https://codereview.chromium.org/209923002/diff/1/src/arm64/macro-assembler-arm64.cc
File src/arm64/macro-assembler-arm64.cc (right):
https://codereview.chromium.org/209923002/diff/1/src/arm64/macro-assembler-arm64.cc#newcode2027
src/arm64/macro-assembler-arm64.cc:2027: // Addresses are 48 bits so we
never need to load the upper 16 bits.
ASSERT_EQ(0, ((imm >> 48) & 0xffff));
https://codereview.chromium.org/209923002/diff/1/src/arm64/macro-assembler-arm64.cc#newcode2035
src/arm64/macro-assembler-arm64.cc:2035: if (((imm >> 32) & 0xffff) !=
0) movk(temp, (imm >> 32) & 0xffff, 32);
v8/test/mjsunit and v8/benchmarks did not hit the cases when we omit
these moves. I'd suggest to remove this optimization because it
complicates the code and does not seem to trigger in tests/benchmarks.
https://codereview.chromium.org/209923002/
--
--
v8-dev mailing list
[email protected]
http://groups.google.com/group/v8-dev
---
You received this message because you are subscribed to the Google Groups "v8-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
