public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug tree-optimization/112678] New: Massive slowdown of compilation time with PGO
@ 2023-11-23 12:57 fkastl at suse dot cz
2023-11-23 13:01 ` [Bug tree-optimization/112678] [14 regression] " sjames at gcc dot gnu.org
` (4 more replies)
0 siblings, 5 replies; 6+ messages in thread
From: fkastl at suse dot cz @ 2023-11-23 12:57 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=112678
Bug ID: 112678
Summary: Massive slowdown of compilation time with PGO
Product: gcc
Version: 14.0
Status: UNCONFIRMED
Keywords: compile-time-hog
Severity: normal
Priority: P3
Component: tree-optimization
Assignee: unassigned at gcc dot gnu.org
Reporter: fkastl at suse dot cz
CC: hubicka at gcc dot gnu.org
Target Milestone: ---
Between g:37183018134049a7 and g:af7fa3135b6b046f there is a big increase of
compilation time when compiling with profile guided optimization. Happens on
both O2 and Ofast with lto both disabled or enabled on x86_64 machines and
aarch64 machine.
Some plots that show this:
https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=474.110.8 (380%
slowdown)
https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=589.337.8 (700%
slowdown)
https://lnt.opensuse.org/db_default/v4/SPEC/graph?plot.0=286.407.8 (660%
slowdown)
All plots show compilation times of SPEC benchmarks.
^ permalink raw reply [flat|nested] 6+ messages in thread* [Bug tree-optimization/112678] [14 regression] Massive slowdown of compilation time with PGO
2023-11-23 12:57 [Bug tree-optimization/112678] New: Massive slowdown of compilation time with PGO fkastl at suse dot cz
@ 2023-11-23 13:01 ` sjames at gcc dot gnu.org
2023-11-23 13:41 ` rguenth at gcc dot gnu.org
` (3 subsequent siblings)
4 siblings, 0 replies; 6+ messages in thread
From: sjames at gcc dot gnu.org @ 2023-11-23 13:01 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=112678
Sam James <sjames at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Target Milestone|--- |14.0
^ permalink raw reply [flat|nested] 6+ messages in thread* [Bug tree-optimization/112678] [14 regression] Massive slowdown of compilation time with PGO
2023-11-23 12:57 [Bug tree-optimization/112678] New: Massive slowdown of compilation time with PGO fkastl at suse dot cz
2023-11-23 13:01 ` [Bug tree-optimization/112678] [14 regression] " sjames at gcc dot gnu.org
@ 2023-11-23 13:41 ` rguenth at gcc dot gnu.org
2023-11-23 14:53 ` hubicka at ucw dot cz
` (2 subsequent siblings)
4 siblings, 0 replies; 6+ messages in thread
From: rguenth at gcc dot gnu.org @ 2023-11-23 13:41 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=112678
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |rguenth at gcc dot gnu.org,
| |sebastian.huber@embedded-br
| |ains.de
--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Hmm, the -fprofile-update mode default didn't change, did it? I'm guessing
that for some reason the generated code is different for
-fprofile-update=single now?
Note IIRC "compile-time" for SPEC with PGO includes the training run itself,
so that might be also the reason for the slowdown (runtime performance
regression of the instrumented binary).
^ permalink raw reply [flat|nested] 6+ messages in thread* [Bug tree-optimization/112678] [14 regression] Massive slowdown of compilation time with PGO
2023-11-23 12:57 [Bug tree-optimization/112678] New: Massive slowdown of compilation time with PGO fkastl at suse dot cz
2023-11-23 13:01 ` [Bug tree-optimization/112678] [14 regression] " sjames at gcc dot gnu.org
2023-11-23 13:41 ` rguenth at gcc dot gnu.org
@ 2023-11-23 14:53 ` hubicka at ucw dot cz
2023-11-23 14:55 ` cvs-commit at gcc dot gnu.org
2023-12-14 10:43 ` [Bug tree-optimization/112678] [14 regression] Massive slowdown of compilation time with PGO since r14-5579-g20a3c74c347429 fkastl at suse dot cz
4 siblings, 0 replies; 6+ messages in thread
From: hubicka at ucw dot cz @ 2023-11-23 14:53 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=112678
--- Comment #2 from Jan Hubicka <hubicka at ucw dot cz> ---
Seems we changed default to locking increments.
jh@ryzen4:/tmp> cat t.C
void
test()
{
}
jh@ryzen4:/tmp> ~/trunk-install/bin/g++ -O2 -fprofile-generate t.C -S ; grep
lock t.s
lock addl $1, __gcov0._Z4testv(%rip)
lock addl %eax, __gcov0._Z4testv+4(%rip)
jh@ryzen4:/tmp> g++ -O2 -fprofile-generate t.C -S ; grep lock t.s
jh@ryzen4:/tmp>
so r:20a3c74c347429c109bc7002285b735be83f6a0b
Honza
^ permalink raw reply [flat|nested] 6+ messages in thread* [Bug tree-optimization/112678] [14 regression] Massive slowdown of compilation time with PGO
2023-11-23 12:57 [Bug tree-optimization/112678] New: Massive slowdown of compilation time with PGO fkastl at suse dot cz
` (2 preceding siblings ...)
2023-11-23 14:53 ` hubicka at ucw dot cz
@ 2023-11-23 14:55 ` cvs-commit at gcc dot gnu.org
2023-12-14 10:43 ` [Bug tree-optimization/112678] [14 regression] Massive slowdown of compilation time with PGO since r14-5579-g20a3c74c347429 fkastl at suse dot cz
4 siblings, 0 replies; 6+ messages in thread
From: cvs-commit at gcc dot gnu.org @ 2023-11-23 14:55 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=112678
--- Comment #3 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Sebastian Huber <sh@gcc.gnu.org>:
https://gcc.gnu.org/g:8674d70ce37ca3249a641fb418c6849d4f95f348
commit r14-5783-g8674d70ce37ca3249a641fb418c6849d4f95f348
Author: Sebastian Huber <sebastian.huber@embedded-brains.de>
Date: Thu Nov 23 14:45:50 2023 +0100
gcov: No atomic ops for -fprofile-update=single
gcc/ChangeLog:
PR tree-optimization/112678
* tree-profile.cc (tree_profiling): Do not use atomic operations
for -fprofile-update=single.
^ permalink raw reply [flat|nested] 6+ messages in thread* [Bug tree-optimization/112678] [14 regression] Massive slowdown of compilation time with PGO since r14-5579-g20a3c74c347429
2023-11-23 12:57 [Bug tree-optimization/112678] New: Massive slowdown of compilation time with PGO fkastl at suse dot cz
` (3 preceding siblings ...)
2023-11-23 14:55 ` cvs-commit at gcc dot gnu.org
@ 2023-12-14 10:43 ` fkastl at suse dot cz
4 siblings, 0 replies; 6+ messages in thread
From: fkastl at suse dot cz @ 2023-12-14 10:43 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=112678
Filip Kastl <fkastl at suse dot cz> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|UNCONFIRMED |RESOLVED
Resolution|--- |FIXED
--- Comment #4 from Filip Kastl <fkastl at suse dot cz> ---
Marking this as fixed.
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2023-12-14 10:43 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-11-23 12:57 [Bug tree-optimization/112678] New: Massive slowdown of compilation time with PGO fkastl at suse dot cz
2023-11-23 13:01 ` [Bug tree-optimization/112678] [14 regression] " sjames at gcc dot gnu.org
2023-11-23 13:41 ` rguenth at gcc dot gnu.org
2023-11-23 14:53 ` hubicka at ucw dot cz
2023-11-23 14:55 ` cvs-commit at gcc dot gnu.org
2023-12-14 10:43 ` [Bug tree-optimization/112678] [14 regression] Massive slowdown of compilation time with PGO since r14-5579-g20a3c74c347429 fkastl at suse dot cz
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).