Skip to content

src: use SIMD for snapshot checksum #47145

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
kvakil wants to merge 1 commit into from
Closed

src: use SIMD for snapshot checksum #47145

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
src: use SIMD for snapshot checksum 
Snapshot checksum verification uses `v8::Internal::Checksum`, which uses
zlib's adler32 checksum. However it was using the slow version of
adler32 for machines with no SIMD. Let's change it to use the SIMD
version when it is available. Fortunately the necessary incantations
were already included in our _other_ copy of Chromium's zlib in
`deps/zlib/zlib.gyp`, so this is mostly just copying from there.

On a very recent Intel machine, this speeds up the snapshot checksumming
from 870us to 300us (a 570us startup saving), according to:

```bash
for i in {1..100}; do
  ./node --profile-deserialization -e 0
done \
  |& grep 'Verifying snapshot checksum' \
  | awk '{s+=$5} END{print s/NR}'
```

P.S.: maybe we should just disable this entirely via
`--no-verify-snapshot-checksum`? It doesn't feel like a super necessary
protection.
  • Loading branch information
@kvakil
kvakil committed Mar 18, 2023
commit aa2a441ce639ff821dd7ad07e3a292a2c3d9ba50
133 changes: 96 additions & 37 deletions tools/v8_gypfiles/v8.gyp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,8 @@
{
'variables': {
'V8_ROOT': '../../deps/v8',
'V8_ZLIB_ROOT': '<(V8_ROOT)/third_party/zlib',
'arm_fpu%': '',
'v8_code': 1,
'v8_random_seed%': 314159265,
'v8_vector_stores%': 0,
Expand Down Expand Up @@ -1884,6 +1886,54 @@
},
}, # postmortem-metadata

{
'target_name': 'v8_zlib_adler32_simd',
'type': 'static_library',
'conditions': [
['target_arch in "ia32 x64" and OS!="ios"', {
'defines': [ 'ADLER32_SIMD_SSSE3' ],
'conditions': [
['OS=="win"', {
'defines': [ 'X86_WINDOWS' ],
},{
'defines': [ 'X86_NOT_WINDOWS' ],
}],
['OS!="win" or llvm_version!="0.0"', {
'cflags': [ '-mssse3' ],
'xcode_settings': {
'OTHER_CFLAGS': [ '-mssse3' ],
},
}],
],
}],

@addaleax addaleax Mar 20, 2023

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's pretty unclear why it would be safe to assume SSSE3 support here.

@kvakil kvakil Mar 22, 2023

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

this is copied from deps/zlib/zlib.gyp; runtime CPU detection is used to decide which SIMD version to call into. (Or are you suggesting that we should add a comment to that effect?)

['arm_fpu=="neon"', {
'defines': [ 'ADLER32_SIMD_NEON' ],
}],
],
'include_dirs': [ '<(V8_ZLIB_ROOT)' ],
'direct_dependent_settings': {
'conditions': [
['target_arch in "ia32 x64" and OS!="ios"', {
'defines': [ 'ADLER32_SIMD_SSSE3' ],
'conditions': [
['OS=="win"', {
'defines': [ 'X86_WINDOWS' ],
},{
'defines': [ 'X86_NOT_WINDOWS' ],
}],
],
}],
['arm_fpu=="neon"', {
'defines': [ 'ADLER32_SIMD_NEON' ],
}],
],
'include_dirs': [ '<(V8_ZLIB_ROOT)' ],
},
'sources': [
'<!@pymod_do_main(GN-scraper "<(V8_ZLIB_ROOT)/BUILD.gn" "\\"zlib_adler32_simd\\".*?sources = ")',
],
}, # v8_zlib_adler32_simd

{
'target_name': 'v8_zlib',
'type': 'static_library',
Expand All @@ -1898,52 +1948,61 @@
}]
]
}],
['target_arch in "ia32 x64" and OS!="ios"', {
'dependencies': [
'v8_zlib_adler32_simd',
],
}],
['arm_fpu=="neon"', {
'dependencies': [
'v8_zlib_adler32_simd',
],
}],
],
'direct_dependent_settings': {
'include_dirs': [
'<(V8_ROOT)/third_party/zlib',
'<(V8_ROOT)/third_party/zlib/google',
'<(V8_ZLIB_ROOT)',
'<(V8_ZLIB_ROOT)/google',
],
},
'defines': [ 'ZLIB_IMPLEMENTATION' ],
'include_dirs': [
'<(V8_ROOT)/third_party/zlib',
'<(V8_ROOT)/third_party/zlib/google',
'<(V8_ZLIB_ROOT)',
'<(V8_ZLIB_ROOT)/google',
],
'sources': [
'<(V8_ROOT)/third_party/zlib/adler32.c',
'<(V8_ROOT)/third_party/zlib/chromeconf.h',
'<(V8_ROOT)/third_party/zlib/compress.c',
'<(V8_ROOT)/third_party/zlib/contrib/optimizations/insert_string.h',
'<(V8_ROOT)/third_party/zlib/contrib/optimizations/insert_string.h',
'<(V8_ROOT)/third_party/zlib/cpu_features.c',
'<(V8_ROOT)/third_party/zlib/cpu_features.h',
'<(V8_ROOT)/third_party/zlib/crc32.c',
'<(V8_ROOT)/third_party/zlib/crc32.h',
'<(V8_ROOT)/third_party/zlib/deflate.c',
'<(V8_ROOT)/third_party/zlib/deflate.h',
'<(V8_ROOT)/third_party/zlib/gzclose.c',
'<(V8_ROOT)/third_party/zlib/gzguts.h',
'<(V8_ROOT)/third_party/zlib/gzlib.c',
'<(V8_ROOT)/third_party/zlib/gzread.c',
'<(V8_ROOT)/third_party/zlib/gzwrite.c',
'<(V8_ROOT)/third_party/zlib/infback.c',
'<(V8_ROOT)/third_party/zlib/inffast.c',
'<(V8_ROOT)/third_party/zlib/inffast.h',
'<(V8_ROOT)/third_party/zlib/inffixed.h',
'<(V8_ROOT)/third_party/zlib/inflate.c',
'<(V8_ROOT)/third_party/zlib/inflate.h',
'<(V8_ROOT)/third_party/zlib/inftrees.c',
'<(V8_ROOT)/third_party/zlib/inftrees.h',
'<(V8_ROOT)/third_party/zlib/trees.c',
'<(V8_ROOT)/third_party/zlib/trees.h',
'<(V8_ROOT)/third_party/zlib/uncompr.c',
'<(V8_ROOT)/third_party/zlib/zconf.h',
'<(V8_ROOT)/third_party/zlib/zlib.h',
'<(V8_ROOT)/third_party/zlib/zutil.c',
'<(V8_ROOT)/third_party/zlib/zutil.h',
'<(V8_ROOT)/third_party/zlib/google/compression_utils_portable.cc',
'<(V8_ROOT)/third_party/zlib/google/compression_utils_portable.h',
'<(V8_ZLIB_ROOT)/chromeconf.h',
'<(V8_ZLIB_ROOT)/adler32.c',
'<(V8_ZLIB_ROOT)/compress.c',
'<(V8_ZLIB_ROOT)/contrib/optimizations/insert_string.h',
'<(V8_ZLIB_ROOT)/cpu_features.c',
'<(V8_ZLIB_ROOT)/cpu_features.h',
'<(V8_ZLIB_ROOT)/crc32.c',
'<(V8_ZLIB_ROOT)/crc32.h',
'<(V8_ZLIB_ROOT)/deflate.c',
'<(V8_ZLIB_ROOT)/deflate.h',
'<(V8_ZLIB_ROOT)/gzclose.c',
'<(V8_ZLIB_ROOT)/gzguts.h',
'<(V8_ZLIB_ROOT)/gzlib.c',
'<(V8_ZLIB_ROOT)/gzread.c',
'<(V8_ZLIB_ROOT)/gzwrite.c',
'<(V8_ZLIB_ROOT)/infback.c',
'<(V8_ZLIB_ROOT)/inffast.c',
'<(V8_ZLIB_ROOT)/inffast.h',
'<(V8_ZLIB_ROOT)/inffixed.h',
'<(V8_ZLIB_ROOT)/inflate.c',
'<(V8_ZLIB_ROOT)/inflate.h',
'<(V8_ZLIB_ROOT)/inftrees.c',
'<(V8_ZLIB_ROOT)/inftrees.h',
'<(V8_ZLIB_ROOT)/trees.c',
'<(V8_ZLIB_ROOT)/trees.h',
'<(V8_ZLIB_ROOT)/uncompr.c',
'<(V8_ZLIB_ROOT)/zconf.h',
'<(V8_ZLIB_ROOT)/zlib.h',
'<(V8_ZLIB_ROOT)/zutil.c',
'<(V8_ZLIB_ROOT)/zutil.h',
'<(V8_ZLIB_ROOT)/google/compression_utils_portable.cc',
'<(V8_ZLIB_ROOT)/google/compression_utils_portable.h',
],
}, # v8_zlib
],
Expand Down