public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug middle-end/55545] New: [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk
@ 2012-11-30 12:44 rguenth at gcc dot gnu.org
  2012-11-30 12:45 ` [Bug middle-end/55545] " rguenth at gcc dot gnu.org
                   ` (10 more replies)
  0 siblings, 11 replies; 12+ messages in thread
From: rguenth at gcc dot gnu.org @ 2012-11-30 12:44 UTC (permalink / raw)
  To: gcc-bugs


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55545

             Bug #: 55545
           Summary: [4.8 Regression] Incredibly large compile-time
                    performance regression on IA64 compiling 253.perlbmk
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: middle-end
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: rguenth@gcc.gnu.org
            Target: ia64-*-linux


compiling SPEC 2000 253.perlbmk:

cpu2000/result/log.17027:  Elapsed compile for '253.perlbmk': 00:00:17 (17)
cpu2000/result/log.17031:  Elapsed compile for '253.perlbmk': 00:00:14 (14)
cpu2000/result/log.17031:  Elapsed compile for '253.perlbmk': 00:00:45 (45)
cpu2000/result/log.17035:  Elapsed compile for '253.perlbmk': 00:00:15 (15)
cpu2000/result/log.17035:  Elapsed compile for '253.perlbmk': 00:00:17 (17)
cpu2000/result/log.17039:  Elapsed compile for '253.perlbmk': 00:00:14 (14)
cpu2000/result/log.17039:  Elapsed compile for '253.perlbmk': 00:00:44 (44)
cpu2000/result/log.17043:  Elapsed compile for '253.perlbmk': 03:00:02 (10802)
cpu2000/result/log.17043:  Elapsed compile for '253.perlbmk': 03:00:04 (10804)
cpu2000/result/log.17047:  Elapsed compile for '253.perlbmk': 03:00:02 (10802)
cpu2000/result/log.17047:  Elapsed compile for '253.perlbmk': 03:00:22 (10822)

> ls -l cpu2000/result/log.17039   
-rw-rw-r-- 1 gcctest suse 747601 Nov 14 10:26 cpu2000/result/log.17039
> ls -l cpu2000/result/log.17043
-rw-rw-r-- 1 gcctest suse 558857 Nov 16 03:18 cpu2000/result/log.17043


^ permalink raw reply	[flat|nested] 12+ messages in thread
* [Bug middle-end/55545] [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk
  2012-11-30 12:44 [Bug middle-end/55545] New: [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk rguenth at gcc dot gnu.org
@ 2012-11-30 12:45 ` rguenth at gcc dot gnu.org
  2012-11-30 13:01 ` steven at gcc dot gnu.org
                   ` (9 subsequent siblings)
  10 siblings, 0 replies; 12+ messages in thread
From: rguenth at gcc dot gnu.org @ 2012-11-30 12:45 UTC (permalink / raw)
  To: gcc-bugs


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55545

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.8.0


^ permalink raw reply	[flat|nested] 12+ messages in thread
* [Bug middle-end/55545] [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk
  2012-11-30 12:44 [Bug middle-end/55545] New: [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk rguenth at gcc dot gnu.org
  2012-11-30 12:45 ` [Bug middle-end/55545] " rguenth at gcc dot gnu.org
@ 2012-11-30 13:01 ` steven at gcc dot gnu.org
  2012-11-30 13:11 ` rguenth at gcc dot gnu.org
                   ` (8 subsequent siblings)
  10 siblings, 0 replies; 12+ messages in thread
From: steven at gcc dot gnu.org @ 2012-11-30 13:01 UTC (permalink / raw)
  To: gcc-bugs


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55545

--- Comment #1 from Steven Bosscher <steven at gcc dot gnu.org> 2012-11-30 13:00:46 UTC ---
Do you also have revision numbers in those logs?


^ permalink raw reply	[flat|nested] 12+ messages in thread
* [Bug middle-end/55545] [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk
  2012-11-30 12:44 [Bug middle-end/55545] New: [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk rguenth at gcc dot gnu.org
  2012-11-30 12:45 ` [Bug middle-end/55545] " rguenth at gcc dot gnu.org
  2012-11-30 13:01 ` steven at gcc dot gnu.org
@ 2012-11-30 13:11 ` rguenth at gcc dot gnu.org
  2012-11-30 13:15 ` rguenth at gcc dot gnu.org
                   ` (7 subsequent siblings)
  10 siblings, 0 replies; 12+ messages in thread
From: rguenth at gcc dot gnu.org @ 2012-11-30 13:11 UTC (permalink / raw)
  To: gcc-bugs


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55545

--- Comment #2 from Richard Biener <rguenth at gcc dot gnu.org> 2012-11-30 13:11:38 UTC ---
(In reply to comment #1)
> Do you also have revision numbers in those logs?

Unfortunately not, but I can see that rev. 193357 was definitely ok
and rev. 193534 definitely broken.  Maybe rev. 193429 also was still ok,
but I'm not sure on that.  Happens with both -O3 (base flags) and -O3
-funroll-loops -fpeel-loops (peak flags).

In the larger revision range the only suspicious patches are the one
for PR51447 and the gcse reg-pressure work from Bin Cheng.

It seems to be compiling build-disjuncts.c btw.


^ permalink raw reply	[flat|nested] 12+ messages in thread
* [Bug middle-end/55545] [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk
  2012-11-30 12:44 [Bug middle-end/55545] New: [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk rguenth at gcc dot gnu.org
                   ` (2 preceding siblings ...)
  2012-11-30 13:11 ` rguenth at gcc dot gnu.org
@ 2012-11-30 13:15 ` rguenth at gcc dot gnu.org
  2012-11-30 13:16 ` rguenth at gcc dot gnu.org
                   ` (6 subsequent siblings)
  10 siblings, 0 replies; 12+ messages in thread
From: rguenth at gcc dot gnu.org @ 2012-11-30 13:15 UTC (permalink / raw)
  To: gcc-bugs


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55545

--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> 2012-11-30 13:15:17 UTC ---
(In reply to comment #2)
> It seems to be compiling build-disjuncts.c btw.

which is from 197.parser - where the compiler currently is after two hours
(but in schedule_block this time).  So maybe that's an entirely different
issue (no compile-time regression recorded for 197.parser in past logfiles).


^ permalink raw reply	[flat|nested] 12+ messages in thread
* [Bug middle-end/55545] [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk
  2012-11-30 12:44 [Bug middle-end/55545] New: [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk rguenth at gcc dot gnu.org
                   ` (3 preceding siblings ...)
  2012-11-30 13:15 ` rguenth at gcc dot gnu.org
@ 2012-11-30 13:16 ` rguenth at gcc dot gnu.org
  2012-11-30 19:09 ` ebotcazou at gcc dot gnu.org
                   ` (5 subsequent siblings)
  10 siblings, 0 replies; 12+ messages in thread
From: rguenth at gcc dot gnu.org @ 2012-11-30 13:16 UTC (permalink / raw)
  To: gcc-bugs


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55545

--- Comment #4 from Richard Biener <rguenth at gcc dot gnu.org> 2012-11-30 13:16:36 UTC ---
Oh, and 1080x seems to be what ulimit -t is for these jobs - thus this are
endless loops.


^ permalink raw reply	[flat|nested] 12+ messages in thread
* [Bug middle-end/55545] [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk
  2012-11-30 12:44 [Bug middle-end/55545] New: [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk rguenth at gcc dot gnu.org
                   ` (4 preceding siblings ...)
  2012-11-30 13:16 ` rguenth at gcc dot gnu.org
@ 2012-11-30 19:09 ` ebotcazou at gcc dot gnu.org
  2012-12-03  8:43 ` ebotcazou at gcc dot gnu.org
                   ` (4 subsequent siblings)
  10 siblings, 0 replies; 12+ messages in thread
From: ebotcazou at gcc dot gnu.org @ 2012-11-30 19:09 UTC (permalink / raw)
  To: gcc-bugs


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55545

Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2012-11-30
                 CC|                            |ebotcazou at gcc dot
                   |                            |gnu.org
     Ever Confirmed|0                           |1

--- Comment #5 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2012-11-30 19:09:10 UTC ---
The last time I built and tested the compiler on IA-64, the C testsuite results
were rather bad, especially at -O3.


^ permalink raw reply	[flat|nested] 12+ messages in thread
* [Bug middle-end/55545] [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk
  2012-11-30 12:44 [Bug middle-end/55545] New: [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk rguenth at gcc dot gnu.org
                   ` (5 preceding siblings ...)
  2012-11-30 19:09 ` ebotcazou at gcc dot gnu.org
@ 2012-12-03  8:43 ` ebotcazou at gcc dot gnu.org
  2012-12-03 21:05 ` ebotcazou at gcc dot gnu.org
                   ` (3 subsequent siblings)
  10 siblings, 0 replies; 12+ messages in thread
From: ebotcazou at gcc dot gnu.org @ 2012-12-03  8:43 UTC (permalink / raw)
  To: gcc-bugs


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55545

--- Comment #6 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2012-12-03 08:42:51 UTC ---
> The last time I built and tested the compiler on IA-64, the C testsuite results
> were rather bad, especially at -O3.

I just retried and there are still a bunch of C/C++/Fortran failures at -O3.


^ permalink raw reply	[flat|nested] 12+ messages in thread
* [Bug middle-end/55545] [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk
  2012-11-30 12:44 [Bug middle-end/55545] New: [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk rguenth at gcc dot gnu.org
                   ` (6 preceding siblings ...)
  2012-12-03  8:43 ` ebotcazou at gcc dot gnu.org
@ 2012-12-03 21:05 ` ebotcazou at gcc dot gnu.org
  2012-12-04  9:01 ` rguenther at suse dot de
                   ` (2 subsequent siblings)
  10 siblings, 0 replies; 12+ messages in thread
From: ebotcazou at gcc dot gnu.org @ 2012-12-03 21:05 UTC (permalink / raw)
  To: gcc-bugs


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55545

Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Depends on|                            |55158

--- Comment #7 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2012-12-03 21:04:46 UTC ---
Don't you have the tentative patch for PR rtl-opt/55158 in your tree?  It
changes ICEs into timeouts in the testsuite.


^ permalink raw reply	[flat|nested] 12+ messages in thread
* [Bug middle-end/55545] [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk
  2012-11-30 12:44 [Bug middle-end/55545] New: [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk rguenth at gcc dot gnu.org
                   ` (7 preceding siblings ...)
  2012-12-03 21:05 ` ebotcazou at gcc dot gnu.org
@ 2012-12-04  9:01 ` rguenther at suse dot de
  2012-12-04  9:04 ` steven at gcc dot gnu.org
  2012-12-07 10:45 ` rguenth at gcc dot gnu.org
  10 siblings, 0 replies; 12+ messages in thread
From: rguenther at suse dot de @ 2012-12-04  9:01 UTC (permalink / raw)
  To: gcc-bugs


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55545

--- Comment #8 from rguenther at suse dot de <rguenther at suse dot de> 2012-12-04 09:00:29 UTC ---
On Mon, 3 Dec 2012, ebotcazou at gcc dot gnu.org wrote:

> 
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55545
> 
> Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:
> 
>            What    |Removed                     |Added
> ----------------------------------------------------------------------------
>          Depends on|                            |55158
> 
> --- Comment #7 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2012-12-03 21:04:46 UTC ---
> Don't you have the tentative patch for PR rtl-opt/55158 in your tree?  It
> changes ICEs into timeouts in the testsuite.

Indeed we have.


^ permalink raw reply	[flat|nested] 12+ messages in thread
* [Bug middle-end/55545] [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk
  2012-11-30 12:44 [Bug middle-end/55545] New: [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk rguenth at gcc dot gnu.org
                   ` (8 preceding siblings ...)
  2012-12-04  9:01 ` rguenther at suse dot de
@ 2012-12-04  9:04 ` steven at gcc dot gnu.org
  2012-12-07 10:45 ` rguenth at gcc dot gnu.org
  10 siblings, 0 replies; 12+ messages in thread
From: steven at gcc dot gnu.org @ 2012-12-04  9:04 UTC (permalink / raw)
  To: gcc-bugs


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55545

Steven Bosscher <steven at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |steven at gcc dot gnu.org

--- Comment #9 from Steven Bosscher <steven at gcc dot gnu.org> 2012-12-04 09:03:12 UTC ---
(In reply to comment #5)
> The last time I built and tested the compiler on IA-64, the C testsuite results
> were rather bad, especially at -O3.

At trunk 20121203, most of these are the same as PR55158:

$ grep -c "internal compiler error" gcc/testsuite/gcc/gcc.log.sent
947
$ egrep -c "sched-rgn.c:299[08]" gcc/testsuite/gcc/gcc.log.sent
676


^ permalink raw reply	[flat|nested] 12+ messages in thread
* [Bug middle-end/55545] [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk
  2012-11-30 12:44 [Bug middle-end/55545] New: [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk rguenth at gcc dot gnu.org
                   ` (9 preceding siblings ...)
  2012-12-04  9:04 ` steven at gcc dot gnu.org
@ 2012-12-07 10:45 ` rguenth at gcc dot gnu.org
  10 siblings, 0 replies; 12+ messages in thread
From: rguenth at gcc dot gnu.org @ 2012-12-07 10:45 UTC (permalink / raw)
  To: gcc-bugs


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55545

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |DUPLICATE

--- Comment #10 from Richard Biener <rguenth at gcc dot gnu.org> 2012-12-07 10:45:20 UTC ---
Dup of 55158 then.

*** This bug has been marked as a duplicate of bug 55158 ***


^ permalink raw reply	[flat|nested] 12+ messages in thread
end of thread, other threads:[~2012-12-07 10:45 UTC | newest]

Thread overview: 12+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2012-11-30 12:44 [Bug middle-end/55545] New: [4.8 Regression] Incredibly large compile-time performance regression on IA64 compiling 253.perlbmk rguenth at gcc dot gnu.org
2012-11-30 12:45 ` [Bug middle-end/55545] " rguenth at gcc dot gnu.org
2012-11-30 13:01 ` steven at gcc dot gnu.org
2012-11-30 13:11 ` rguenth at gcc dot gnu.org
2012-11-30 13:15 ` rguenth at gcc dot gnu.org
2012-11-30 13:16 ` rguenth at gcc dot gnu.org
2012-11-30 19:09 ` ebotcazou at gcc dot gnu.org
2012-12-03  8:43 ` ebotcazou at gcc dot gnu.org
2012-12-03 21:05 ` ebotcazou at gcc dot gnu.org
2012-12-04  9:01 ` rguenther at suse dot de
2012-12-04  9:04 ` steven at gcc dot gnu.org
2012-12-07 10:45 ` rguenth at gcc dot gnu.org

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).