We can't coalesce a non-WB store with a subsequent Move, as the
result of the store might be the source of the move.
There's a simple codegen test. Not sure how we might do a real test,
as all the repro's I've come up with are very expensive and unreliable.
Fixes#71228
Change-Id: If18bf181a266b9b90964e2591cd2e61a7168371c
Reviewed-on: https://go-review.googlesource.com/c/go/+/642197
Reviewed-by: Keith Randall <khr@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
Adds ACVP test coverage for the hmacDRBG algorithm based on the NIST
spec:
https://pages.nist.gov/ACVP/draft-vassilev-acvp-drbg.html#section-7.2
The HMAC DRBG algorithm in our fips module is a minimal implementation
tailored for use for generating ECDSA nonces and so lives in
crypto/internal/fips140/ecdsa.
In order to be testable by crypto/internal/fips140test this changeset
exports a ecdsa.TestingOnlyNewDrbg() constructor to support the ACVP use-case.
All FIPS-compatible SHA2 and SHA3 digests are tested.
The ACVP capability registration is customized to match the limited
capabilities of our ecdsa-focused impl. Most notably:
* reseedImplemented is false - we expect this impl to be invoked
only once or twice per instantiation and do not support explicit
reseeding.
* predResistanceEnabled is false - this requires reseeding.
* Per mode:
* derFuncEnabled is always false - this is only used by ctrDRBG.
* additionalInputLen is 0 for all modes - this is only used with
preResistanceEnabled.
The other capability values are chosen based on Table 4:
https://pages.nist.gov/ACVP/draft-vassilev-acvp-drbg.html#section-7.4
Updates #69642
Change-Id: Ia58979d691f912e2ed739a05efb719f580fbbf89
Reviewed-on: https://go-review.googlesource.com/c/go/+/639775
Reviewed-by: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Adds ACVP test coverage for ML-KEM based on the NIST spec:
https://pages.nist.gov/ACVP/draft-celi-acvp-ml-kem.html
Notably we need to update the BoringSSL module version because the
acvptool was only recently updated to support testing ML-KEM.
A few non-test updates are also required for the
crypto/internal/fips140/mlkem package:
* For keyGen tests a new ExpandedBytes768() function is added that
converts a DecapsualtionKey768 struct into the expanded NIST
serialization. The existing Bytes() function returns the
key's seed, while ACVP testing requires the more cumbersome format.
* For decap tests a new TestingOnlyNewDecapsulationKey768()
constructor is added to produce a DecapsulationKey768 struct from the
expanded FIPS 203 serialization provided by the ACVP test vector. The
pre-existing NewDecapsulationKey768() function expects a seed as
input.
The generate1024.go helper is updated to translate the above changes to
the generated mlkem1024.go implementation.
Both of these new functions are exclusively for ACVP usage and so not
present in the public mlkem API. End users should always prefer to work
with seeds.
Updates #69642
Change-Id: I79784f8a8db00a2ddefdcece4b8de50b033c8f69
Reviewed-on: https://go-review.googlesource.com/c/go/+/637439
Reviewed-by: Carlos Amedee <carlos@golang.org>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: David Chase <drchase@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Could have fixed this some other ways, including inside the FIPS 140-3
module, but this is small and self-contained, clearly not affecting
production non-toy key sizes. This late in the freeze, a surgical fix
felt best.
Fixes#71185
Change-Id: I6a6a465641357c9d6b076c8a520b221be4210ed5
Reviewed-on: https://go-review.googlesource.com/c/go/+/641755
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
Heap profiles hide "runtime" frames like runtime.mapassign. This broke
in 1.24 because the map implementation moved to internal/runtime/maps,
and runtime/pprof only considered literal "runtime." when looking for
runtime frames.
It would be nice to use cmd/internal/objabi.PkgSpecial to find runtime
packages, but that is hidden away in cmd.
Fixes#71174.
Change-Id: I6a6a636cb42aa17539e47da16854bd3fd8cb1bfe
Reviewed-on: https://go-review.googlesource.com/c/go/+/641775
Auto-Submit: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
The Mmap function returns a Data struct containing a slice with the
mapped contents of the file. Before this change, on Windows, the slice
contained the contents of all the pages of the mapping, including past
the end of the file. Re-slice the slice to the length of the file (if
if the slice is longer) so that the slice contains only the data in the
file.
For #71059
Change-Id: I389b752505b6fa1252b5c6d836a37bc7e662a45d
Reviewed-on: https://go-review.googlesource.com/c/go/+/640155
Reviewed-by: Russ Cox <rsc@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Both fips140only and the service indicator checks in
crypto/internal/fips140/... expect to type assert to
crypto/internal/fips140/{sha256,sha512,sha3}.Digest.
However, crypto/sha3 returns a wrapper concrete type around sha3.Digest.
Add a new fips140hash.Unwrap function to turn the wrapper into the
underlying sha3.Digest, and use it consistently before calling into
fips140only or the FIPS 140-3 module.
In crypto/rsa, also made the fips140only checks apply consistently after
the Go+BoringCrypto shims, so we can instantiate the hash, and avoid
having to wrap the New function. Note that fips140=only is incompatible
with Go+BoringCrypto.
Fixes#70879
Change-Id: I6a6a4656ec55c3e13f6cbfadb9cf89c0f9183bdc
Reviewed-on: https://go-review.googlesource.com/c/go/+/640855
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
In particular, cover the behavior of unmarshaling a JSON string
into a Number type regardless of whether the `string` option
is specified or not.
Change-Id: Ibc55f16860442240bcfeea1fd51aaa76f7e50f67
Reviewed-on: https://go-review.googlesource.com/c/go/+/641416
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Joseph Tsai <joetsai@digital-static.net>
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
The constants appeared badly ordered and grouped in the godoc before
const (
CiphertextSize1024 = 1568
EncapsulationKeySize1024 = 1568
)
const (
SharedKeySize = 32
SeedSize = 64
CiphertextSize768 = 1088
EncapsulationKeySize768 = 1184
)
while now they are a single group with the good size first
const (
SharedKeySize = 32
SeedSize = 64
CiphertextSize768 = 1088
EncapsulationKeySize768 = 1184
CiphertextSize1024 = 1568
EncapsulationKeySize1024 = 1568
)
No code changes.
Change-Id: I6a6a4656961b1e8c8bca3992aafa33e0575af8a2
Reviewed-on: https://go-review.googlesource.com/c/go/+/640997
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Russ Cox <rsc@golang.org>
Reviewed-by: Daniel McCarney <daniel@binaryparadox.net>
The original implementation of the type checkers accepted any boolean
result type for yield, but the compiler's front-end had a problem with
it (#71131).
As a temporary fix (for 1.24), adjust the type checkers to insist on the
spec's literal wording and avoid the compiler panic.
Fixes#71131.
For #71164.
Change-Id: Ie25f9a892e58b5e489d399b0bce2d0af55dc3c48
Reviewed-on: https://go-review.googlesource.com/c/go/+/640599
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Tim King <taking@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
New is called to get a Hash which can then be rejected with an error
(instead of a panic) from fips140only.ApprovedHash.
Also, it's reasonable to call New().Size() and then not use the hash.
Change-Id: I6a6a4656c43528d169c4b28c8b6de48448236d4f
Reviewed-on: https://go-review.googlesource.com/c/go/+/641317
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Daniel McCarney <daniel@binaryparadox.net>
Reviewed-by: Russ Cox <rsc@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Having a global lock on the random state (used only in FIPS-140 mode)
introduces contention in concurrent programs. Use an approximately
per-P random state instead, using sync.Pool to manage per-P state.
This code is important to land for the Go 1.24 release because it is
part of the FIPS-140 module that will be validated and certified,
so it will live for a long time. We otherwise wouldn't be able to
correct this contention for at least a year, perhaps more.
At the same time, the code is only used in the FIPS-140 mode,
so there is no risk to normal programs.
Fixes#71155.
Change-Id: I6b779f15ddfdf232f608f5cda08f75906e58114f
Reviewed-on: https://go-review.googlesource.com/c/go/+/641097
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
The backslashes on the windows paths will be escaped, so when checking
for them in the regular expression we'd have to have quadruple
backslashes '\\\\'. Since it's difficult to escape $GOCACHEPROG properly
for both json and regexp, just check for a string that ends in
cacheprog$GOEXE. We already check that the proper value is reported in
go env and go env -changed, and the json test case is mostly useful to
verify that GOCACHEPROG shows up in the json output.
For #71059
Change-Id: I52d49de61f2309a139f84c4d232b4cd94546ec8c
Cq-Include-Trybots: luci.golang.try:gotip-windows-amd64-longtest,gotip-linux-amd64-longtest
Reviewed-on: https://go-review.googlesource.com/c/go/+/641375
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Sam Thanawalla <samthanawalla@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Currently injectglist emits all the trace events before actually calling
casgstatus on each goroutine. This is a problem, since tracing can
observe an inconsistent state (gstatus does not match tracer's 'emitted
an event' state).
This change fixes the problem by having injectglist do what every other
scheduler function does, and that's wrap each call to casgstatus in
traceAcquire/traceRelease.
Fixes#70883.
Change-Id: I857e96cec01688013597e8efc0c4c3d0b72d3a70
Reviewed-on: https://go-review.googlesource.com/c/go/+/638558
Reviewed-by: Michael Pratt <mpratt@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Change-Id: I3e7a8498514da1d278acd566d526fdf6278f7d41
Reviewed-on: https://go-review.googlesource.com/c/go/+/638916
Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
For #67434
Change-Id: Iebcfc2559a62405fea7e3ceff8cf6c2f50b61def
Reviewed-on: https://go-review.googlesource.com/c/go/+/641176
Reviewed-by: Ian Lance Taylor <iant@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Damien Neil <dneil@google.com>
On Windows, we can't open a file that's already been opened. Before this
change, we'd try to write an index entry if mmapping the entry failed.
But that could happen either if the file doesn't exist or if there was a
problem mmapping an already opened file. Pass through information about
whether the file was actually opened so that we don't try to write to an
already opened file.
For #71059
Change-Id: I6adabe1093fed9ec37e7fafb13384c102786cbce
Reviewed-on: https://go-review.googlesource.com/c/go/+/640577
Reviewed-by: Sam Thanawalla <samthanawalla@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
These break if the tools are run with GODEBUG=fips140=only,
which happens if someone sets that during 'go test' (and a test
binary must be built).
The easiest fix is to make the tools compatible with this GODEBUG
by just using sha256 as the underlying hash always. Just in case,
I made the wrappers select different sections of the hash, but
none of the call sites really care.
This CL is for the Go 1.24 release, but a follow-up during the Go 1.25
dev cycle could change all the usage sites to only use Sum32/New32.
For #70514Fixes#70878
Change-Id: Id5fea779c83df51d1680dbe561e0949c56e8d1e5
Reviewed-on: https://go-review.googlesource.com/c/go/+/641096
Reviewed-by: Keith Randall <khr@golang.org>
Auto-Submit: Russ Cox <rsc@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Keith Randall <khr@google.com>
Allow a gocacheprog to not respond to close. The intention of the code
is that after we send the close message we'd ignore errors reading from
the cacheprog's stdout. But before this change if a cacheprog
did not respond to close and we got an EOF reading from the cacheprog's
stdout we'd just ignore all pending requests. The send operation would
then block forever waiting for a response. With this change, we close
all response channels for pending responses if there's an error reading
from the cacheprog's stdout while we're closing. The receives from the
response channels would then proceed (but now have to handle a nil
value). Then the send operation would return and the (*ProgCache).Close
function can proceed.
Fixes#70848
Change-Id: I6631d317ba7aea3f25f714f31cd2aeef0f4d4e3e
Cq-Include-Trybots: luci.golang.try:gotip-linux-amd64-longtest,gotip-windows-amd64-longtest
Reviewed-on: https://go-review.googlesource.com/c/go/+/640516
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: Sam Thanawalla <samthanawalla@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Currently weak.Pointer.Value will panic if the weak.Pointer is
uninitialized (zero value) which goes against it's documentation. Fix
this and add a test. While we're here, also add a test to ensure
weak.Make[T](nil) is equivalent to the zero value of weak.Pointer[T].
Fixes#71153.
Change-Id: I4d9196026360bc42a5bfcb33ce449131ec251dba
Reviewed-on: https://go-review.googlesource.com/c/go/+/641095
Reviewed-by: David Finkel <david.finkel@gmail.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Carlos Amedee <carlos@golang.org>
This case is not properly handled by the type checkers (see issue)
but the compiler uses the parser's label checking so it works as
expected.
For #70974.
Change-Id: I0849376bf7514a9a7730846649c3fe28c91f44ca
Reviewed-on: https://go-review.googlesource.com/c/go/+/640895
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Alan Donovan <adonovan@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Function types for function (and method) declarations do not
appear in Info.Types maps, only Info.Defs maps, because the
function type is implicit in the declaration and not a proper
(function) type expression. This is true even though the AST
represents these types via an (artificial) FuncType node.
Document this explicitly in the API.
No functional code changes.
Fixes#70908.
Change-Id: I2aa897daed04e7ad0fa8b625d9adc7b423c57387
Reviewed-on: https://go-review.googlesource.com/c/go/+/640776
Reviewed-by: Alan Donovan <adonovan@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
The spec indicates that if a client sends an invalid ECHClientHello.type
in ClientHelloOuter, the server will abort the handshake with a
decode_error alert.
Define errInvalidECHExt for invalid ECHClientHello.type. If parseECHExt
returns an errInvalidECHExt error, Conn now sends an illegal_parameter
alert.
Fixes#71061.
Change-Id: I240241fe8bbe3e77d6ad1af989794647bfa2ff87
GitHub-Last-Rev: 3d6c233ccd
GitHub-Pull-Request: golang/go#71062
Reviewed-on: https://go-review.googlesource.com/c/go/+/639235
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
There is no point to zeroise anything here because there are no secrets,
but there is a strict FIPS 140-3 test requirement for it.
> TE05.08.02 (Levels 1, 2, 3, and 4): verify that any temporary values
> generated during the integrity test are zeroised upon completion of
> the integrity test
Change-Id: I7b0db075dae9910f8e825a22ca2caa2b4c918980
Reviewed-on: https://go-review.googlesource.com/c/go/+/636556
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Daniel McCarney <daniel@binaryparadox.net>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
disasm_riscv64 currently always returns an instruction length of four,
which is not correct if compressed instructions are in use. Return the
length of the decoded instruction, defaulting to two bytes if the
instruction is unknown.
With this change it is possible to correctly objdump a binary that is
written in C and includes compressed instructions:
$ go tool objdump ./hello
TEXT _start(SB)
:0 0x5b0 ef002002 CALL 8(PC)
:0 0x5b4 aa87 ADD X10, X0, X15
:0 0x5b6 17250000 AUIPC $2, X10
:0 0x5ba 033525a3 MOV -1486(X10), X10
:0 0x5be 8265 MOV (X2), X11
:0 0x5c0 3000 ADDI $8, X2, X12
...
Fixes#71102
Change-Id: Ia99eb114a98c6d535de872ce8a526cd5e6203fff
Reviewed-on: https://go-review.googlesource.com/c/go/+/639995
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Reviewed-by: Jorropo <jorropo.pgm@gmail.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
This has no practical advantage, and requires extra variable time code,
but is an explicit FIPS 186-5 requirement.
Note that the new behavior is consistent with Go+BoringCrypto, but not
with Go 1.23. The resulting keys are essentially interchangeable, but
it's not impossible for applications to notice (google/go-tpm#383).
gcd_lcm_tests.txt is from BoringSSL.
Change-Id: I6a6a4656fd5e92912c87bedc667456d0e8787023
Reviewed-on: https://go-review.googlesource.com/c/go/+/639936
Reviewed-by: Russ Cox <rsc@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
This test checks a use-case of sync.Map that's expected to be more
common in Go 1.24 and beyond, as a concurrent weak cache.
The test will also fail if CompareAndSwap is not properly atomic with
CompareAndDelete, which is what #70970 is actually about. We should have
more explicit tests checking mutual atomicity of operations, but
for now this is OK, and still useful.
For #70970.
Change-Id: I6db508660691586a8af9ad511c9a96432d333343
Reviewed-on: https://go-review.googlesource.com/c/go/+/640737
Reviewed-by: David Chase <drchase@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Fixes#70874
Change-Id: Icbcfc95e6b45521880287dcc3bc8609461a3b401
GitHub-Last-Rev: 05276c56b0
GitHub-Pull-Request: golang/go#70877
Reviewed-on: https://go-review.googlesource.com/c/go/+/637035
Reviewed-by: qiu laidongfeng2 <2645477756@qq.com>
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: David Chase <drchase@google.com>
Auto-Submit: Ian Lance Taylor <iant@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
This reverts CL 606462.
Reason for revert: Breaks atomicity between operations. See #70970.
Change-Id: I1a899f2784da5a0f9da3193e3267275c23aea661
Reviewed-on: https://go-review.googlesource.com/c/go/+/638615
Auto-Submit: Michael Knyszek <mknyszek@google.com>
Commit-Queue: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: David Chase <drchase@google.com>
FIPS 186-4 used to defer to ANSI X9.62-2005, which is not freely
available, so we were referring to SEC 1 instead. Our new reference,
FIPS 186-5, actually specifies the full algorithm, so there is no need
to refer to SEC 1 anymore.
Change-Id: Ief499d0f7778f3221547993e9e8951ae873aacef
Reviewed-on: https://go-review.googlesource.com/c/go/+/640115
Reviewed-by: Daniel McCarney <daniel@binaryparadox.net>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
In documentation, we've usually but not always referred to a
context with a closed Done channel as "done" rather than
"canceled", to avoid ambiguity between a context canceled
by calling a CancelFunc and one past its deadline.
This actually adds ambiguity, however, since it's common to
see references to a "canceled context" that are intended to
cover contexts past their deadline. If you see "function F
returns if its context is canceled", you can reasonably
assume that F will return if its context passes its
deadline, unless something says otherwise.
Update the context package docs to explicitly state that
a context is canceled when its deadline passes. Drop references
to contexts becoming "done" and just use "canceled" throughout.
Fixes#70945
Change-Id: I99fbd800c6049deaa37015a304f7f9d9a84100e1
Reviewed-on: https://go-review.googlesource.com/c/go/+/640095
Reviewed-by: Ian Lance Taylor <iant@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Damien Neil <dneil@google.com>
For #71112
Change-Id: Ifda4fc8de148c42a2154da54b53d7215b9a6faa0
Reviewed-on: https://go-review.googlesource.com/c/go/+/640175
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Ian Lance Taylor <iant@golang.org>
Auto-Submit: Ian Lance Taylor <iant@google.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Since OpenBSD 7.3, external linking uses -fexecute-only, which breaks
the integrity check. Since we are not validating on OpenBSD anyway,
mark it as unsupported at least for now.
Fixes#70880
Change-Id: I6a6a4656b6c7a97c0962b4158d920f9e6b19678e
Reviewed-on: https://go-review.googlesource.com/c/go/+/639337
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Daniel McCarney <daniel@binaryparadox.net>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Joel Sing <joel@sing.id.au>
Reviewed-by: صادق <sadq04724@gmail.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Move the logic duplicated in multiple places to a central function.
Change-Id: I6a6a4656469c91dd62b0be716ec8367358f4a3e1
Reviewed-on: https://go-review.googlesource.com/c/go/+/639336
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Daniel McCarney <daniel@binaryparadox.net>
Reviewed-by: Roland Shoemaker <roland@golang.org>
It's not being used anywhere, remove it.
Change-Id: I9c3eecacd4e5d44b18243bdec24ad88bc38c82bf
Reviewed-on: https://go-review.googlesource.com/c/go/+/639957
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Roland Shoemaker <roland@golang.org>
Change-Id: I9703c6a4a2ae4a608d33cf706106c92e9bd2aef7
GitHub-Last-Rev: 3eee41957d
GitHub-Pull-Request: golang/go#71096
Reviewed-on: https://go-review.googlesource.com/c/go/+/637962
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
check.Enabled, internal/fips140.Enabled, and crypto/fips140.Enabled were
redundant. Package check can just use internal/fips140.Enabled.
check.Verified is still there for the tests and belt-and-suspenders
assurance in crypto/fips140.Enabled, although it's implied by Enabled.
For #69536
Change-Id: I83921cc925da841aba4da79a9a5e9ac526a3f2bf
Reviewed-on: https://go-review.googlesource.com/c/go/+/638855
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Daniel McCarney <daniel@binaryparadox.net>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
Change-Id: Ib750438107db6c82020cfb4abbab52435012b7fc
GitHub-Last-Rev: 3fa9b8c7bc
GitHub-Pull-Request: golang/go#71082
Reviewed-on: https://go-review.googlesource.com/c/go/+/639217
Auto-Submit: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Michael Matloob <matloob@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Sam Thanawalla <samthanawalla@google.com>
This adds GOCACHEPROG to the list of environment variables in "go help
environment" and points to the cacheprog package documentation for
details of the protocol.
Fixes#71032
Updates #59719
Change-Id: Ib8f5804926a8fa59237661076d129c2852665ac3
Reviewed-on: https://go-review.googlesource.com/c/go/+/638997
Auto-Submit: Austin Clements <austin@google.com>
Reviewed-by: Mauri de Souza Meneguzzo <mauri870@gmail.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Now that these types are in their own package, drop the unnecessary
Prog prefixes from everything.
Updates #71032
Updates #59719
Change-Id: Id54edf0473754e3b21a71beb72803fb5481206c1
Reviewed-on: https://go-review.googlesource.com/c/go/+/638996
Reviewed-by: Mauri de Souza Meneguzzo <mauri870@gmail.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Austin Clements <austin@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This is a step toward making it easy to point to them in
documentation. The other option is that we copy-paste all of these
type definitions wholesale, which seems ridiculous.
Updates #71032
Updates #59719
Change-Id: I7117e03308ae0adc721ed7a57792c33ba68ce827
Reviewed-on: https://go-review.googlesource.com/c/go/+/638995
Auto-Submit: Austin Clements <austin@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
When ECH is rejected, properly take retry configs from the encrypted
extensions message. Also fix the bogo shim to properly test for this
behavior.
We should properly map the full BoringSSL -> Go errors so that we don't
run into a similar failure in the future, but this is left for a follow
up CL.
Fixes#70915
Change-Id: Icc1878ff6f87df059e7b83e0a431f50f1fea833c
Reviewed-on: https://go-review.googlesource.com/c/go/+/638583
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
This is in preparation for adding a "go help" topic for GOCACHEPROG.
Updates #71032
Updates #59719
Change-Id: I9dbbe56fa328dffe89207b5b41a0f37afd51e2b5
Reviewed-on: https://go-review.googlesource.com/c/go/+/638566
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Austin Clements <austin@google.com>
Reviewed-by: Michael Matloob <matloob@golang.org>
Keith Randall
Daniel McCarney
Filippo Valsorda
Michael Pratt
Sean Liao
Michael Matloob
Joe Tsai
Robert Griesemer
Russ Cox
Michael Anthony Knyszek
Koichi Shiraishi
Damien Neil
thekuwayama
Joel Sing
yincong
Ian Lance Taylor
Jakub Ciolek
yaxum62
Mateusz Poliwczak
linmaolin
Austin Clements
Roland Shoemaker