https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122041
Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Last reconfirmed| |2025-09-23
CC| |ebotcazou at gcc dot gnu.org
Ever confirmed|0 |1
Status|UNCONFIRMED |WAITING
--- Comment #2 from Eric Botcazou <ebotcazou at gcc dot gnu.org> ---
Note that -O4 does not exist as an optimization level (-O3 is the maximum) and
that -finline-functions -fstrict-aliasing are automatically enabled at -O2 or
above. Can you post results for just -O3 and -O3 -funroll-loops?
Since you pass different options to the compilers, it's a bit like comparing
apples and oranges. GCC supports -mcpu=native/-mtune=native and other options
that could help, see the manual.
You'd need to look at the relocation to see the target of the call (for example
objdump -dr) but this may indeed be a missed inlining opportunity.
And please avoid posting garbage from AI, this wastes everyone's time.
