commit | d7eb93ffcbc2112023dd5ed4bcfb44dcef23dcfb | [log] [tgz] |
---|---|---|
author | Jeremiah Griffin <jzgriffin@google.com> | Tue Sep 12 11:27:57 2023 -0700 |
committer | Jeremiah Griffin <jzgriffin@google.com> | Wed Sep 13 09:00:53 2023 -0700 |
tree | f65983c07e7f220df262148206e22d0cc4025ec2 | |
parent | 1fc4e250e0122e75c3dd0510c1627db4767c87d6 [diff] |
code_gen_lib: Improve floating-point code gen This change replaces manual NaN boxing of float32 arguments, which involved two moves in memory, with the optimized version from MacroAssembler. This optimized version does NaN boxing in-place in the XMM register and then moves to memory once. This change also uses vmov to copy floating-point arguments and results when the host supports AVX. Test: berberis_all Bug: 282063730 Change-Id: I90cf2d0f90d00027dd2ac0db54e16a4c7b0a53f3
Berberis: dynamic binary translator to run Android apps with riscv64 native code on x86_64 devices or emulators.