Branch: refs/heads/main
Home: https://github.com/WebKit/WebKit
Commit: c4582598ea7d3cd6ada4f179a63166743ce88495
https://github.com/WebKit/WebKit/commit/c4582598ea7d3cd6ada4f179a63166743ce88495
Author: Keith Miller <[email protected]>
Date: 2026-05-01 (Fri, 01 May 2026)
Changed paths:
R JSTests/wasm/stress/inline-simd-function.js
Log Message:
-----------
Remove redundant inline-simd-function.js test
https://bugs.webkit.org/show_bug.cgi?id=313846
rdar://176050194
Reviewed by Yusuke Suzuki.
JSTests/wasm/stress/inline-simd-function.js is a 3,331-line / 208 KB
fuzzer-derived test added in 292288@main (bug 289900, "Non-SIMD function
should use SIMD for OMG catch OSR if an inlined function uses SIMD") to
exercise the OMG catch-OSR entrypoint when a non-SIMD function inlines
a SIMD function. It is annotated slow! + memoryHog!, allocates ~24 MB
of shared wasm memory, and runs 14 different JSC configuration modes,
so uses roughly 522s overall (45-55 per test runs).
The fix the test originally guarded has been superseded twice. Commit
3a46f626fff1 (bug 291594) moved SIMD-ness tracking to per-catch-handler.
Commit c410d58148a6 (bug 301380) then unified scratch-buffer slot sizing
to always fit a vector, deleted the SIMD/non-SIMD fork entirely
(buildEntryBufferForCatchSIMD / buildEntryBufferForCatchNoSIMD,
scratchBufferSlotsPerValue), and shipped a focused hand-written
regression test, JSTests/wasm/stress/simd-inline-exceptions.js
Canonical link: https://commits.webkit.org/312451@main
To unsubscribe from these emails, change your notification settings at
https://github.com/WebKit/WebKit/settings/notifications
