https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119015
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Last reconfirmed| |2025-02-26
Known to fail| |14.2.0, 15.0
Known to work| |13.3.0
Status|UNCONFIRMED |NEW
Target Milestone|--- |14.3
Ever confirmed|0 |1
Summary|[14 Regression] g++ -O2 |[14/15 Regression] g++ -O2
|uses huge amounts of time |uses huge amounts of time
|and memory |and memory
--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
On x86-64 while -O2 is a lot slower it seems to not use a lot more memory
(just 30% more than -Os or -O1).
> /usr/bin/time /space/rguenther/install/gcc-14.2/bin/g++ -S
> MarbleAboutDialog.ii -Os
1.92user 0.07system 0:02.02elapsed 99%CPU (0avgtext+0avgdata
352996maxresident)k
2688inputs+1904outputs (4major+75433minor)pagefaults 0swaps
> /usr/bin/time /space/rguenther/install/gcc-14.2/bin/g++ -S
> MarbleAboutDialog.ii -O2
12.58user 0.17system 0:12.77elapsed 99%CPU (0avgtext+0avgdata
457364maxresident)k
1944inputs+2096outputs (6major+105247minor)pagefaults 0swaps
offenders are:
alias stmt walking : 1.71 ( 13%)
tree DSE : 2.32 ( 18%)
store merging : 1.65 ( 13%)
TOTAL : 13.10
given -Os isn't affected this is likely triggered by inlining (-fno-inline
brings down compile-time again). GCC 15 behaves similarly.
