Description[turbofan/x64] Load word64 followed by a shift right 32 -> load (and sign-extend if necessary) high 32bit.
We were missing this optimization in a few cases because TruncateInt64ToInt32 was also interfering.
Also removed the equivalent from simplified-lowering.cc, as the arm64 instruction selector has a similar optimization.
R=jarin@chromium.org
Committed: https://crrev.com/06cde13e7cf47adf7c3763d417ce510d3d5fd6f8
Cr-Commit-Position: refs/heads/master@{#38711}
Patch Set 1 #
Total comments: 2
Patch Set 2 : Update. #
Messages
Total messages: 16 (10 generated)
|