[X86] Set MaxAtomicSizeInBitsSupported. (#75112)
commited4194bb8dbca5222628c2cddbc032fff57193b5
authorJames Y Knight <jyknight@google.com>
Tue, 12 Dec 2023 13:16:55 +0000 (12 08:16 -0500)
committerGitHub <noreply@github.com>
Tue, 12 Dec 2023 13:16:55 +0000 (12 08:16 -0500)
treed2ccbbb146cea36a3f61bdb75f9710c3a499e642
parent19546861772f225d2648d7a5f3fea7798a13be70
[X86] Set MaxAtomicSizeInBitsSupported. (#75112)

This will result in larger atomic operations getting expanded to
`__atomic_*` libcalls via AtomicExpandPass, which matches what Clang
already does in the frontend.
llvm/lib/Target/X86/X86ISelLowering.cpp
llvm/test/CodeGen/X86/atomic-idempotent.ll
llvm/test/CodeGen/X86/atomic-nocx16.ll
llvm/test/CodeGen/X86/atomic-ops-ancient-64.ll
llvm/test/CodeGen/X86/atomic-oversize.ll [new file with mode: 0644]
llvm/test/CodeGen/X86/atomic-xor.ll