projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: bab44a6) | patch
[CodeGen] fix test to be (mostly) independent of LLVM optimizer; NFC
authorSanjay Patel <spatel@rotateright.com>
Sun, 10 May 2020 15:19:43 +0000 (11:19 -0400)
committerSanjay Patel <spatel@rotateright.com>
Sun, 10 May 2020 15:19:43 +0000 (11:19 -0400)
commitbcc5ed7b24e921c8902d0d0db614576bd249f128
tree59d47c0e9bc6324680219480d38d60115b211d1dtree | snapshot
parentbab44a698e821292ecce4ee01a61964a0a40c33ecommit | diff
[CodeGen] fix test to be (mostly) independent of LLVM optimizer; NFC

This test would break with the proposed change to IR canonicalization
in D79171. The raw unoptimized IR from clang is massive, so I've
replaced -instcombine with -mem2reg to make it more manageable,
but still be unlikely to break with unrelated changed to optimization.
clang/test/CodeGen/aarch64-neon-fp16fml.c diff | blob | history
Domain: System / Toolchain;