1
0
mirror of https://github.com/golang/go synced 2024-11-11 17:51:49 -07:00
Commit Graph

60897 Commits

Author SHA1 Message Date
Robert Griesemer
e4f9cb5f77 go/types, types2: remove unused argument from Checker.compositeLit
Change-Id: I4f52503bab569d85f28ae8920079de26a331b7ff
Reviewed-on: https://go-review.googlesource.com/c/go/+/610560
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Tim King <taking@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
2024-09-05 22:12:07 +00:00
Robert Griesemer
0fc3b127a0 go/types, types2: factor type checking of basic literals and generate go/types code
Move the code for type checking of basic literals into literals.go.

In go/types, the respective code is now generated from the types2 source.

Change-Id: Ib21eb7a87e11b77bcb2469985f9844964d35df57
Reviewed-on: https://go-review.googlesource.com/c/go/+/610540
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Tim King <taking@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
2024-09-05 22:12:04 +00:00
Robert Griesemer
0f12e51ce7 go/types: factor out opPos computation
Adjust Checker.overflow call sites to match types2
where possible.

Change-Id: Iaa0d423f2ebf642428c745c4ac4f712e4136dffb
Reviewed-on: https://go-review.googlesource.com/c/go/+/610956
Reviewed-by: Robert Griesemer <gri@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Tim King <taking@google.com>
2024-09-05 22:10:13 +00:00
Robert Griesemer
a17356cd23 go/types, types2: factor type checking of func literals and generate go/types code
Move the code for type checking of function literals into
literals.go.

In go/types, the respective code is now generated from the types2 source.

Change-Id: Ic81ab3c0d3c66d99bc0f2e21d66bf9a896ef9375
Reviewed-on: https://go-review.googlesource.com/c/go/+/610996
Reviewed-by: Tim King <taking@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Robert Griesemer <gri@google.com>
2024-09-05 22:09:30 +00:00
Robert Griesemer
829d1b346f go/types, types2: factor out endPos call
Preparation for generation of function literal type checking code
from types2 source.

Change-Id: I6b3029c34c2507d356ac8874154537bc6c38a715
Reviewed-on: https://go-review.googlesource.com/c/go/+/610995
Reviewed-by: Tim King <taking@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
2024-09-05 22:09:28 +00:00
Robert Griesemer
e3f3083113 go/types: generate literals.go from corresponding types2 source
Change-Id: I0635101b984725ee24c2207ebfdb413d29212b67
Reviewed-on: https://go-review.googlesource.com/c/go/+/610558
Reviewed-by: Tim King <taking@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
2024-09-05 22:09:24 +00:00
Robert Griesemer
aca0e2bb27 go/types, types2: factor out isdddArray and inNode helper functions
Preparation for generation of go/types/literals.go from types2 sources.

Change-Id: I9af23fbe1e448976394ddd7b348188c2595d8afe
Reviewed-on: https://go-review.googlesource.com/c/go/+/610557
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Tim King <taking@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
2024-09-05 22:09:22 +00:00
Robert Griesemer
0fb35efe5b go/types, types2: factor out code for type checking composite literals
Move code into separate function in separate file.
Replace "goto Error" statements with "x.mode = invalid; return".
No other semantic changes.

Change-Id: I2d5e858e8df3dc1011fa79cdac3db9d3e7b1dfe5
Reviewed-on: https://go-review.googlesource.com/c/go/+/610556
Reviewed-by: Tim King <taking@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Robert Griesemer <gri@google.com>
2024-09-05 22:09:19 +00:00
Robert Griesemer
1e2114349d cmd/compile/internal/syntax: handle parentheses around constraints consistently
Generally, the parser strips (i.e., does not record in the syntax tree)
unnecessary parentheses. Specifically, given a type parameter list of
the form

        [P (C),]

it records it as

        [P C]

and then no comma is required when printing. However it did only strip
one level of parentheses, and

        [P ((C)),]

made it through, causing a panic when printing. Somewhat related,
the printer stripped parentheses around constraints as well.

This CL implements a more consistent behavior:

1) The parser strips all parentheses around constraints. For testing
   purposes, a local flag (keep_parens) can be set to retain the
   parentheses.

2) The printer code now correctly intruces a comma if parentheses
   are present (e.g., when testing with keep_parens). This case does
   not occur in normal operation.

3) The printer does not strip parentheses around constraints since
   the parser does it already.

For #69206.

Change-Id: I974a800265625e8daf9477faa9ee4dd74dbd17ce
Reviewed-on: https://go-review.googlesource.com/c/go/+/610758
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Robert Findley <rfindley@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Robert Griesemer <gri@google.com>
2024-09-05 20:52:22 +00:00
Cuong Manh Le
32bd777a37 cmd/compile: use slices.SortStableFunc
Now that we're bootstrapping from a toolchain that has the slices
package.

Updates #64751

Change-Id: I876ec6d261466344faf33f8c5cda229dd1e4185f
Reviewed-on: https://go-review.googlesource.com/c/go/+/610602
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
2024-09-05 18:56:40 +00:00
Cuong Manh Le
f15095f543 cmd/compile: use slices.{Sort,SortFunc}
Now that we're bootstrapping from a toolchain that has the slices
package.

Updates #64751

Change-Id: I2e63d95577d058670d3dc75bd45d6e050c6f0e25
Reviewed-on: https://go-review.googlesource.com/c/go/+/610601
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
2024-09-05 18:56:37 +00:00
Tobias Klauser
634363e3ca cmd/cgo: use slices.Index
Now that Go 1.22.6 is the minimum bootstrap toolchain (cf. CL 606156),
the slices package (introduced in Go 1.21) can be used in packages built
using the bootstrap toolchain.

For #64751

Change-Id: Ife0daa37c0982d9ec1afab07b9d40a1dfee9b7d4
Reviewed-on: https://go-review.googlesource.com/c/go/+/610575
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Tobias Klauser <tobias.klauser@gmail.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Ian Lance Taylor <iant@google.com>
2024-09-05 18:54:24 +00:00
Mateusz Poliwczak
c3f16307b3 go/printer: do not treat comments inside a ast.Decl as godoc
This change makes sure that we do not format comments
as doc comments inside of a declaration and makes the
go doc formatter idempotent:

Previously:

	// test comment
	//go:directive2
	// test comment
func main() {
}

was formatted to:

// test comment
//go:directive2
// test comment
func main() {
}

after another formatting, it got formatted with doc rules into:

// test comment
// test comment
//
//go:directive2
func main() {
}

With this change it gets directly to the correct form (last one).

Change-Id: Id7d8f03e43474357cd714e0672e886652c3fce86
GitHub-Last-Rev: 9833b87536
GitHub-Pull-Request: golang/go#69134
Reviewed-on: https://go-review.googlesource.com/c/go/+/609077
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
2024-09-05 18:53:36 +00:00
Kir Kolyshkin
89958ab4ae cmd/go/internal: use t.TempDir in tests
Change-Id: I8b4c19ed1085d2ffb07e2c8db33a10b6d70988eb
Reviewed-on: https://go-review.googlesource.com/c/go/+/611015
Reviewed-by: Michael Matloob <matloob@golang.org>
Reviewed-by: Ian Lance Taylor <iant@google.com>
Auto-Submit: Ian Lance Taylor <iant@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
2024-09-05 18:50:16 +00:00
Roland Shoemaker
f22d731976 go/build/constraint: add parsing limits
Limit the size of build constraints that we will parse. This prevents a
number of stack exhaustions that can be hit when parsing overly complex
constraints. The imposed limits are unlikely to ever be hit in real
world usage.

Fixes #69141
Fixes CVE-2024-34158

Change-Id: I38b614bf04caa36eefc6a4350d848588c4cef3c4
Reviewed-on: https://go-internal-review.googlesource.com/c/go/+/1540
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: Russ Cox <rsc@google.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/611240
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>
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
2024-09-05 16:42:11 +00:00
Roland Shoemaker
08c84420bc encoding/gob: cover missed cases when checking ignore depth
This change makes sure that we are properly checking the ignored field
recursion depth in decIgnoreOpFor consistently. This prevents stack
exhaustion when attempting to decode a message that contains an
extremely deeply nested struct which is ignored.

Thanks to Md Sakib Anwar of The Ohio State University (anwar.40@osu.edu)
for reporting this issue.

Fixes #69139
Fixes CVE-2024-34156

Change-Id: Iacce06be95a5892b3064f1c40fcba2e2567862d6
Reviewed-on: https://go-internal-review.googlesource.com/c/go/+/1440
Reviewed-by: Russ Cox <rsc@google.com>
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/611239
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
2024-09-05 16:42:09 +00:00
Roland Shoemaker
dd2019528b go/parser: track depth in nested element lists
Prevents stack exhaustion with extremely deeply nested literal values,
i.e. field values in structs.

Fixes #69138
Fixes CVE-2024-34155

Change-Id: I2e8e33b44105cc169d7ed1ae83fb56df0c10f1ee
Reviewed-on: https://go-internal-review.googlesource.com/c/go/+/1520
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: Russ Cox <rsc@google.com>
Reviewed-on: https://go-review.googlesource.com/c/go/+/611238
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org>
2024-09-05 16:42:07 +00:00
Cuong Manh Le
73fa90e290 types2, go/types: use slices.SortFunc
Now that we're bootstrapping from a toolchain that has the slices
package.

Updates #64751

Change-Id: I3227e55f87e033dae63a2d1712b7f9373fe49731
Reviewed-on: https://go-review.googlesource.com/c/go/+/610603
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Robert Findley <rfindley@google.com>
2024-09-05 16:39:57 +00:00
Meng Zhuo
2707d42966 runtime: use regabi for riscv64 cputicks
goos: linux
goarch: riscv64
pkg: runtime
cpu: Spacemit(R) X60
                     │ select.old.log │           select.new.log           │
                     │     sec/op     │   sec/op     vs base               │
SelectUncontended         490.5n ± 0%   486.8n ± 0%  -0.77% (p=0.000 n=10)
SelectSyncContended       2.754µ ± 0%   2.726µ ± 0%  -1.02% (p=0.000 n=10)
SelectAsyncContended      488.2n ± 0%   484.2n ± 0%  -0.84% (p=0.000 n=10)
SelectNonblock            112.2n ± 0%   111.5n ± 0%  -0.58% (p=0.000 n=10)
SelectProdCons            1.420µ ± 0%   1.417µ ± 0%       ~ (p=0.069 n=10)
GoroutineSelect           10.79m ± 3%   10.74m ± 3%       ~ (p=0.529 n=10)
geomean                   3.228µ        3.208µ       -0.63%

Change-Id: Idb519ef8b2872284dca6dbf1cf94c3fff65bfd37
Reviewed-on: https://go-review.googlesource.com/c/go/+/610095
Reviewed-by: abner chenc <chenguoqi@loongson.cn>
Reviewed-by: Joel Sing <joel@sing.id.au>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
2024-09-05 01:59:23 +00:00
Xiaolin Zhao
db07c8607a cmd/internal/obj/loong64: add support for instructions ANDN and ORN
Go asm syntax:
	ANDN/ORN	RK, RJ, RD
    or  ANDN/ORN	RK, RD

Equivalent platform assembler syntax:
	andn/orn	rd, rj, rk
    or  andn/orn	rd, rd, rk

Ref: https://loongson.github.io/LoongArch-Documentation/LoongArch-Vol1-EN.html

Change-Id: I6d240ecae8f9443811ca450aed3574f13f0f4a81
Reviewed-on: https://go-review.googlesource.com/c/go/+/610475
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Commit-Queue: abner chenc <chenguoqi@loongson.cn>
Reviewed-by: abner chenc <chenguoqi@loongson.cn>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: abner chenc <chenguoqi@loongson.cn>
2024-09-05 00:48:33 +00:00
Xiaolin Zhao
501b389efc runtime: optimize the function memclrNoHeapPointers on loong64
The relevant performance improved by 54.61%.

benchmark:
goos: linux
goarch: loong64
pkg: runtime
cpu: Loongson-3A6000 @ 2500.00MHz
                        │     old      │                 new                 │
                        │    sec/op    │   sec/op     vs base                │
Memclr/5                   4.803n ± 0%   2.801n ± 0%  -41.68% (p=0.000 n=20)
Memclr/16                  4.803n ± 0%   3.202n ± 0%  -33.33% (p=0.000 n=20)
Memclr/64                  9.605n ± 0%   5.061n ± 1%  -47.30% (p=0.000 n=20)
Memclr/256                 29.22n ± 0%   10.24n ± 0%  -64.96% (p=0.000 n=20)
Memclr/4096                413.4n ± 0%   106.9n ± 0%  -74.14% (p=0.000 n=20)
Memclr/65536               6.566µ ± 0%   1.673µ ± 0%  -74.52% (p=0.000 n=20)
Memclr/1M                 104.95µ ± 0%   52.51µ ± 0%  -49.97% (p=0.000 n=20)
Memclr/4M                  419.8µ ± 0%   209.9µ ± 0%  -49.99% (p=0.000 n=20)
Memclr/8M                  839.6µ ± 0%   419.9µ ± 0%  -49.98% (p=0.000 n=20)
Memclr/16M                1687.6µ ± 0%   845.3µ ± 0%  -49.91% (p=0.000 n=20)
Memclr/64M                 6.725m ± 0%   3.389m ± 0%  -49.61% (p=0.000 n=20)
MemclrUnaligned/0_5        6.003n ± 0%   4.581n ± 0%  -23.69% (p=0.000 n=20)
MemclrUnaligned/0_16       6.005n ± 0%   5.084n ± 0%  -15.33% (p=0.000 n=20)
MemclrUnaligned/0_64      10.810n ± 0%   6.229n ± 0%  -42.38% (p=0.000 n=20)
MemclrUnaligned/0_256      30.43n ± 0%   10.68n ± 0%  -64.90% (p=0.000 n=20)
MemclrUnaligned/0_4096     414.8n ± 0%   107.1n ± 0%  -74.18% (p=0.000 n=20)
MemclrUnaligned/0_65536    6.566µ ± 0%   1.700µ ± 0%  -74.11% (p=0.000 n=20)
MemclrUnaligned/1_5        6.003n ± 0%   4.582n ± 0%  -23.67% (p=0.000 n=20)
MemclrUnaligned/1_16      11.610n ± 0%   5.080n ± 0%  -56.24% (p=0.000 n=20)
MemclrUnaligned/1_64      16.810n ± 0%   7.370n ± 0%  -56.16% (p=0.000 n=20)
MemclrUnaligned/1_256      36.42n ± 0%   12.95n ± 0%  -64.44% (p=0.000 n=20)
MemclrUnaligned/1_4096     420.6n ± 0%   114.6n ± 0%  -72.75% (p=0.000 n=20)
MemclrUnaligned/1_65536    6.573µ ± 0%   1.708µ ± 0%  -74.01% (p=0.000 n=20)
MemclrUnaligned/4_5        6.003n ± 0%   4.582n ± 0%  -23.67% (p=0.000 n=20)
MemclrUnaligned/4_16      10.410n ± 0%   5.069n ± 0%  -51.30% (p=0.000 n=20)
MemclrUnaligned/4_64      15.610n ± 0%   7.372n ± 0%  -52.77% (p=0.000 n=20)
MemclrUnaligned/4_256      35.22n ± 0%   12.95n ± 0%  -63.23% (p=0.000 n=20)
MemclrUnaligned/4_4096     419.4n ± 0%   114.6n ± 0%  -72.68% (p=0.000 n=20)
MemclrUnaligned/4_65536    6.571µ ± 0%   1.708µ ± 0%  -74.01% (p=0.000 n=20)
MemclrUnaligned/7_5        6.003n ± 0%   4.581n ± 0%  -23.69% (p=0.000 n=20)
MemclrUnaligned/7_16       8.855n ± 0%   5.079n ± 0%  -42.65% (p=0.000 n=20)
MemclrUnaligned/7_64      14.010n ± 0%   7.370n ± 0%  -47.39% (p=0.000 n=20)
MemclrUnaligned/7_256      33.62n ± 0%   12.95n ± 0%  -61.48% (p=0.000 n=20)
MemclrUnaligned/7_4096     417.8n ± 0%   114.7n ± 0%  -72.56% (p=0.000 n=20)
MemclrUnaligned/7_65536    6.570µ ± 0%   1.708µ ± 0%  -74.00% (p=0.000 n=20)
MemclrUnaligned/0_1M      104.96µ ± 0%   52.51µ ± 0%  -49.97% (p=0.000 n=20)
MemclrUnaligned/0_4M       419.8µ ± 0%   209.9µ ± 0%  -49.99% (p=0.000 n=20)
MemclrUnaligned/0_8M       839.5µ ± 0%   419.8µ ± 0%  -49.99% (p=0.000 n=20)
MemclrUnaligned/0_16M     1687.9µ ± 0%   844.9µ ± 0%  -49.94% (p=0.000 n=20)
MemclrUnaligned/0_64M      6.725m ± 0%   3.382m ± 0%  -49.72% (p=0.000 n=20)
MemclrUnaligned/1_1M      104.97µ ± 0%   52.51µ ± 0%  -49.97% (p=0.000 n=20)
MemclrUnaligned/1_4M       419.8µ ± 0%   210.0µ ± 0%  -49.97% (p=0.000 n=20)
MemclrUnaligned/1_8M       839.5µ ± 0%   419.8µ ± 0%  -50.00% (p=0.000 n=20)
MemclrUnaligned/1_16M     1687.6µ ± 0%   844.2µ ± 0%  -49.97% (p=0.000 n=20)
MemclrUnaligned/1_64M      6.724m ± 0%   3.367m ± 0%  -49.93% (p=0.000 n=20)
MemclrUnaligned/4_1M      104.97µ ± 0%   52.51µ ± 0%  -49.97% (p=0.000 n=20)
MemclrUnaligned/4_4M       419.8µ ± 0%   210.0µ ± 0%  -49.97% (p=0.000 n=20)
MemclrUnaligned/4_8M       839.5µ ± 0%   419.8µ ± 0%  -50.00% (p=0.000 n=20)
MemclrUnaligned/4_16M     1687.5µ ± 0%   844.4µ ± 0%  -49.96% (p=0.000 n=20)
MemclrUnaligned/4_64M      6.725m ± 0%   3.366m ± 0%  -49.95% (p=0.000 n=20)
MemclrUnaligned/7_1M      104.97µ ± 0%   52.51µ ± 0%  -49.97% (p=0.000 n=20)
MemclrUnaligned/7_4M       419.8µ ± 0%   210.0µ ± 0%  -49.97% (p=0.000 n=20)
MemclrUnaligned/7_8M       839.5µ ± 0%   419.8µ ± 0%  -50.00% (p=0.000 n=20)
MemclrUnaligned/7_16M     1687.9µ ± 0%   844.3µ ± 0%  -49.98% (p=0.000 n=20)
MemclrUnaligned/7_64M      6.724m ± 0%   3.362m ± 0%  -50.00% (p=0.000 n=20)
geomean                    4.659µ        2.114µ       -54.61%

Change-Id: If0174e4cd8be5e17ad146698508a966158fe83e8
Reviewed-on: https://go-review.googlesource.com/c/go/+/589539
Reviewed-by: Qiqi Huang <huangqiqi@loongson.cn>
Reviewed-by: Michael Pratt <mpratt@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: abner chenc <chenguoqi@loongson.cn>
Reviewed-by: Meidan Li <limeidan@loongson.cn>
2024-09-05 00:41:13 +00:00
Xiaolin Zhao
1bc795633e crypto/sha256: implement sha256block in hardware on loong64
goos: linux
goarch: loong64
pkg: crypto/sha256
cpu: Loongson-3A6000 @ 2500.00MHz
                  │  bench.old  │              bench.new              │
                  │   sec/op    │   sec/op     vs base                │
Hash8Bytes/New      674.0n ± 0%   433.3n ± 0%  -35.71% (p=0.000 n=20)
Hash8Bytes/Sum224   695.3n ± 0%   445.8n ± 0%  -35.88% (p=0.000 n=20)
Hash8Bytes/Sum256   676.0n ± 0%   475.5n ± 0%  -29.66% (p=0.000 n=20)
Hash1K/New          8.982µ ± 0%   4.963µ ± 0%  -44.74% (p=0.000 n=20)
Hash1K/Sum224       8.999µ ± 0%   4.975µ ± 0%  -44.72% (p=0.000 n=20)
Hash1K/Sum256       8.979µ ± 0%   5.005µ ± 0%  -44.26% (p=0.000 n=20)
Hash8K/New          66.97µ ± 0%   36.78µ ± 0%  -45.08% (p=0.000 n=20)
Hash8K/Sum224       66.99µ ± 0%   36.79µ ± 0%  -45.09% (p=0.000 n=20)
Hash8K/Sum256       66.97µ ± 0%   36.82µ ± 0%  -45.03% (p=0.000 n=20)
geomean             7.431µ        4.357µ       -41.37%

                  │  bench.old   │              bench.new               │
                  │     B/s      │     B/s       vs base                │
Hash8Bytes/New      11.32Mi ± 0%   17.60Mi ± 0%  +55.52% (p=0.000 n=20)
Hash8Bytes/Sum224   10.98Mi ± 0%   17.11Mi ± 0%  +55.86% (p=0.000 n=20)
Hash8Bytes/Sum256   11.28Mi ± 0%   16.04Mi ± 0%  +42.18% (p=0.000 n=20)
Hash1K/New          108.7Mi ± 0%   196.8Mi ± 0%  +80.98% (p=0.000 n=20)
Hash1K/Sum224       108.5Mi ± 0%   196.3Mi ± 0%  +80.89% (p=0.000 n=20)
Hash1K/Sum256       108.8Mi ± 0%   195.1Mi ± 0%  +79.42% (p=0.000 n=20)
Hash8K/New          116.7Mi ± 0%   212.4Mi ± 0%  +82.09% (p=0.000 n=20)
Hash8K/Sum224       116.6Mi ± 0%   212.4Mi ± 0%  +82.09% (p=0.000 n=20)
Hash8K/Sum256       116.7Mi ± 0%   212.2Mi ± 0%  +81.90% (p=0.000 n=20)
geomean             52.15Mi        88.95Mi       +70.55%

goos: linux
goarch: loong64
pkg: crypto/sha256
cpu: Loongson-3A5000 @ 2500.00MHz
                  │  bench.old   │              bench.new              │
                  │    sec/op    │   sec/op     vs base                │
Hash8Bytes/New       855.9n ± 0%   521.1n ± 0%  -39.12% (p=0.000 n=20)
Hash8Bytes/Sum224    875.2n ± 0%   532.7n ± 1%  -39.13% (p=0.000 n=20)
Hash8Bytes/Sum256    909.4n ± 0%   553.9n ± 0%  -39.09% (p=0.000 n=20)
Hash1K/New          11.304µ ± 0%   5.684µ ± 0%  -49.72% (p=0.000 n=20)
Hash1K/Sum224       11.323µ ± 0%   5.690µ ± 0%  -49.75% (p=0.000 n=20)
Hash1K/Sum256       11.341µ ± 0%   5.714µ ± 0%  -49.62% (p=0.000 n=20)
Hash8K/New           84.26µ ± 0%   41.97µ ± 0%  -50.19% (p=0.000 n=20)
Hash8K/Sum224        84.27µ ± 0%   41.99µ ± 0%  -50.18% (p=0.000 n=20)
Hash8K/Sum256        84.32µ ± 0%   42.01µ ± 0%  -50.18% (p=0.000 n=20)
geomean              9.434µ        5.041µ       -46.56%

                  │  bench.old   │               bench.new                │
                  │     B/s      │      B/s       vs base                 │
Hash8Bytes/New      8.917Mi ± 0%   14.639Mi ± 0%   +64.17% (p=0.000 n=20)
Hash8Bytes/Sum224   8.717Mi ± 0%   14.319Mi ± 1%   +64.28% (p=0.000 n=20)
Hash8Bytes/Sum256   8.392Mi ± 0%   13.771Mi ± 0%   +64.09% (p=0.000 n=20)
Hash1K/New          86.39Mi ± 0%   171.81Mi ± 0%   +98.89% (p=0.000 n=20)
Hash1K/Sum224       86.25Mi ± 0%   171.64Mi ± 0%   +99.00% (p=0.000 n=20)
Hash1K/Sum256       86.11Mi ± 0%   170.92Mi ± 0%   +98.49% (p=0.000 n=20)
Hash8K/New          92.72Mi ± 0%   186.13Mi ± 0%  +100.75% (p=0.000 n=20)
Hash8K/Sum224       92.71Mi ± 0%   186.07Mi ± 0%  +100.71% (p=0.000 n=20)
Hash8K/Sum256       92.65Mi ± 0%   185.99Mi ± 0%  +100.74% (p=0.000 n=20)
geomean             41.08Mi         76.87Mi        +87.12%

Change-Id: Ib41d19d136b8593339af94a822942c102238891b
Reviewed-on: https://go-review.googlesource.com/c/go/+/590155
Reviewed-by: Meidan Li <limeidan@loongson.cn>
Reviewed-by: abner chenc <chenguoqi@loongson.cn>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Qiqi Huang <huangqiqi@loongson.cn>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
2024-09-05 00:41:00 +00:00
Xiaolin Zhao
137e91832a crypto/sha512: implement sha512block in hardware on loong64
goos: linux
goarch: loong64
pkg: crypto/sha512
cpu: Loongson-3A6000 @ 2500.00MHz
                  │  bench.old   │              bench.new              │
                  │    sec/op    │   sec/op     vs base                │
Hash8Bytes/New       942.3n ± 0%   627.5n ± 0%  -33.41% (p=0.000 n=20)
Hash8Bytes/Sum384    951.3n ± 0%   661.1n ± 0%  -30.51% (p=0.000 n=20)
Hash8Bytes/Sum512   1032.0n ± 0%   631.9n ± 0%  -38.77% (p=0.000 n=20)
Hash1K/New           6.355µ ± 0%   3.285µ ± 0%  -48.31% (p=0.000 n=20)
Hash1K/Sum384        6.333µ ± 0%   3.320µ ± 0%  -47.58% (p=0.000 n=20)
Hash1K/Sum512        6.416µ ± 0%   3.293µ ± 0%  -48.68% (p=0.000 n=20)
Hash8K/New           43.91µ ± 0%   22.01µ ± 0%  -49.89% (p=0.000 n=20)
Hash8K/Sum384        43.77µ ± 0%   22.05µ ± 0%  -49.61% (p=0.000 n=20)
Hash8K/Sum512        43.87µ ± 0%   22.01µ ± 0%  -49.83% (p=0.000 n=20)
geomean              6.480µ        3.596µ       -44.50%

                  │  bench.old   │               bench.new               │
                  │     B/s      │      B/s       vs base                │
Hash8Bytes/New      8.097Mi ± 0%   12.159Mi ± 0%  +50.18% (p=0.000 n=20)
Hash8Bytes/Sum384   8.020Mi ± 0%   11.539Mi ± 0%  +43.88% (p=0.000 n=20)
Hash8Bytes/Sum512   7.391Mi ± 0%   12.074Mi ± 0%  +63.35% (p=0.000 n=20)
Hash1K/New          153.7Mi ± 0%    297.2Mi ± 0%  +93.43% (p=0.000 n=20)
Hash1K/Sum384       154.2Mi ± 0%    294.1Mi ± 0%  +90.74% (p=0.000 n=20)
Hash1K/Sum512       152.2Mi ± 0%    296.6Mi ± 0%  +94.84% (p=0.000 n=20)
Hash8K/New          177.9Mi ± 0%    355.0Mi ± 0%  +99.55% (p=0.000 n=20)
Hash8K/Sum384       178.5Mi ± 0%    354.3Mi ± 0%  +98.46% (p=0.000 n=20)
Hash8K/Sum512       178.1Mi ± 0%    355.0Mi ± 0%  +99.34% (p=0.000 n=20)
geomean             59.81Mi         107.8Mi       +80.19%

goos: linux
goarch: loong64
pkg: crypto/sha512
cpu: Loongson-3A5000 @ 2500.00MHz
                  │  bench.old   │              bench.new              │
                  │    sec/op    │   sec/op     vs base                │
Hash8Bytes/New      1224.0n ± 0%   730.1n ± 0%  -40.36% (p=0.000 n=20)
Hash8Bytes/Sum384   1265.0n ± 0%   763.2n ± 0%  -39.67% (p=0.000 n=20)
Hash8Bytes/Sum512   1257.0n ± 0%   753.5n ± 0%  -40.06% (p=0.000 n=20)
Hash1K/New           8.218µ ± 0%   3.785µ ± 0%  -53.94% (p=0.000 n=20)
Hash1K/Sum384        8.248µ ± 0%   3.813µ ± 0%  -53.77% (p=0.000 n=20)
Hash1K/Sum512        8.235µ ± 0%   3.807µ ± 0%  -53.77% (p=0.000 n=20)
Hash8K/New           56.83µ ± 0%   25.37µ ± 0%  -55.35% (p=0.000 n=20)
Hash8K/Sum384        56.85µ ± 0%   25.39µ ± 0%  -55.34% (p=0.000 n=20)
Hash8K/Sum512        56.84µ ± 0%   25.38µ ± 0%  -55.36% (p=0.000 n=20)
geomean              8.360µ        4.165µ       -50.18%

                  │  bench.old   │               bench.new                │
                  │     B/s      │      B/s       vs base                 │
Hash8Bytes/New      6.232Mi ± 0%   10.452Mi ± 0%   +67.71% (p=0.000 n=20)
Hash8Bytes/Sum384   6.027Mi ± 0%    9.995Mi ± 0%   +65.82% (p=0.000 n=20)
Hash8Bytes/Sum512   6.065Mi ± 0%   10.123Mi ± 0%   +66.90% (p=0.000 n=20)
Hash1K/New          118.8Mi ± 0%    258.0Mi ± 0%  +117.12% (p=0.000 n=20)
Hash1K/Sum384       118.4Mi ± 0%    256.1Mi ± 0%  +116.29% (p=0.000 n=20)
Hash1K/Sum512       118.6Mi ± 0%    256.5Mi ± 0%  +116.30% (p=0.000 n=20)
Hash8K/New          137.5Mi ± 0%    307.9Mi ± 0%  +123.96% (p=0.000 n=20)
Hash8K/Sum384       137.4Mi ± 0%    307.8Mi ± 0%  +123.93% (p=0.000 n=20)
Hash8K/Sum512       137.4Mi ± 0%    307.9Mi ± 0%  +124.01% (p=0.000 n=20)
geomean             46.35Mi         93.05Mi       +100.76%

Change-Id: I0d764df16872598b8d2fd92d7253cf3fdbfdfdf2
Reviewed-on: https://go-review.googlesource.com/c/go/+/590156
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: abner chenc <chenguoqi@loongson.cn>
Reviewed-by: Meidan Li <limeidan@loongson.cn>
Reviewed-by: Qiqi Huang <huangqiqi@loongson.cn>
2024-09-05 00:40:39 +00:00
Ian Lance Taylor
d102c2ed01 os: don't fail TestGetwdDeepWithPWDSet if TMPDIR has a symlink
When testing with PWD set, it's possible for the stat of PWD to fail
with ENAMETOOLONG, and for syscall.Getwd to fail for the same reason.
If PWD contains symlinks, the fallback code won't know about them.
If Getwd returns the same result as PWD with resolved symlinks,
the test should not fail.

Change-Id: I39587ddb826d4e18339e185aad0cdd60167b1079
Reviewed-on: https://go-review.googlesource.com/c/go/+/610759
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>
Commit-Queue: Ian Lance Taylor <iant@google.com>
Reviewed-by: Tim King <taking@google.com>
2024-09-04 23:26:25 +00:00
Kir Kolyshkin
a2891aef75 cmd/link/internal/ld: rm os.Getwd from elf_test
When specifying the package to build, a relative path is sufficient.

Change-Id: I1ae08065b5cd77ec25be42dc1e664720a07baa62
Reviewed-on: https://go-review.googlesource.com/c/go/+/610039
Auto-Submit: Ian Lance Taylor <iant@golang.org>
Commit-Queue: Ian Lance Taylor <iant@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
2024-09-04 20:44:48 +00:00
Kir Kolyshkin
9ba8ac253e cmd/link/internal/ld: rm os.Getwd from dwarf_test
Calls to os.Getwd were needed to set the cwd for go build to an absolute
path. Since CL 401340 os/exec takes care of setting PWD to a
filepath.Abs(cmd.Dir), so it looks like an absolute path is not
really required.

Change-Id: Ib3abffc9087a3329d8f40f81eb65f1b2c1a03a9c
Reviewed-on: https://go-review.googlesource.com/c/go/+/610038
Auto-Submit: Ian Lance Taylor <iant@golang.org>
Commit-Queue: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
2024-09-04 20:44:15 +00:00
Alan Donovan
de0aafa3c7 vendor/golang.org/x/tools: update to v0.24.1-0.20240904143311-70f56264139c
Among other things, this should fix a regression in printf
whereby materialized aliases caused "any" and "interface{}"
in printf signatures not to be recognized as identical.

It also updates ureader.go used by vendored x/tools during
some tests, including cmd/internal/moddeps.TestAllDependencies.
This test uses golang.org/x/tools/cmd/bundle which uses x/reader.

Fixes #68796

Change-Id: I9f0711e66a5c4daaffe695c515aea3b8fb3d01e0
Reviewed-on: https://go-review.googlesource.com/c/go/+/610736
Reviewed-by: Robert Findley <rfindley@google.com>
Auto-Submit: Alan Donovan <adonovan@google.com>
Reviewed-by: Tim King <taking@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
2024-09-04 20:42:05 +00:00
Garrett Bodley
004686b6e5 crypto/internal/nistec: Avo port of p256_asm_amd64.s
This implementation utilizes the same registers found in the reference
implementation, aiming to produce a minimal semantic diff between the
Avo-generated output and the original hand-written assembly.

To verify the Avo implementation, the reference and Avo-generated
assembly files are fed to `go tool asm`, capturing the debug output into
corresponding temp files. The debug output contains supplementary
metadata (line numbers, instruction offsets, and source file references)
that must be removed in order to obtain a semantic diff of the two
files. This is accomplished via a small utility script written in awk.

The reference assembly file does not specify a frame size for a number
of the defined assembly functions. Avo automatically infers the frame
size when generating the TEXT directive, leading to a diff on those
lines.

Commands used to verify Avo output:

GOROOT=$(go env GOROOT)
ASM_PATH="src/crypto/internal/nistec/p256_asm_amd64.s"
REFERENCE="54fe0fd43fcf8609666c16ae6d15ed92873b1564"

go tool asm -o /dev/null -I "$GOROOT"/src/runtime -debug \
  <(git cat-file -p "$REFERENCE:$ASM_PATH") \
  > /tmp/reference.s

go tool asm -o /dev/null -I "$GOROOT"/src/runtime -debug \
  "$ASM_PATH" \
  > /tmp/avo.s

normalize(){
  awk '{
    $1=$2=$3="";
    print substr($0,4)
  }'
}

diff <(normalize < /tmp/reference.s) <(normalize < /tmp/avo.s)

1c1
< TEXT <unlinkable>.p256OrdLittleToBig(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.p256OrdLittleToBig(SB), NOSPLIT, $0-16
3c3
< TEXT <unlinkable>.p256OrdBigToLittle(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.p256OrdBigToLittle(SB), NOSPLIT, $0-16
5c5
< TEXT <unlinkable>.p256LittleToBig(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.p256LittleToBig(SB), NOSPLIT, $0-16
7c7
< TEXT <unlinkable>.p256BigToLittle(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.p256BigToLittle(SB), NOSPLIT, $0-16
23c23
< TEXT <unlinkable>.p256MovCond(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.p256MovCond(SB), NOSPLIT, $0-32
74c74
< TEXT <unlinkable>.p256NegCond(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.p256NegCond(SB), NOSPLIT, $0-16
99c99
< TEXT <unlinkable>.p256Sqr(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.p256Sqr(SB), NOSPLIT, $0-24
234c234
< TEXT <unlinkable>.p256Mul(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.p256Mul(SB), NOSPLIT, $0-24
401c401
< TEXT <unlinkable>.p256FromMont(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.p256FromMont(SB), NOSPLIT, $0-16
465c465
< TEXT <unlinkable>.p256Select(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.p256Select(SB), NOSPLIT, $0-24
513c513
< TEXT <unlinkable>.p256SelectAffine(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.p256SelectAffine(SB), NOSPLIT, $0-24
566c566
< TEXT <unlinkable>.p256OrdMul(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.p256OrdMul(SB), NOSPLIT, $0-24
806c806
< TEXT <unlinkable>.p256OrdSqr(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.p256OrdSqr(SB), NOSPLIT, $0-24

Change-Id: I610b097c573b9d9018f0e26bc2afde5edb3f954b
Reviewed-on: https://go-review.googlesource.com/c/go/+/599875
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
2024-09-04 20:37:00 +00:00
Garrett Bodley
d61b73c1d1 crypto/aes: Avo port of asm_amd64.s
This implementation utilizes the same registers found in the reference
implementation, aiming to produce a minimal semantic diff between the
Avo-generated output and the original hand-written assembly.

To verify the Avo implementation, the reference and Avo-generated
assembly files are fed to `go tool asm`, capturing the debug output into
corresponding temp files. The debug output contains supplementary
metadata (line numbers, instruction offsets, and source file references)
that must be removed in order to obtain a semantic diff of the two
files. This is accomplished via a small utility script written in awk.

The reference assembly file does not specify a frame size for some of
the defined assembly functions. Avo automatically infers the frame size
when generating TEXT directives, leading to a diff on those lines.

Commands used to verify Avo output:

GOROOT=$(go env GOROOT)
ASM_PATH="src/crypto/aes/asm_amd64.s"
REFERENCE="54fe0fd43fcf8609666c16ae6d15ed92873b1564"

go tool asm -o /dev/null -I "$GOROOT"/src/runtime -debug \
  <(git cat-file -p "$REFERENCE:$ASM_PATH") \
  > /tmp/reference.s

go tool asm -o /dev/null -I "$GOROOT"/src/runtime -debug \
  "$ASM_PATH" \
  > /tmp/avo.s

normalize(){
  awk '{
    $1=$2=$3="";
    print substr($0,4)
  }'
}

diff <(normalize < /tmp/reference.s) <(normalize < /tmp/avo.s)

1c1
< TEXT <unlinkable>.encryptBlockAsm(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.encryptBlockAsm(SB), NOSPLIT, $0-32
45c45
< TEXT <unlinkable>.decryptBlockAsm(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.decryptBlockAsm(SB), NOSPLIT, $0-32
89c89
< TEXT <unlinkable>.expandKeyAsm(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.expandKeyAsm(SB), NOSPLIT, $0-32

Change-Id: If647584df4137146d355f91ac0f6a8285d07c932
Reviewed-on: https://go-review.googlesource.com/c/go/+/600375
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
2024-09-04 20:36:51 +00:00
Garrett Bodley
6ee8c07c3c crypto/aes: Avo port of gcm_amd64.s
This implementation utilizes the same registers found in the reference
implementation, aiming to produce a minimal semantic diff between the
Avo-generated output and the original hand-written assembly.

To verify the Avo implementation, the reference and Avo-generated
assembly files are fed to `go tool asm`, capturing the debug output into
corresponding temp files. The debug output contains supplementary
metadata (line numbers, instruction offsets, and source file references)
that must be removed in order to obtain a semantic diff of the two
files. This is accomplished via a small utility script written in awk.

The reference assembly file does not specify a frame size for some of
the defined assembly functions. Avo automatically infers the frame size
when generating TEXT directives, leading to a diff on those lines. Some
metadata not included in the reference assembly has also been added,
which leads to a diff in the lines where that parameter symbol is
referenced.

Commands used to verify Avo output:

GOROOT=$(go env GOROOT)
ASM_PATH="src/crypto/aes/gcm_amd64.s"
REFERENCE="54fe0fd43fcf8609666c16ae6d15ed92873b1564"

go tool asm -o /dev/null -I "$GOROOT"/src/runtime -debug \
  <(git cat-file -p "$REFERENCE:$ASM_PATH") \
  > /tmp/reference.s

go tool asm -o /dev/null -I "$GOROOT"/src/runtime -debug \
  "$ASM_PATH" \
  > /tmp/avo.s

normalize(){
  awk '{
    $1=$2=$3="";
    print substr($0,4)
  }'
}

diff <(normalize < /tmp/reference.s) <(normalize < /tmp/avo.s)

1c1
< TEXT <unlinkable>.gcmAesFinish(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.gcmAesFinish(SB), NOSPLIT, $0-40
44c44
< TEXT <unlinkable>.gcmAesInit(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.gcmAesInit(SB), NOSPLIT, $0-32
131c131
< TEXT <unlinkable>.gcmAesData(SB), NOSPLIT, $0
---
> TEXT <unlinkable>.gcmAesData(SB), NOSPLIT, $0-40
325c325
< MOVQ dst+8(FP), DX
---
> MOVQ dst_base+8(FP), DX
1207c1207
< MOVQ dst+8(FP), SI
---
> MOVQ dst_base+8(FP), SI

Change-Id: Iad8f8c6ea5d50ac093c8535adc9d23fbf2612fc2
Reviewed-on: https://go-review.googlesource.com/c/go/+/601462
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
2024-09-04 20:36:42 +00:00
Garrett Bodley
1e0cba8ee1 crypto/md5: Avo port of md5block_amd64.s
This implementation utilizes the same registers found in the reference
implementation, aiming to produce a minimal semantic diff between the
Avo-generated output and the original hand-written assembly.

To verify the Avo implementation, the reference and Avo-generated
assembly files are fed to `go tool asm`, capturing the debug output into
corresponding temp files. The debug output contains supplementary
metadata (line numbers, instruction offsets, and source file references)
that must be removed in order to obtain a semantic diff of the two
files. This is accomplished via a small utility script written in awk.

Metadata not found in the reference assembly file has been added to one
parameter symbol, resulting in a single line diff.

Commands used to verify Avo output:

GOROOT=$(go env GOROOT)
ASM_PATH="src/crypto/md5/md5block_amd64.s"
REFERENCE="54fe0fd43fcf8609666c16ae6d15ed92873b1564"

go tool asm -o /dev/null -I "$GOROOT"/src/runtime -debug \
  <(git cat-file -p "$REFERENCE:$ASM_PATH") \
  > /tmp/reference.s

go tool asm -o /dev/null -I "$GOROOT"/src/runtime -debug \
  "$ASM_PATH" \
  > /tmp/avo.s

normalize(){
  awk '{
    $1=$2=$3="";
    print substr($0,4)
  }'
}

diff <(normalize < /tmp/reference.s) <(normalize < /tmp/avo.s)

3c3
< MOVQ p+8(FP), SI
---
> MOVQ p_base+8(FP), SI

Change-Id: Ifecc84fd0f5a39a88350e6eaffb45ed3fdacf2fb
Reviewed-on: https://go-review.googlesource.com/c/go/+/599935
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
2024-09-04 20:36:18 +00:00
Garrett Bodley
3b78ca0626 crypto/sha512: Avo port of sha512block_amd64.s
This implementation utilizes the same registers found in the reference
implementation, aiming to produce a minimal semantic diff between the
Avo-generated output and the original hand-written assembly.

To verify the Avo implementation, the reference and Avo-generated
assembly files are fed to `go tool asm`, capturing the debug output into
corresponding temp files. The debug output contains supplementary
metadata (line numbers, instruction offsets, and source file references)
that must be removed in order to obtain a semantic diff of the two
files. This is accomplished via a small utility script written in awk.

Commands used to verify Avo output:

GOROOT=$(go env GOROOT)
ASM_PATH="src/crypto/sha512/sha512block_amd64.s"
REFERENCE="54fe0fd43fcf8609666c16ae6d15ed92873b1564"

go tool asm -o /dev/null -I "$GOROOT"/src/runtime -debug \
  <(git cat-file -p "$REFERENCE:$ASM_PATH") \
  > /tmp/reference.s

go tool asm -o /dev/null -I "$GOROOT"/src/runtime -debug \
  "$ASM_PATH" \
  > /tmp/avo.s

normalize(){
  awk '{
    $1=$2=$3="";
    print substr($0,4)
  }'
}

diff <(normalize < /tmp/reference.s) <(normalize < /tmp/avo.s)

Change-Id: I172f0cb97252635c657efe82d1b547e6b6f40ebb
Reviewed-on: https://go-review.googlesource.com/c/go/+/598958
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
2024-09-04 20:36:05 +00:00
Garrett Bodley
f9eb06c302 crypto/sha1: Avo port of sha1block_amd64.s
This implementation utilizes the same registers found in the reference
implementation, aiming to produce a minimal semantic diff between the
Avo-generated output and the original hand-written assembly.

To verify the Avo implementation, the reference and Avo-generated
assembly files are fed to `go tool asm`, capturing the debug output into
corresponding temp files. The debug output contains supplementary
metadata (line numbers, instruction offsets, and source file references)
that must be removed in order to obtain a semantic diff of the two
files. This is accomplished via a small utility script written in awk.

Commands used to verify Avo output:

GOROOT=$(go env GOROOT)
ASM_PATH="src/crypto/sha1/sha1block_amd64.s"
REFERENCE="54fe0fd43fcf8609666c16ae6d15ed92873b1564"

go tool asm -o /dev/null -I "$GOROOT"/src/runtime -debug \
  <(git cat-file -p "$REFERENCE:$ASM_PATH") \
  > /tmp/reference.s

go tool asm -o /dev/null -I "$GOROOT"/src/runtime -debug \
  "$ASM_PATH" \
  > /tmp/avo.s

normalize(){
  awk '{
    $1=$2=$3="";
    print substr($0,4)
  }'
}

diff <(normalize < /tmp/reference.s) <(normalize < /tmp/avo.s)

1273c1273
< MOVQ $K_XMM_AR<>(SB), R8
---
> LEAQ K_XMM_AR<>(SB), R8

Change-Id: I39168fadb01baa9a96bc2b432fc94b492d036ce4
Reviewed-on: https://go-review.googlesource.com/c/go/+/598795
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Roland Shoemaker <roland@golang.org>
2024-09-04 20:35:55 +00:00
Garrett Bodley
dbd50a16dc crypto/sha256: Avo port of sha256block_amd64.s
This implementation utilizes the same registers found in the reference
implementation, aiming to produce a minimal semantic diff between the
Avo-generated output and the original hand-written assembly.

To verify the Avo implementation, the reference and Avo-generated
assembly files are fed to `go tool asm`, capturing the debug output into
corresponding temp files. The debug output contains supplementary
metadata (line numbers, instruction offsets, and source file references)
that must be removed in order to obtain a semantic diff of the two
files. This is accomplished via a small utility script written in awk.

Commands used to verify Avo output:

GOROOT=$(go env GOROOT)
ASM_PATH="src/crypto/sha256/sha256block_amd64.s"
REFERENCE="54fe0fd43fcf8609666c16ae6d15ed92873b1564"

go tool asm -o /dev/null -I "$GOROOT"/src/runtime -debug \
  <(git cat-file -p "$REFERENCE:$ASM_PATH") \
  > /tmp/reference.s

go tool asm -o /dev/null -I $GOROOT/src/runtime -debug \
  "$ASM_PATH" \
  > /tmp/avo.s

normalize(){
  awk '{
    $1=$2=$3="";
    print substr($0,4)
  }'
}

diff <(normalize < /tmp/reference.s) <(normalize < /tmp/avo.s)

3513c3513
< MOVQ $K256<>(SB), BP
---
> LEAQ K256<>(SB), BP
4572c4572
< MOVQ $K256<>(SB), BP
---
> LEAQ K256<>(SB), BP

Change-Id: I637c01d746ca775b8a09f874f7925ffc3b4965ad
Reviewed-on: https://go-review.googlesource.com/c/go/+/595559
Reviewed-by: Russell Webb <russell.webb@protonmail.com>
Reviewed-by: Roland Shoemaker <roland@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
2024-09-04 20:35:45 +00:00
Cuong Manh Le
d6d5d8c8fb cmd/compile: remove NameSet.Sorted
The only usage of it was removed in CL 517617

Change-Id: If1898b7cde4f8c7ab906a2c27a01125415b463c3
Reviewed-on: https://go-review.googlesource.com/c/go/+/610600
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
2024-09-04 20:29:15 +00:00
Cuong Manh Le
e7047bd14d cmd/compile/internal/liveness: use slices.Reverse
Now that we're bootstrapping from a toolchain that has the slices
package.

Updates #64751

Change-Id: Id50d76de05e353ef06d64b47ad6400b2b7572205
Reviewed-on: https://go-review.googlesource.com/c/go/+/610775
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
2024-09-04 20:28:38 +00:00
Roland Shoemaker
a5708231ac crypto/internal/boring: disable LFS64 interfaces
Comment out the definition in the libcrypto I/O code which enables
the LFS64 interfaces. We don't use any of the I/O bits and pieces, and
it's outside of the FIPS module, and it fixes some breakage in certain
scenarios.

Change-Id: Ie6597813726f94e23780b77d907cc1b9ccef36f0
Reviewed-on: https://go-review.googlesource.com/c/go/+/609976
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Russ Cox <rsc@golang.org>
2024-09-04 19:22:52 +00:00
Kir Kolyshkin
dfae83c9db cmd/link/internal/ld: simplify gobuildTestdata
Drop the second argument, which is is always a one-time temporary
directory, thus it can be created right here.

Change-Id: I73e5be2ccd4bddec249c7cb2a8ff9242d99f7e20
Reviewed-on: https://go-review.googlesource.com/c/go/+/610037
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Ian Lance Taylor <iant@golang.org>
Commit-Queue: Ian Lance Taylor <iant@golang.org>
2024-09-04 18:50:48 +00:00
Mateusz Poliwczak
fbe74d8ddb go/parser: convert *ast.CallExpr into *ast.ParenExpr in extractName
We are loosing a bit of the AST information, i believe we should
convert *ast.CallExpr into *ast.ParenExpr.

See https://github.com/golang/go/issues/69206#issuecomment-2324592744

Change-Id: I2d9ad8a3dead664a4fa9ac324e8d8a955a4d97c8
GitHub-Last-Rev: e5db56d5ca
GitHub-Pull-Request: golang/go#69209
Reviewed-on: https://go-review.googlesource.com/c/go/+/610078
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Robert Griesemer <gri@google.com>
Auto-Submit: Robert Griesemer <gri@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
2024-09-04 18:29:14 +00:00
Cuong Manh Le
c541aabfc3 cmd: remove notsha256 package
All of its usages were removed in CL 610596.

Updates #51940
Updates #64751

Change-Id: I72a0ea4bd44a2f671e032bffa1facf84822a90f2
Reviewed-on: https://go-review.googlesource.com/c/go/+/610599
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
2024-09-04 18:27:49 +00:00
Cuong Manh Le
5ca785d48d cmd: use 16 bytes hash when possible
CL 402595 changes all usages of 16 bytes hash to 32 bytes hash by using
notsha256.

However, since CL 454836, notsha256 is not necessary anymore, so this CL
reverts those changes to 16 bytes hash using cmd/internal/hash package.

Updates #51940
Updates #64751

Change-Id: Ic015468ca4a49d0c3b1fb9fdbed93fddef3c838f
Reviewed-on: https://go-review.googlesource.com/c/go/+/610598
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
2024-09-04 18:27:07 +00:00
Cuong Manh Le
f033bc1819 cmd: use 20 bytes hash when possible
CL 402595 changes all usages of 20 bytes hash to 32 bytes hash by using
notsha256.

However, since CL 454836, notsha256 is not necessary anymore, so this CL
reverts those changes to 20 bytes hash using cmd/internal/hash package.

Updates #51940
Updates #64751

Change-Id: Icb08d5a0d8032a3c4d050ff7b2298d31c483b88b
Reviewed-on: https://go-review.googlesource.com/c/go/+/610597
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Keith Randall <khr@golang.org>
2024-09-04 18:25:21 +00:00
Cuong Manh Le
4fd73e5d4c cmd: do not use notsha256
CL 402595 used notsha256 to prevent the compiler from depending on
cgo-based implementations of sha1 and sha256.

However, since CL 454836, cmd is built with CGO_ENABLED=0, which
will disable boringcrypto. Thus all usages of notsha256 is not necessary
anymore.

Updates #51940
Updates #64751

Change-Id: I503090f7a2efb5723e8a79523b143dc7cdb4edd0
Reviewed-on: https://go-review.googlesource.com/c/go/+/610596
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
2024-09-04 18:23:49 +00:00
Cuong Manh Le
ad8b5f7fe9 cmd/internal: add hash package
To be used in compiler toolchain instead of notsha256.

Change-Id: Iceeacb6df7dfa7111ec98f070eba8e27a4ddbe8b
Reviewed-on: https://go-review.googlesource.com/c/go/+/610595
Reviewed-by: Keith Randall <khr@golang.org>
Auto-Submit: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Reviewed-by: Keith Randall <khr@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
2024-09-04 18:22:26 +00:00
Alan Donovan
cd9a300afc all: fix printf(var) mistakes detected by latest printf checker
These will cause build failures once we vendor x/tools.

In once case I renamed a function err to errf to indicate
that it is printf-like.

Updates golang/go#68796

Change-Id: I04d57b34ee5362f530554b7e8b817f70a9088d12
Reviewed-on: https://go-review.googlesource.com/c/go/+/610739
Commit-Queue: Alan Donovan <adonovan@google.com>
Reviewed-by: Robert Findley <rfindley@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Tim King <taking@google.com>
Auto-Submit: Alan Donovan <adonovan@google.com>
2024-09-04 18:16:59 +00:00
Michael Anthony Knyszek
21ac23a96f unique: don't retain uncloned input as key
Currently the unique package tries to clone strings that get stored in
its internal map to avoid retaining large strings.

However, this falls over entirely due to the fact that the original
string is *still* stored in the map as a key. Whoops. Fix this by
storing the cloned value in the map instead.

This change also adds a test which fails without this change.

Change-Id: I1a6bb68ed79b869ea12ab6be061a5ae4b4377ddb
Reviewed-on: https://go-review.googlesource.com/c/go/+/610738
Reviewed-by: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
2024-09-04 18:13:29 +00:00
Michael Anthony Knyszek
79fd633632 internal/weak: shade pointer in weak-to-strong conversion
There's a bug in the weak-to-strong conversion in that creating the
*only* strong pointer to some weakly-held object during the mark phase
may result in that object not being properly marked.

The exact mechanism for this is that the new strong pointer will always
point to a white object (because it was only weakly referenced up until
this point) and it can then be stored in a blackened stack, hiding it
from the garbage collector.

This "hide a white pointer in the stack" problem is pretty much exactly
what the Yuasa part of the hybrid write barrier is trying to catch, so
we need to do the same thing the write barrier would do: shade the
pointer.

Added a test and confirmed that it fails with high probability if the
pointer shading is missing.

Fixes #69210.

Change-Id: Iaae64ae95ea7e975c2f2c3d4d1960e74e1bd1c3f
Reviewed-on: https://go-review.googlesource.com/c/go/+/610396
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Auto-Submit: Michael Knyszek <mknyszek@google.com>
2024-09-04 18:13:24 +00:00
Keith Randall
1b4cf43e42 runtime: size maps.Clone destination bucket array safely
In rare situations, like during same-sized grows, the source map for
maps.Clone may be overloaded (has more than 6.5 entries per
bucket). This causes the runtime to allocate a larger bucket array for
the destination map than for the source map. The maps.Clone code
walks off the end of the source array if it is smaller than the
destination array.

This is a pretty simple fix, ensuring that the destination bucket
array is never longer than the source bucket array. Maybe a better fix
is to make the Clone code handle shorter source arrays correctly, but
this fix is deliberately simple to reduce the risk of backporting this
fix.

Fixes #69110

Change-Id: I824c93d1db690999f25a3c43b2816fc28ace7509
Reviewed-on: https://go-review.googlesource.com/c/go/+/609757
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
2024-09-04 16:06:05 +00:00
Michael Pratt
a4eba85fad cmd/internal/obj: drop NOFRAME TODO
NOFRAME is long since implemented beyond ppc64x.

Change-Id: Ia02c732badc3330bf876723bb64eff390f3e6622
Reviewed-on: https://go-review.googlesource.com/c/go/+/610695
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Cherry Mui <cherryyz@google.com>
Auto-Submit: Michael Pratt <mpratt@google.com>
2024-09-04 15:31:27 +00:00
komuw
c55a170083 crypto/cipher: update documentation for aead Open
The remaining capacity of dst should not overlap ciphertext.
The previous wording was probably a copy paste mistake from aead Seal.

Change-Id: Iaa28073f9ea90cbe2032c0c1149a78feab6c9239
GitHub-Last-Rev: fb54bc84c4
GitHub-Pull-Request: golang/go#69108
Reviewed-on: https://go-review.googlesource.com/c/go/+/609075
Reviewed-by: Filippo Valsorda <filippo@golang.org>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-by: Michael Pratt <mpratt@google.com>
Auto-Submit: Filippo Valsorda <filippo@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
2024-09-04 14:37:02 +00:00