--- Comment #60 from laurent at guerby dot net 2009-09-13 21:44 ---
Ok I filed
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41349
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #59 from ebotcazou at gcc dot gnu dot org 2009-09-13 21:31
---
> Is there a separate PR or should I reopen this one?
This one is already overloaded, I'd suggest opening a new one.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #58 from laurent at guerby dot net 2009-09-13 11:36 ---
I'm still seeing a bootstrap comparison failure on sparc-linux as of revision
151650 so after the last commit:
Comparing stages 2 and 3
warning: gcc/cc1-checksum.o differs
Bootstrap comparison failure!
gcc/sel-sched.o d
--- Comment #57 from hjl at gcc dot gnu dot org 2009-09-10 18:58 ---
Subject: Bug 41241
Author: hjl
Date: Thu Sep 10 18:57:46 2009
New Revision: 151611
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=151611
Log:
2009-09-10 H.J. Lu
Backport from mainline:
2009-
--- Comment #56 from ebotcazou at gcc dot gnu dot org 2009-09-06 21:17
---
This should be fixed everywhere.
--
ebotcazou at gcc dot gnu dot org changed:
What|Removed |Added
--
--- Comment #55 from ebotcazou at gcc dot gnu dot org 2009-09-06 21:16
---
Subject: Bug 41241
Author: ebotcazou
Date: Sun Sep 6 21:15:45 2009
New Revision: 151463
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=151463
Log:
PR bootstrap/41241
* combine-stack-adj.
--- Comment #54 from jakub at gcc dot gnu dot org 2009-09-06 19:32 ---
Subject: Bug 41241
Author: jakub
Date: Sun Sep 6 19:31:55 2009
New Revision: 151462
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=151462
Log:
PR bootstrap/41241
* combine-stack-adj.c (struct
--- Comment #53 from ebotcazou at gcc dot gnu dot org 2009-09-06 19:08
---
> FYI, here is the updated patch that I've already bootstrapped on x86_64-linux
> and i686-linux and am regtesting ATM.
This is fine, thanks (missing "records" after "This structure").
--
http://gcc.gnu.org
--- Comment #52 from jakub at gcc dot gnu dot org 2009-09-06 18:38 ---
Created an attachment (id=18518)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18518&action=view)
gcc45-pr41241.patch
FYI, here is the updated patch that I've already bootstrapped on x86_64-linux
and i686-linux
--- Comment #51 from ebotcazou at gcc dot gnu dot org 2009-09-06 18:20
---
> I tried several times to reproduce it without success. It is suspicious
> that
> only gnat generated files are different. May be the bug is in gnat specific
> files.
No, it's again CSA, I'm testing a patc
--- Comment #50 from vmakarov at redhat dot com 2009-09-06 13:46 ---
I tried several times to reproduce it without success. It is suspicious that
only gnat generated files are different. May be the bug is in gnat specific
files.
I found that gnat1 generated by a system compiler (g
--- Comment #49 from ebotcazou at gcc dot gnu dot org 2009-09-05 16:11
---
> gcc/ada/exp_util.o differs
--- exp_util-stage2.txt 2009-09-05 18:09:00.0 +0200
+++ exp_util-stage3.txt 2009-09-05 18:08:06.0 +0200
@@ -1,5 +1,5 @@
-exp_util-stage2.o: file format elf32-i38
--- Comment #48 from ebotcazou at gcc dot gnu dot org 2009-09-05 15:03
---
> This patch along with Vladamir's patch fixes the
> problem on FreeBSD. Thanks for the quick response.
The patch fixes the cfgloopmanip.o comparison failure on i586-linux as well.
Unfortunately that's pretty
--- Comment #47 from kargl at gcc dot gnu dot org 2009-09-05 05:33 ---
(In reply to comment #44)
> Created an attachment (id=18497)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18497&action=view) [edit]
> gcc45-pr41241.patch
>
> New version of the csa bugfix.
>
This patch along
--- Comment #46 from rguenth at gcc dot gnu dot org 2009-09-04 20:07
---
*** Bug 41247 has been marked as a duplicate of this bug. ***
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #45 from vmakarov at gcc dot gnu dot org 2009-09-04 19:36
---
Subject: Bug 41241
Author: vmakarov
Date: Fri Sep 4 19:36:26 2009
New Revision: 151440
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=151440
Log:
2009-09-04 Vladimir Makarov
PR bootstrap/4124
--- Comment #44 from jakub at gcc dot gnu dot org 2009-09-04 19:24 ---
Created an attachment (id=18497)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18497&action=view)
gcc45-pr41241.patch
New version of the csa bugfix.
--
jakub at gcc dot gnu dot org changed:
What
--- Comment #43 from ubizjak at gmail dot com 2009-09-04 17:58 ---
Vlad, your patch also fixes alpha. Bootstrap finished OK, I'm running testsuite
now.
Thanks!
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #42 from ro at techfak dot uni-bielefeld dot de 2009-09-04
17:36 ---
Subject: Re: [4.5 regression] bootstrap comparison failure
> --- Comment #39 from vmakarov at redhat dot com 2009-09-04 16:05 ---
> Uros, Eric, Rainer, could somebody of you check the following
--- Comment #41 from jakub at gcc dot gnu dot org 2009-09-04 16:58 ---
Created an attachment (id=18496)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18496&action=view)
gcc45-pr41241.patch
Untested fix for the cfgloopmanip.c issue. We have (mem (mem (sp + cst))
in the debug_insn,
--- Comment #40 from kargl at gcc dot gnu dot org 2009-09-04 16:56 ---
(In reply to comment #39)
> Uros, Eric, Rainer, could somebody of you check the following patch works.
> I
> tried to reproduce the bug on several available to me machines since yesterday
> night and I failed.
It
--- Comment #39 from vmakarov at redhat dot com 2009-09-04 16:05 ---
Uros, Eric, Rainer, could somebody of you check the following patch works. I
tried to reproduce the bug on several available to me machines since yesterday
night and I failed.
Index: ira.c
==
--- Comment #38 from vmakarov at redhat dot com 2009-09-04 15:25 ---
Finally, I think I found what is the reason for all these bootstrap failures.
I've changed code dealing with register preferences. Now regnfo::reg_pref is
not null from the start and some passes (now it is regmove) us
--- Comment #37 from jakub at gcc dot gnu dot org 2009-09-04 14:31 ---
Created an attachment (id=18495)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18495&action=view)
pr41241.c
Self-contained testcase for the i586 combine-stack-adj.c bug.
--
http://gcc.gnu.org/bugzilla/show
--- Comment #36 from ubizjak at gmail dot com 2009-09-04 14:17 ---
Actual differences for alpha (very minor, only different registers actually):
--- tree-vect-slp-stage2.s 2009-09-04 16:09:55.0 +0200
+++ tree-vect-slp-stage3.s 2009-09-04 16:10:04.0 +0200
@@ -52
--- Comment #35 from ubizjak at gmail dot com 2009-09-04 14:08 ---
(In reply to comment #30)
> Regarding Alpha, there seem to be code generation differences, at least in
> tree-vect-slp.o. Can you check e.g. if there are the code generation
> differences even when you compile it with st
--- Comment #34 from ro at techfak dot uni-bielefeld dot de 2009-09-04
13:39 ---
Subject: Re: [4.5 regression] bootstrap comparison failure
> --- Comment #28 from jakub at gcc dot gnu dot org 2009-09-04 12:13
> ---
> In #c21/#c22 there are code generation differences. Raine
--- Comment #33 from ebotcazou at gcc dot gnu dot org 2009-09-04 13:04
---
Created an attachment (id=18492)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18492&action=view)
Preprocessed source of stage3 cfgloopmanip.o
/home/eric/build/gcc/native32/./prev-gcc/cc1 -fpreprocessed c
--- Comment #32 from jakub at gcc dot gnu dot org 2009-09-04 13:02 ---
Re: #c29, I've managed to reproduce myself with -march=i586 -mtune=i586,
debugging csa pass now.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #31 from ebotcazou at gcc dot gnu dot org 2009-09-04 12:55
---
> Does it work if you revert
>
> 2009-09-03 Richard Guenther
>
> * dwarf2out.c (dwarf2out_do_cfi_asm): Remove check of
> eh_personality_libfunc.
>
> ?
No, it doesn't, same failure.
--
--- Comment #30 from jakub at gcc dot gnu dot org 2009-09-04 12:50 ---
Regarding Alpha, there seem to be code generation differences, at least in
tree-vect-slp.o. Can you check e.g. if there are the code generation
differences even when you compile it with stage1 in both cases, just onc
--- Comment #29 from jakub at gcc dot gnu dot org 2009-09-04 12:31 ---
In the cfgloopmanip.o case, it seems that in addition to stripping .eh_frame
section we'd need to strip also .rel.eh_frame and .rela.eh_frame.
Plus, there is a code generation difference between -g and -g0, apparently
--- Comment #28 from jakub at gcc dot gnu dot org 2009-09-04 12:13 ---
In #c21/#c22 there are code generation differences. Rainer, can you please
build it using stage1 and stage2 compiler with -da -fdump-noaddr
-fdump-unnumbered and find out in which pass they start to diverge?
--
--- Comment #27 from ubizjak at gmail dot com 2009-09-04 12:01 ---
Created an attachment (id=18490)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18490&action=view)
stage 2/3 object filles that are different
These object files are different between stage2/3 on alpha-linux-gnu.
-
--- Comment #26 from rguenther at suse dot de 2009-09-04 11:58 ---
Subject: Re: [4.5 regression] bootstrap comparison
failure
On Fri, 4 Sep 2009, ebotcazou at gcc dot gnu dot org wrote:
>
>
> --- Comment #23 from ebotcazou at gcc dot gnu dot org 2009-09-04 11:51
> ---
> >
--- Comment #25 from ebotcazou at gcc dot gnu dot org 2009-09-04 11:54
---
Created an attachment (id=18489)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18489&action=view)
stripped stage3 object file on i586-linux
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #24 from ebotcazou at gcc dot gnu dot org 2009-09-04 11:54
---
Created an attachment (id=18488)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18488&action=view)
stage2 object file on i586-linux
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #23 from ebotcazou at gcc dot gnu dot org 2009-09-04 11:51
---
> Seems to work for me now. (three of three bootstraps succeeded)
Reconfirmed at revision 151417:
Comparing stages 2 and 3
warning: gcc/cc1-checksum.o differs
warning: gcc/cc1plus-checksum.o differs
Bootstrap
--- Comment #22 from ro at gcc dot gnu dot org 2009-09-04 11:10 ---
Created an attachment (id=18487)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18487&action=view)
stage3 sparc-sun-solaris2.11 object file
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #21 from ro at gcc dot gnu dot org 2009-09-04 11:10 ---
Created an attachment (id=18486)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18486&action=view)
stage2 sparc-sun-solaris2.11 object file
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #20 from ro at gcc dot gnu dot org 2009-09-04 11:09 ---
As of r151416, there's one remaining comparison failure on
sparc-sun-solaris2.11:
gcc/c-common.o differs
I'll attach the stage2 and stage3 object files.
Rainer
--
ro at gcc dot gnu dot org changed:
W
--- Comment #19 from rguenth at gcc dot gnu dot org 2009-09-04 10:40
---
Seems to work for me now. (three of three bootstraps succeeded)
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #18 from jakub at gcc dot gnu dot org 2009-09-04 06:48 ---
For each target built after r151388, can you please attach one stage2 object
that differs and corresponding stage3 object?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #17 from kargl at gcc dot gnu dot org 2009-09-04 03:13 ---
Same problem on i686-*-*freebsd at Revision: 151401
gmake[3]: Entering directory `/usr/home/kargl/gcc/obj4x'
rm -f stage_current
gmake[3]: Leaving directory `/usr/home/kargl/gcc/obj4x'
Comparing stages 2 and 3
warnin
--- Comment #16 from ubizjak at gmail dot com 2009-09-03 21:30 ---
Alpha fails bootstrap too, r151391:
Comparing stages 2 and 3
warning: gcc/cc1plus-checksum.o differs
warning: gcc/cc1-checksum.o differs
Bootstrap comparison failure!
gcc/tree-vect-slp.o differs
gcc/c-common.o differs
gc
--- Comment #15 from ebotcazou at gcc dot gnu dot org 2009-09-03 20:55
---
The problem is not fixed at revision 151391:
Comparing stages 2 and 3
warning: gcc/cc1-checksum.o differs
warning: gcc/cc1plus-checksum.o differs
Bootstrap comparison failure!
gcc/cfgloopmanip.o differs
gcc/ada/
--- Comment #14 from ebotcazou at gcc dot gnu dot org 2009-09-03 20:00
---
> Oh, just curious - if you disable address space randomization, does the build
> succeed? (echo 0 > /proc/sys/kernel/randomize-va-space)
Disabling it is the first thing I do when I get a new box. :-)
--
h
--- Comment #13 from vmakarov at gcc dot gnu dot org 2009-09-03 18:33
---
Subject: Bug 41241
Author: vmakarov
Date: Thu Sep 3 18:33:25 2009
New Revision: 151388
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=151388
Log:
2009-09-03 Vladimir Makarov
PR bootstrap/4124
--- Comment #12 from vmakarov at redhat dot com 2009-09-03 15:01 ---
It looks as an old IRA rare hidden bug which was triggered by the new patches.
I check IRA on valgrind on a lot of tests but never saw it.
Update_equiv_reg imported from the old RA uses pseudo class but it was never
s
--- Comment #11 from rguenth at gcc dot gnu dot org 2009-09-03 14:51
---
Created an attachment (id=18476)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18476&action=view)
testcase
testcase, _umoddi_s.i from libgcc2.c
> valgrind ./cc1 -quiet -O2 -m32 t.i
==1232== Invalid read of
--- Comment #10 from hjl dot tools at gmail dot com 2009-09-03 14:25
---
I got my first compare failure at revision 151353.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #9 from rguenth at gcc dot gnu dot org 2009-09-03 14:13 ---
-g is not needed after all. It's just very random.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #8 from hjl dot tools at gmail dot com 2009-09-03 13:56 ---
It may be caused by VTA merge. I also got random compare failures.
--
hjl dot tools at gmail dot com changed:
What|Removed |Added
--
--- Comment #7 from hjl dot tools at gmail dot com 2009-09-03 13:54 ---
*** Bug 41243 has been marked as a duplicate of this bug. ***
--
hjl dot tools at gmail dot com changed:
What|Removed |Added
---
--- Comment #6 from rguenth at gcc dot gnu dot org 2009-09-03 13:30 ---
First differing dump is 183r.ira.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #5 from rguenth at gcc dot gnu dot org 2009-09-03 13:19 ---
I get different code with two invocations of stage2 gcc. -g is needed, and
you have to be lucky with address space randomization.
--
rguenth at gcc dot gnu dot org changed:
What|Removed
--- Comment #4 from rguenth at gcc dot gnu dot org 2009-09-03 13:06 ---
I get different .text sizes for _umoddi3_s.o:
- Start of section headers: 676 (bytes into file)
+ Start of section headers: 684 (bytes into file)
...
- [ 1] .text PROGBITS000
--- Comment #3 from rguenth at gcc dot gnu dot org 2009-09-03 13:00 ---
I get
Comparing stages 2 and 3
Bootstrap comparison failure!
x86_64-unknown-linux-gnu/32/libgcc/bid_binarydecimal.o differs
x86_64-unknown-linux-gnu/32/libgcc/_paritydi2.o differs
or another try
Comparing stages 2
--- Comment #2 from rguenth at gcc dot gnu dot org 2009-09-03 12:40 ---
Oh, just curious - if you disable address space randomization, does the build
succeed? (echo 0 > /proc/sys/kernel/randomize-va-space)
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241
--- Comment #1 from rguenth at gcc dot gnu dot org 2009-09-03 12:32 ---
I see similar failure with r151376 on x86_64 with the 32bit multilibs. Trying
to reproduce now.
--
rguenth at gcc dot gnu dot org changed:
What|Removed |Added
60 matches
Mail list logo
