https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64567
Daniel Henrique Barboza <daniel.barboza at oss dot qualcomm.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |daniel.barboza at oss dot
qualcomm
| |.com
--- Comment #5 from Daniel Henrique Barboza <daniel.barboza at oss dot
qualcomm.com> ---
Hi,
Just sent a fix proposal to the ML:
https://gcc.gnu.org/pipermail/gcc-patches/2026-February/708492.html
My idea is to use phiopt to assess whether we're dealing with the cases
mentioned here:
- if someone checks if a bit/bitmask is not enabled, and if true enable it ->
just enable the bit/bitmask;
- if someone checks if a bit/bitmask is enabled, and if true disable it -> just
enable the bit/bitmask.
The end result is that the gcond is eliminated and, if the bitmask comparison
isn't used anywhere else, the bit_and operation also goes away.
