded-For.
Fixes#50465.
Change-Id: If8ed1f48d83f8ea0389c53519bc7994cb53891db
Reviewed-on: https://go-review.googlesource.com/c/go/+/407414
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
The current implementation ignores cookies where the
cookie name starts or ends with a space. For example,
name =value
is ignored.
I have come across pages that send cookies in this weird format.
I tested with the latest versions of Firefox, Safari and Chrome,
all of which accept cookies in this format.
To do this, I remove leading and trailing spaces from the
cookie name after cutting at '='.
Change-Id: I8fd0c37a2113b6ce75712dd43607d1ea55e86c68
GitHub-Last-Rev: 368f50fcb4
GitHub-Pull-Request: golang/go#52121
Reviewed-on: https://go-review.googlesource.com/c/go/+/397734
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Damien Neil <dneil@google.com>
CL 310850 dropped work re-checks on non-spinning Ms to fix#43997.
This introduced a new race condition: a non-spinning M may drop its P
and then park at the same time a spinning M attempts to wake a P to
handle some new work. The spinning M fails to find an idle P (because
the non-spinning M hasn't quite made its P idle yet), and does nothing
assuming that the system is fully loaded. This results in loss of work
conservation. In the worst case we could have a complete deadlock if
injectglist fails to wake anything just as all Ps are going idle.
sched.needspinning adds new synchronization to cover this case. If work
submission fails to find a P, it sets needspinning to indicate that a
spinning M is required. When non-spinning Ms prepare to drop their P,
they check needspinning and abort going idle to become a spinning M
instead. This addresses the race without extra spurious wakeups. In the
normal (non-racing case), an M will become spinning via the normal path
and clear the flag.
injectglist must change in addition to wakep because it is a similar
form of work submission, notably used following netpoll at a point when
we might not have a P that would guarantee the work runs.
Fixes#45867
Change-Id: Ieb623a6d4162fb8c2be7b4ff8acdebcc3a0d69a8
Reviewed-on: https://go-review.googlesource.com/c/go/+/389014
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
Note that this converts npidle from uint32 to int32 for consistency with
the other count fields in schedt and the type of p.id.
Note that this changes previously unsynchronized operations to
synchronized operations in:
* handoffp
* injectglist
* schedtrace
* schedEnableUser
* sync_runtime_canSpin
For #53821.
Change-Id: I36d1b3b4a28131c9d47884fade6bc44439dd6937
Reviewed-on: https://go-review.googlesource.com/c/go/+/419445
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
Note that this changes the type from uint64 to int64, the type used by
nanotime(). It also adds an atomic load in pollWork(), which used to use
a non-atomic load.
For #53821.
Change-Id: I6173c90f20bfdc0e0a4bc3a7b1c798d1c429fff5
Reviewed-on: https://go-review.googlesource.com/c/go/+/419442
Run-TryBot: Michael Pratt <mpratt@google.com>
Reviewed-by: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
We currently print these as -1, but some are technically uint64. We can
be more explicit about their irrelevance by printing 'nil' rather than
-1.
Change-Id: I267fd8830564c75032bfe9176af59047f5a90202
Reviewed-on: https://go-review.googlesource.com/c/go/+/419441
Reviewed-by: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Michael Pratt <mpratt@google.com>
schedt.goidgen and p.goidcache are already uint64, this makes all cases
consistent.
The only oddball here is schedtrace which prints -1 as an equivalent for
N/A or nil. A future CL will make this more explicit.
Change-Id: I489626f3232799f6ca333d0d103b71d9d3aa7494
Reviewed-on: https://go-review.googlesource.com/c/go/+/419440
Reviewed-by: Austin Clements <austin@google.com>
Run-TryBot: Michael Pratt <mpratt@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
If a program only uses ecdh.P256(), the implementation of the other
curves shouldn't end up in the binary. This mostly required moving some
operations from init() time. Small performance hit in uncompressed
Bytes/SetBytes, but not big enough to show up in higher-level
benchmarks. If it becomes a problem, we can fix it by pregenerating the
p-1 bytes representation in generate.go.
For #52182
Updates #52221
Change-Id: I64460973b59ee3df787d7e967a6c2bcbc114ba65
Reviewed-on: https://go-review.googlesource.com/c/go/+/402555
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Fernando Lobato Meeser <felobato@google.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
We use crypto/internal/edwards25519/field to implement X25519 directly,
so that golang.org/x/crypto/curve25519 can be dropped from the src
module dependencies, and eventually replaced with a crypto/ecdh wrapper,
removing the need to keep golang.org/x/crypto/curve25519/internal/field
in sync with crypto/internal/edwards25519/field.
In crypto/internal/nistec, we add BytesX to serialize only the x
coordinate, which we'll need for the horrible ECDSA x-coord-to-scalar
operation, too.
In crypto/tls, we replace the ECDHE implementation with crypto/ecdh,
dropping the X25519 special cases and related scaffolding.
Finally, FINALLY, we deprecate the ~white whale~ big.Int-based APIs of
the crypto/elliptic package. •_•) ( •_•)>⌐■-■ (⌐■_■)
Fixes#52182Fixes#34648Fixes#52221
Change-Id: Iccdda210319cc892e96bb28a0e7b7123551982c7
Reviewed-on: https://go-review.googlesource.com/c/go/+/398914
Reviewed-by: Fernando Lobato Meeser <felobato@google.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
ONAME nodes are shared, so using their position for anything is almost
always a mistake. There are probably more instances of this mistake
elsewhere. For now, handle the case of map key temporaries, where it's
been a problem.
Fixes#53456.
Change-Id: Id44e845d08d428592ad3ba31986635b6b87b0041
Reviewed-on: https://go-review.googlesource.com/c/go/+/417076
Run-TryBot: David Chase <drchase@google.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Now that we've moved the trace locks to the leaf of the lock graph, we
can safely annotate that any trace event may acquire trace.lock even
if dynamically it turns out a particular event doesn't need to flush
and acquire this lock.
This reveals a new edge where we can trace while holding the mheap
lock, so we add this to the lock graph.
For #53789.
Updates #53979.
Change-Id: I13e2f6cd1b621cca4bed0cc13ef12e64d05c89a7
Reviewed-on: https://go-review.googlesource.com/c/go/+/418720
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Now that trace.lock cannot be held over a stack split, we can move
that lock and traceStackTab to the leaf of the lock graph. We add a
couple edges to STACKGROW that were previously passing through trace.
Fixes#53979.
Change-Id: Ie664ff7bb33973745f991f7516dc6106e60f5892
Reviewed-on: https://go-review.googlesource.com/c/go/+/418957
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently, trace.lock can be acquired while on a user G and stack
splits can happen while holding trace.lock. That means every lock used
by the stack allocator must be okay to acquire while holding
trace.lock, including various locks related to span allocation. In
turn, we cannot safely emit trace events while holding any
allocation-related locks because this would cause a cycle in the lock
rank graph.
To fix this, require that trace.lock only be acquired on the system
stack, like mheap.lock. This pushes it into the "bottom half" and
eliminates the lock rank relationship between tracing and stack
allocation, making it safe to emit trace events in many more places.
One subtlety is that the trace code has race annotations and uses
maps, which have race annotations. By default, we can't have race
annotations on the system stack, so we borrow the user race context
for these situations.
We'll update the lock graph itself in the next CL.
For #53979. This CL technically fixes the problem, but the lock rank
checker doesn't know that yet.
Change-Id: I9f5187a9c52a67bee4f7064db124b1ad53e5178f
Reviewed-on: https://go-review.googlesource.com/c/go/+/418956
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
We're about to require that all uses of trace.lock be on the system
stack. That's mostly easy, except that it's involving parking the
trace reader. Fix this by changing that parking protocol so it instead
synchronizes through an atomic.
For #53979.
Change-Id: Icd6db8678dd01094029d7ad1c612029f571b4cbb
Reviewed-on: https://go-review.googlesource.com/c/go/+/418955
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Writing out the trace footer currently manages trace buffers
differently from the rest of trace code. Rearrange it so it looks like
the rest of the code. In particular, we now write the frequency event
out to the trace buffer rather than returning it in a special byte
slice, and (*traceStackTable).dump threads a traceBufPtr like most
other functions that write to the trace buffers.
Change-Id: I3d0e108e56df884e7bd19823310dfbc0e21af9a5
Reviewed-on: https://go-review.googlesource.com/c/go/+/422974
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Following up on the previous CL, this CL removes a unnecessary stack
copy of a large object in a range loop. This drops another 64 KiB from
(*traceStackTable).dump's stack frame so it is now roughly 80 bytes
depending on architecture, which will easily fit on the system stack.
For #53979.
Change-Id: I16f642f6f1982d0ed0a62371bf2e19379e5870eb
Reviewed-on: https://go-review.googlesource.com/c/go/+/422955
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Austin Clements <austin@google.com>
Currently, the stack frame of (*traceStackTable).dump is 68KiB. We're
about to move (*traceStackTable).dump to the system stack, where we
often don't have this much room.
5140 bytes of this is an on-stack temporary buffer for constructing
potentially large trace events before copying these out to the actual
trace buffer.
Reduce the stack frame size by writing these events directly to the
trace buffer rather than temporary space. This introduces a couple
complications:
- The trace event starts with a varint encoding the event payload's
length in bytes. These events are large and somewhat complicated, so
it's hard to know the size ahead of time. That's not a problem with
the temporary buffer because we can just construct the event and see
how long it is. In order to support writing directly to the trace
buffer, we reserve enough bytes for a maximum size varint and add
support for populating a reserved space after the fact.
- Emitting a stack event calls traceFrameForPC, which can itself emit
string events. If these were emitted in the middle of the stack
event, it would corrupt the stream. We already allocate a []Frame to
convert the PC slice to frames, and then convert each Frame into a
traceFrame with trace string IDs, so we address this by combining
these two steps into one so that all trace string events are emitted
before we start constructing the stack event.
For #53979.
Change-Id: Ie60704be95199559c426b551f8e119b14e06ddac
Reviewed-on: https://go-review.googlesource.com/c/go/+/422954
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
In go.dev/cl/413396, I implemented implicit conversions by setting the
conversion's position to the enclosing statement that necessitated the
conversion. However, users actually want the position information to
be at the expression itself, and this seems sensible anyway.
This was noticed because x/tools had a test for:
fmt.Println(42)
and it was checking where the escape analysis diagnostic for
`42` (really `any(42)`) was reported.
Historically, we reported the column of the `4`; but CL 413396 caused
unified IR to instead report the column of the `(` instead (the
position associated with the call expression, which forced `42` to be
implicitly converted from `int` to `any`).
I chalk this mistake up to being accustomed to working with ir, where
we can't reliably use n.Pos() because of how ONAME positions work, so
I was trying to avoid relying on the implicitly converted expression's
own position.
Change-Id: I762076af6f65ebe6d444d64630722a5016dc2698
Reviewed-on: https://go-review.googlesource.com/c/go/+/422976
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
I structured the test for issue54343.go after issue46725.go, where I
was careful to use `[4]int`, which is a type large enough to avoid the
tiny object allocator (which interferes with finalizer semantics). But
in that test, I didn't note the importance of that type, so I
mistakenly used just `int` in issue54343.go.
This CL switches issue54343.go to use `[4]int` too, and then adds
comments to both pointing out the significance of this type.
Updates #54343.
Change-Id: I699b3e64b844ff6d8438bbcb4d1935615a6d8cc4
Reviewed-on: https://go-review.googlesource.com/c/go/+/423115
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: David Chase <drchase@google.com>
With GOEXPERIMENT=unified, the order variables are printed in "live at
entry to f.func1" is sensitive to whether regabi is enabled for some
reason. The order shouldn't matter to correctness, but it is odd.
For now, this CL just relaxes the test expectation order to unblock
enabling GOEXPERIMENT=unified by default. I've filed #54402 to
investigate further to confirm this a concern.
Updates #54402.
Change-Id: Iddfbb12c6cf7cc17b2aec8102b33761abd5f93ad
Reviewed-on: https://go-review.googlesource.com/c/go/+/422975
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
For selector expression "x.M" where "M" is a promoted method, irgen is using
the type of receiver "x" for determining the typeparams for instantiation.
However, because M is a promoted method, so its associated receiver is
not "x", but "x.T" where "T" is the embedded field of "x". That casues a
mismatch when converting non-shape types arguments.
Fixing it by using the actual receiver which has the method, instead of
using the base receiver.
Fixes#53982
Change-Id: I1836fc422d734df14e9e6664d4bd014503960bfc
Reviewed-on: https://go-review.googlesource.com/c/go/+/419294
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
When making the recursive call to fprint,
we just need to know how many bytes were written
and whether multiple lines were written.
We don't need a buffer to accomplish those;
a custom writer can keep track of the two in a cheap way,
avoiding the allocation of a byte slice and copying bytes.
name old time/op new time/op delta
Print-16 6.28ms ± 2% 6.12ms ± 1% -2.50% (p=0.000 n=10+9)
name old speed new speed delta
Print-16 8.26MB/s ± 3% 8.47MB/s ± 1% +2.56% (p=0.000 n=10+9)
name old alloc/op new alloc/op delta
Print-16 483kB ± 0% 443kB ± 0% -8.20% (p=0.000 n=10+10)
name old allocs/op new allocs/op delta
Print-16 17.8k ± 0% 17.3k ± 0% -2.31% (p=0.000 n=9+10)
Change-Id: Ib8411ae6738a2acae6af6d185da71727ce2eb97a
Reviewed-on: https://go-review.googlesource.com/c/go/+/412555
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
The current implementation of compilerVersion incorrectly gives an
error message that the compiler version is too old even though the
system has a recent compiler. This happens for specifically for the
gcc compiler and causes ASAN tests to be skipped.
Replacing -v with gcc dump version options seems to fix it. Running
./testsanitizers.test -test.v now shows the ASAN tests being run.
--- PASS: TestASAN (16.81s)
--- PASS: TestASAN/asan_useAfterReturn (0.60s)
--- PASS: TestASAN/asan_global5 (0.61s)
--- PASS: TestASAN/asan_unsafe_fail1 (0.73s)
--- PASS: TestASAN/asan_unsafe_fail3 (0.73s)
--- PASS: TestASAN/asan_unsafe_fail2 (0.74s)
--- PASS: TestASAN/asan_global4_fail (0.74s)
--- PASS: TestASAN/asan5_fail (0.74s)
--- PASS: TestASAN/asan3_fail (0.88s)
--- PASS: TestASAN/asan4_fail (0.89s)
--- PASS: TestASAN/asan2_fail (0.99s)
--- PASS: TestASAN/asan_global3_fail (1.00s)
--- PASS: TestASAN/asan_global1_fail (1.01s)
--- PASS: TestASAN/asan1_fail (1.01s)
--- PASS: TestASAN/asan_global2_fail (1.02s)
PASS
Fixes#54370
Change-Id: Iac13a1cf37de54432a6e49555f61e9ec1d781ab8
Reviewed-on: https://go-review.googlesource.com/c/go/+/422574
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
Run-TryBot: Archana Ravindar <aravind5@in.ibm.com>
Reviewed-by: Than McIntosh <thanm@google.com>
The comment claims that reflect users cannot obtain an unsafe.Pointer
without also importing the unsafe package explicitly.
This is no longer true now that the Value.UnsafePointer method
directly returns an unsafe.Pointer.
Change-Id: Ia5bf2e8aead681c8fac5b011129954d075ae5a43
Reviewed-on: https://go-review.googlesource.com/c/go/+/404396
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
When trampolines are needed (e.g. Darwin ARM64), the DWARF LPT (Line
Program Table - see DWARF section 6.1) generation fails because the
replacement symbols are marked as external symbols and skipped during
the DWARF LPT generation phase.
Fixes#54320
Change-Id: I6c93f5378f50e5edf30d5121402a48214abb1ce2
GitHub-Last-Rev: 085bbc55db
GitHub-Pull-Request: golang/go#54321
Reviewed-on: https://go-review.googlesource.com/c/go/+/422154
Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Than McIntosh <thanm@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Change the type of Conn.handshakeStatus from an atomically
accessed uint32 to an atomic.Bool. Change its name to
Conn.isHandshakeComplete to indicate it is a boolean value.
Eliminate the handshakeComplete() helper function, which checks
for equality with 1, in favor of the simpler
c.isHandshakeComplete.Load().
Change-Id: I084c83956fff266e2145847e8645372bef6ae9df
Reviewed-on: https://go-review.googlesource.com/c/go/+/422296
Auto-Submit: Filippo Valsorda <filippo@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Run-TryBot: Filippo Valsorda <filippo@golang.org>
This CL adds a helper expression code for receiver addressing; i.e.,
the implicit addressing, dereferencing, and field selections involved
in changing the `x` in `x.M()` into an appropriate expression to pass
as an argument to the method.
Change-Id: I9be933e2a38c8f94f6a85d95b54f34164e5efb0a
Reviewed-on: https://go-review.googlesource.com/c/go/+/421820
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Unified IR uses static dictionaries for some itabs and function/method
expressions, and they're roughly the right idea. But at the same time,
they're actually somewhat brittle and I need to reorganize some ideas
anyway to get shaped-based stenciling working. So this CL just rips
them out entirely.
Note: the code for emitting runtime dictionaries with *runtime._type
symbols is still present, and continues to demonstrate that basic
runtime dictionary handling is working.
Change-Id: I44eb1c7974fb397909ad5db12987659e7505c2ad
Reviewed-on: https://go-review.googlesource.com/c/go/+/421819
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Than McIntosh <thanm@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
This CL separates out the handling of selector expressions for field
values, method values, and method expressions. Again part of
refactoring to make it possible to access runtime dictionaries where
needed.
No behavioral change; just duplicating and then streamlining the
existing code paths.
Change-Id: I53b2a344f4bdba2c9f37ef370dc9a091a3941021
Reviewed-on: https://go-review.googlesource.com/c/go/+/421818
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
There was a TODO about quadratic performance, and indeed,
it can get bad. Added a map, made some integers that are
unlikely to exceed a few million into 32-bit integers.
Change-Id: I6facf2eabc00483e943b326ca8dcae2f778093da
Reviewed-on: https://go-review.googlesource.com/c/go/+/422297
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
This CL applies the same change to test/live.go that was previously
applied to test/live_regabi.go in golang.org/cl/415240. This wasn't
noticed at the time though, because GOEXPERIMENT=unified was only
being tested on linux-amd64, which is a regabi platform.
Change-Id: I0c75c2b7097544305e4174c2f5ec6ec283c81a8e
Reviewed-on: https://go-review.googlesource.com/c/go/+/422254
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
`go env GOEXPERIMENT` prints what experiments are enabled relative to
the baseline configuration, so it's not a very robust way to detect
what experiments have been statically enabled at bootstrap time.
Instead, we can check build.Default.ToolTags, which has goexperiment.*
for all currently enabled experiments, independent of baseline.
Change-Id: I6132deaa73b1e79ac24176ef4de5af67a507ee26
Reviewed-on: https://go-review.googlesource.com/c/go/+/422234
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: David Chase <drchase@google.com>
I've needed this more than once in the past, I hack it in,
then throw it away, seems sensible to make the change and
save it.
Fixes#53937.
Change-Id: I7fe886b1c93d73cbf553bed587f2c30f0f5d5a0b
Reviewed-on: https://go-review.googlesource.com/c/go/+/418015
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
For signed comparisons, the following four optimization rules hold:
(CMPconst [0] z:(AND x y)) && z.Uses == 1 => (TST x y)
(CMPWconst [0] z:(AND x y)) && z.Uses == 1 => (TSTW x y)
(CMPconst [0] x:(ANDconst [c] y)) && x.Uses == 1 => (TSTconst [c] y)
(CMPWconst [0] x:(ANDconst [c] y)) && x.Uses == 1 => (TSTWconst [int32(c)] y)
But currently they only apply to jump instructions, not to conditional
instructions within a block, such as cset, csel, etc. This CL extends
the above rules into blocks so that conditional instructions can also be
optimized.
name old time/op new time/op delta
DivisiblePow2constI64-160 1.04ns ± 0% 0.86ns ± 0% -17.30% (p=0.008 n=5+5)
DivisiblePow2constI32-160 1.04ns ± 0% 0.87ns ± 0% -16.16% (p=0.016 n=4+5)
DivisiblePow2constI16-160 1.04ns ± 0% 0.87ns ± 0% -16.03% (p=0.008 n=5+5)
DivisiblePow2constI8-160 1.04ns ± 0% 0.86ns ± 0% -17.15% (p=0.008 n=5+5)
Change-Id: I6bc34bff30862210e8dd001e0340b8fe502fe3de
Reviewed-on: https://go-review.googlesource.com/c/go/+/420434
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Eric Fang <eric.fang@arm.com>
When a generic interface method is used, we use a special
relocation R_USEGENERICIFACEMETHOD to tell the linker the name of
the generic interface method, so it can keep methods with that
name live. The relocation references a symbol whose content is the
name. Currently this is a string symbol, which is content
addessable and may have trailing zero bytes (for better
deduplication). The trailing bytes can cause confusion for the
linker. This symbol doesn't need to be in the final binary and
doesn't need to be deduplicated with other symbol. So we don't use
content addressable symbol but make an (unnamed) symbol
specifically for this.
May fix#54346.
Change-Id: If0c34f7844c3553a7be3846b66cf1c103bc231c4
Reviewed-on: https://go-review.googlesource.com/c/go/+/422300
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
We claim to not maintain pointer bits for noscan objects. But in fact
we do, since whenever we switch a page from scannable to noscan, we
call heapBits.initSpan which zeroes the heap bits.
Switch to ensure that we never scan noscan objects. This ensures that
we don't depend on the ptrbits for noscan objects. That fixes a bug
in the 1-bit bitmap CL which depended on that fact.
Change-Id: I4e66f582605b53732f8fca310c1f6bd2892963cb
Reviewed-on: https://go-review.googlesource.com/c/go/+/422435
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Keith Randall <khr@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
The Go 1.18 frontend handles package-scope generic method values by
spilling the receiver value to a global temporary variable, which pins
it into memory. This issue isn't present in unified IR, which uses
OMETHVALUE when the receiver type is statically known.
Updates #54343.
Change-Id: I2c4ffeb125a3cf338f949a93b0baac75fff6cd31
Reviewed-on: https://go-review.googlesource.com/c/go/+/422198
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
For interface method values, we nil check the receiver value at the
point of evaluating the method value. Currently this is inserted by
the backend during walk, but in some cases it's useful to emit them
upfront instead.
OITAB is essentially a field selection operation, like ODOT, OIDATA,
and OSPTR.
OCHECKNIL is a statement that simply evaluates its unary operand, and
discards the result (after testing for nil).
Change-Id: I583b5170539caa9a87aec661d5c293080fd87fbb
Reviewed-on: https://go-review.googlesource.com/c/go/+/422197
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
ir.ClosureExpr implements ir.InitNode, so ir.InitExpr can prepend init
statements to it. However, CalleeEffects wasn't aware of this and
could cause the init statements to get dropped when inlining a call to
a closure.
This isn't an issue today, because we don't create closures with init
statements. But I ran into this within unified IR.
Easy and robust solution: just take advantage that ir.TakeInit can
handle any node.
Change-Id: Ica05fbf6a8c5be4b11927daf84491a1140da5431
Reviewed-on: https://go-review.googlesource.com/c/go/+/422196
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
NewClosureVar should only be called to capture locally declared
variables in the enclosing function scope. This CL adds a check to
make sure it's used that way, in particular to make sure it's not
called to capture global variables.
This came up because for generic method values, we desugar the method
value into a function literal that captures the receiver value after
evaluating it. However, due to compiler backend limitations, for
package-scope generic method values we spill the receiver value into a
global variable rather than capturing it normally.
To prevent confusing backend issues when misusing NewClosureVar with
global variables, this CL adds an extra check.
Change-Id: I80f0f083dc80f70c7f0298020efe56dba00b67d7
Reviewed-on: https://go-review.googlesource.com/c/go/+/422195
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Skip a collection of -buildmode=pie tests on alpine, which are
currently failing on the linux-amd64-alpine builder. Once #54354 has
been investigated and resolved we can turn these tests back on.
Updates #54354.
Change-Id: I99d4016a40873ee6bb4eda571a64eddbe719c76a
Reviewed-on: https://go-review.googlesource.com/c/go/+/422295
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Currently the linux/arm64 and linux/riscv64 ports wrap the fstatat
syscall twice, once in func fstatat and once in func Fstatat. Change the
latter to be a simple wrapper around the former to deduplicate some
code.
Change-Id: I82ec9374e2bcfe116eabf01f021eed1ee810ec15
Reviewed-on: https://go-review.googlesource.com/c/go/+/412934
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
Reviewed-by: Than McIntosh <thanm@google.com>
As it can't appear in user package paths.
There is a hack for handling "go:buildid" and "type:*" on windows/386.
Previously, windows/386 requires underscore prefix on external symbols,
but that's only applied for SHOSTOBJ/SUNDEFEXT or cgo export symbols.
"go.buildid" is STEXT, "type.*" is STYPE, thus they are not prefixed
with underscore.
In external linking mode, the external linker can't resolve them as
external symbols. But we are lucky that they have "." in their name,
so the external linker see them as Forwarder RVA exports. See:
- https://docs.microsoft.com/en-us/windows/win32/debug/pe-format#export-address-table
- https://sourceware.org/git/?p=binutils-gdb.git;a=blob;f=ld/pe-dll.c;h=e7b82ba6ffadf74dc1b9ee71dc13d48336941e51;hb=HEAD#l972)
This CL changes "." to ":" in symbols name, so theses symbols can not be
found by external linker anymore. So a hacky way is adding the
underscore prefix for these 2 symbols. I don't have enough knowledge to
verify whether adding the underscore for all STEXT/STYPE symbols are
fine, even if it could be, that would be done in future CL.
Fixes#37762
Change-Id: I92eaaf24c0820926a36e0530fdb07b07af1fcc35
Reviewed-on: https://go-review.googlesource.com/c/go/+/317917
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Doing the test at link time lets us distribute one Linux toolchain
that works on both glibc-based and musl-based Linux systems.
The old way built a toolchain that only ran on one or the other.
Fixes#54197.
Change-Id: Iaae8c274c78e1091eee828a720b49646be9bfffe
Reviewed-on: https://go-review.googlesource.com/c/go/+/420774
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Arrange for tests that call setMimeInit to fully restore the old values,
by clearing the sync.Once that controls initialization.
Once we've done that, call initMime in initMimeUnixTest because
otherwise the test types loaded there will be cleared by the call to
initMime that previously was not being done.
For golang/go#51648
Change-Id: I8bf92b305fc4499337db06113817c9decdc5aedb
Reviewed-on: https://go-review.googlesource.com/c/go/+/421442
Reviewed-by: Than McIntosh <thanm@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Split should only split strings and not perform mangling
of invalid UTF-8 into ut8.RuneError.
The prior behavior is clearly a bug since mangling is not
performed in all other situations (e.g., separator is non-empty).
Fixes#53511
Change-Id: I112a2ef15ee46ddecda015ee14bca04cd76adfbf
Reviewed-on: https://go-review.googlesource.com/c/go/+/413715
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Run-TryBot: Ian Lance Taylor <iant@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Currently there is a an ANDconst and an ANDCCconst op in PPC64,
which is confusing since they map onto the same instruction.
One of these ops sets the result of the AND operation, and the
other sets the flag (condition register).
This converts ANDCCconst into an op with the 2 expected results:
the integer result of the AND and the flag setting. The ANDconst
op has been removed.
Note that in the PPC64 ISA the only variation of the 'and immediate'
is the one that sets the condition bit, which probably led to the
original (confusing) implementation.
This also adds a few rules to improve the use of ANDCCconst with
ISELB and some testcases to verify those improvements.
Change-Id: I523703fa4da2098eb995dc3ba744d36fa28e41d4
Reviewed-on: https://go-review.googlesource.com/c/go/+/422015
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Paul Murphy <murp@ibm.com>
Duff's device was disabled on darwin/arm64 because the darwin
linker couldn't handle a branch relocation with non-zero addend.
This is no longer the case now. The darwin linker can handle it
just fine. So enable it.
Fixes#54189.
Change-Id: Ida7ebafe6eb01db1af5bb8ae60a62491da5eabdf
Reviewed-on: https://go-review.googlesource.com/c/go/+/420894
Reviewed-by: Eric Fang <eric.fang@arm.com>
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
When types2 type checks a method expression or method value that
selects a type parameter method, the Selection.Index is indexed based
on the method's index within the type parameter's constraint
interface.
However, with a fully-stenciled implementation, naively using the
index would result in picking a method from the corresponding type
argument's full method set, which could select a different method.
Unified IR currently avoids this because it selects methods based on
name, not index; but experimenting with index-based selection revealed
that there are no test cases that would have caught this failure case.
Change-Id: Idbc39e1ee741714203d4749e47f5bc015af25020
Reviewed-on: https://go-review.googlesource.com/c/go/+/421815
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Mark the assembly routines as not escaping their arguments.
Add a special case to NewInt that, when inlined, can do all
of its allocations (a big.Int and a [1]Word) on the stack.
Update #29951
Change-Id: I9bd38c262eb97df98c0ed9874da7daac381243ea
Reviewed-on: https://go-review.googlesource.com/c/go/+/411254
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Robert Griesemer <gri@google.com>
Calling TypeOf to initialize variables forces any import of "reflect"
to link in the declared types of "reflect" even if they are unused.
TypeOf operates on Type and which will pull in
all transitive dependencies of Type, which includes Value as well.
Avoid this problem by declaring a rtypeOf function that
directly extracts the *rtype from an interface value
without going through Type as an intermediate type.
For a program that blank imports "reflect",
this reduces the binary size by ~34 KiB.
Updates #54097
Change-Id: I8dc7d8da8fedc48cc0dd842b69f510d17144827e
Reviewed-on: https://go-review.googlesource.com/c/go/+/419757
Run-TryBot: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Bryan Mills <bcmills@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
There's no need for a and b to match types. The typechecker already
ensured that a and b are both slices with the same base type, or
a and b are (possibly named) []byte and string.
The optimization to treat append(b, make([], ...)) as a zeroing
slice extension doesn't fire when there's a OCONVNOP wrapping the make.
Fixes#53888
Change-Id: Ied871ed0bbb8e4a4b35d280c71acbab8103691bc
Reviewed-on: https://go-review.googlesource.com/c/go/+/418475
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Read the bitmaps one uintptr at a time instead of one byte at a time.
Performance so far:
Allocation heavy, no retention: ~30% faster in heapBitsSetType
Scan heavy, ~no allocation: ~even in scanobject
Change-Id: I40d492b50d7f89d1b4261c2de58f6d255fa5e93e
Reviewed-on: https://go-review.googlesource.com/c/go/+/407036
Reviewed-by: Keith Randall <khr@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Use just 1 bit per word to record the ptr/nonptr bitmap.
Use word-sized operations to manipulate the bitmap, so we can operate
on up to 64 ptr/nonptr bits at a time.
Use a separate bitmap, one bit per word of the ptr/nonptr bitmap,
to encode a no-more-pointers signal. Since we can check 64 ptr/nonptr
bits at once, knowing the exact last pointer location is not necessary.
This cleans up the bitmap implementation significantly, which will
hopefully make it faster. TODO: measure
As a followon CL, we should make the gcdata bitmap an array of
uintptr instead of an array of byte, so we can load 64 bits of it at once.
Similarly for the processing of gc programs.
Change-Id: I18151b1876d9543599800dec51e2a1b19df97d49
Reviewed-on: https://go-review.googlesource.com/c/go/+/407035
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Keith Randall <khr@google.com>
Use the newly added atomic.Pointer[T] type for atomically
loading and storing type *T pointers. This has the advantage of
avoiding runtime type assertions required by its predecessor,
atomic.Value.
To fix build failures uncovered by TryBots (caused by "panic:
unaligned 64-bit atomic operation"), also change conn.curState to
type atomic.Uint64 so that it is 64-bit aligned.
Change-Id: I6024d12cd581adfdccc01be7eb0faa7482036614
Reviewed-on: https://go-review.googlesource.com/c/go/+/420901
Reviewed-by: Damien Neil <dneil@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Than McIntosh <thanm@google.com>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
OpenBSD 7.1 onwards expose the aarch64 ISAR0 and ISAR1 registers via sysctl:
$ sysctl machdep
machdep.compatible=apple,j274
machdep.id_aa64isar0=153421459058925856
machdep.id_aa64isar1=1172796674562
Implement CPU feature detection for openbsd/arm64 based on this information.
Fixes#31746
Change-Id: If8a9b2b8fc557e1aaefbcb52f4d1bd9efc43856d
Reviewed-on: https://go-review.googlesource.com/c/go/+/421875
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Tobias Klauser <tobias.klauser@gmail.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Run-TryBot: Joel Sing <joel@sing.id.au>
Run-TryBot: Ian Lance Taylor <iant@google.com>
Implement proposal #45454, providing build tags based on the
sub-architecture information in the GO$GOARCH variable
(for example, GOARM for GOARCH=arm).
For example, when GOAMD64=v2, the additional build tags
amd64.v1 and amd64.v2 are defined to be true.
Fixes#45454.
Change-Id: I7be56060d47fc61843b97fd8a78498e8202c1ee7
Reviewed-on: https://go-review.googlesource.com/c/go/+/421434
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Auto-Submit: Russ Cox <rsc@golang.org>
Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Current init() implementation in `encoding/gob/decode.go` checks int/uint/uintptr bit size with reflection in runtime. We could replace it with values available on compile stage. This should reduce time and allocations on binary start.
Results from GODEBUG=inittrace=1:
before:
init encoding/gob @4.4 ms, 0.21 ms clock, 43496 bytes, 652 allocs
after:
init encoding/gob @4.4 ms, 0.15 ms clock, 41672 bytes, 643 allocs
Updates #54184
Change-Id: I46dda2682fb92519da199415e29401d61edce697
Reviewed-on: https://go-review.googlesource.com/c/go/+/420455
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Rob Pike <r@golang.org>
Reviewed-by: Rob Pike <r@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
The RType field isn't needed when performing type assertions from
non-empty interface types, because we use the ITab field instead. But
the inline body exporter didn't know to expect this.
It's possible we could use a single bool to distinguish whether
we're serializing the RType or ITab field, but using two is simpler
and seems safer.
Fixes#54302.
Change-Id: I9ddac72784fb2241fee0a0dee30493d868a2c259
Reviewed-on: https://go-review.googlesource.com/c/go/+/421755
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Sometimes when implementing a Formatter it's helpful to use the fmt
package without invoking the formatter. This new function, FormatString,
makes that easier in some cases by recreating the original formatting
directive (such as "%3.2f") that caused Formatter.Format to be
called.
The original Formatter interface is probably not what we would
design today, but we're stuck with it. FormatString, although it
takes a State as an argument, compensates by making Formatter a
little more flexible.
The State does not include the verb so (unlike in the issue), we
must provide it explicitly in the call to FormatString. Doing it there
minimizes allocations by returning the complete format string.
Fixes#51668
Updates #51195
Change-Id: Ie31c8256515864b2f460df45fbd231286b8b7a28
Reviewed-on: https://go-review.googlesource.com/c/go/+/400875
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Run-TryBot: Russ Cox <rsc@golang.org>
The setting of the path for the dynamic loader when building for
linux/ppc64le ELF v2 was incorrectly set to the path for
PPC64 ELF v1. This has not caused issues in the common cases
because this string can be set based on the default GO_LDSO setting.
It does result in an incorrect value when cross compiling binaries
with -buildmode=pie.
Updates #53813
Change-Id: I84de1c97b42e0434760b76a57c5a05e055fbb730
Reviewed-on: https://go-review.googlesource.com/c/go/+/417614
Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
For a type definition like `type I interface{ M() }`, the go/types API
traditionally sets `M`'s receiver parameter type to `I`, whereas
Unified IR was (intentionally) leaving it as `interface{ M() }`.
I still think `interface{ M() }` is the more consistent and
semantically correct type to use in this scenario, but I concede that
users want `I` instead, as evidenced by existing tooling and tests.
Updates #49906.
Change-Id: I74ba5e8b08e4e98ed9dc49f72b7834d5b552058b
Reviewed-on: https://go-review.googlesource.com/c/go/+/421355
Reviewed-by: David Chase <drchase@google.com>
Auto-Submit: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
