Branch: refs/heads/main
Home: https://github.com/WebKit/WebKit
Commit: dbf8cb7e3eccb57417d8846bc3d659d9a8530838
https://github.com/WebKit/WebKit/commit/dbf8cb7e3eccb57417d8846bc3d659d9a8530838
Author: Yusuke Suzuki <[email protected]>
Date: 2026-03-15 (Sun, 15 Mar 2026)
Changed paths:
M Source/JavaScriptCore/assembler/ARM64Assembler.h
M Source/JavaScriptCore/assembler/MacroAssemblerARM64.h
M Source/JavaScriptCore/b3/B3LowerToAir.cpp
M Source/JavaScriptCore/b3/B3ReduceStrength.cpp
M Source/JavaScriptCore/b3/air/AirOpcode.opcodes
M Source/JavaScriptCore/b3/testb3.h
M Source/JavaScriptCore/b3/testb3_1.cpp
M Source/JavaScriptCore/b3/testb3_7.cpp
Log Message:
-----------
[JSC] Add more SIMD lowering
https://bugs.webkit.org/show_bug.cgi?id=309933
rdar://172525410
Reviewed by Justin Michaud.
This patch adds a bit more SIMD lowering which can be found in the
benchmarks.
1. eor3
ARM64 SHA3 has eor3, which is three-way xor. We use it if we find the
pattern.
2. VectorDotProduct -> VectorExtaddPairwise
When the input is constant unit vector (all-one), it can be converted to
VectorExtaddPairwise.
Tests: Source/JavaScriptCore/b3/testb3_1.cpp
Source/JavaScriptCore/b3/testb3_7.cpp
* Source/JavaScriptCore/assembler/ARM64Assembler.h:
* Source/JavaScriptCore/assembler/MacroAssemblerARM64.h:
(JSC::MacroAssemblerARM64::vectorXor3):
* Source/JavaScriptCore/b3/B3LowerToAir.cpp:
* Source/JavaScriptCore/b3/B3ReduceStrength.cpp:
* Source/JavaScriptCore/b3/air/AirOpcode.opcodes:
* Source/JavaScriptCore/b3/testb3.h:
* Source/JavaScriptCore/b3/testb3_1.cpp:
(run):
* Source/JavaScriptCore/b3/testb3_7.cpp:
(testVectorDotProductSplatOne):
(testVectorXor3):
Canonical link: https://commits.webkit.org/309298@main
To unsubscribe from these emails, change your notification settings at
https://github.com/WebKit/WebKit/settings/notifications
