* g++.pt/repo3.C regression between 3.1.1 and 3.2?
@ 2002-07-31 10:36 Franz Sirl
2002-08-06 12:30 ` Franz Sirl
0 siblings, 1 reply; 10+ messages in thread
From: Franz Sirl @ 2002-07-31 10:36 UTC (permalink / raw)
To: gcc; +Cc: Mark Mitchell
Hi,
while comparing i686-linux-gnu testsuite results, I noticed that
g++.pt/repo3.C fails at -O2 in 3.2, is this a regression or expected?
Ah, while we are at potential ABI problems, g++.abi/bitfields.C fails with
-funsigned-char on i686-linux-gnu.
Franz.
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: g++.pt/repo3.C regression between 3.1.1 and 3.2?
2002-07-31 10:36 g++.pt/repo3.C regression between 3.1.1 and 3.2? Franz Sirl
@ 2002-08-06 12:30 ` Franz Sirl
2002-08-06 12:34 ` Mark Mitchell
0 siblings, 1 reply; 10+ messages in thread
From: Franz Sirl @ 2002-08-06 12:30 UTC (permalink / raw)
To: gcc; +Cc: Mark Mitchell
On Mittwoch, 31. Juli 2002 17:54, Franz Sirl wrote:
> Hi,
>
> while comparing i686-linux-gnu testsuite results, I noticed that
> g++.pt/repo3.C fails at -O2 in 3.2, is this a regression or expected?
>
> Ah, while we are at potential ABI problems, g++.abi/bitfields.C fails with
> -funsigned-char on i686-linux-gnu.
Mark, what about these? Can we safely ignore them for 3.2?
Franz.
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: g++.pt/repo3.C regression between 3.1.1 and 3.2?
2002-08-06 12:30 ` Franz Sirl
@ 2002-08-06 12:34 ` Mark Mitchell
2002-08-06 12:48 ` Franz Sirl
0 siblings, 1 reply; 10+ messages in thread
From: Mark Mitchell @ 2002-08-06 12:34 UTC (permalink / raw)
To: Franz Sirl, gcc
--On Tuesday, August 06, 2002 09:30:30 PM +0200 Franz Sirl
<Franz.Sirl-kernel@lauterbach.com> wrote:
On Mittwoch, 31. Juli 2002 17:54, Franz Sirl wrote:
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: g++.pt/repo3.C regression between 3.1.1 and 3.2?
2002-08-06 12:34 ` Mark Mitchell
@ 2002-08-06 12:48 ` Franz Sirl
2002-08-06 12:54 ` Mark Mitchell
0 siblings, 1 reply; 10+ messages in thread
From: Franz Sirl @ 2002-08-06 12:48 UTC (permalink / raw)
To: Mark Mitchell, gcc
On Dienstag, 6. August 2002 21:33, Mark Mitchell wrote:
> --On Tuesday, August 06, 2002 09:30:30 PM +0200 Franz Sirl
>
> <Franz.Sirl-kernel@lauterbach.com> wrote:
> > On Mittwoch, 31. Juli 2002 17:54, Franz Sirl wrote:
> >> Hi,
> >>
> >> while comparing i686-linux-gnu testsuite results, I noticed that
> >> g++.pt/repo3.C fails at -O2 in 3.2, is this a regression or expected?
> >>
> >> Ah, while we are at potential ABI problems, g++.abi/bitfields.C fails
> >> with -funsigned-char on i686-linux-gnu.
> >
> > Mark, what about these? Can we safely ignore them for 3.2?
>
> I believe so. I thought Jason reviewed them.
There were no replies that reached me or the archives.
> I will take a look too -- can you point me at the original message?
I quoted all of it. Just compile g++.pt/repo3.C with -O2.
I get this in g++.log:
Executing on host: /home/gnu/obj/gcc32/gcc/testsuite/../g++
-B/home/gnu/obj/gcc32/gcc/testsuite/../ -nostdinc++ -I/home/gnu/obj/gcc32/i
686-redhat-linux/libstdc++-v3/include/i686-redhat-linux
-I/home/gnu/obj/gcc32/i686-redhat-linux/libstdc++-v3/include
-I/home/gnu/gcc32/l
ibstdc++-v3/libsupc++ -I/home/gnu/gcc32/libstdc++-v3/libio
-I/home/gnu/gcc32/libstdc++-v3/include/backward -I/home/gnu/gcc32/libstdc++-v
3/testsuite -fmessage-length=0 -frepo -c -O2 -o
/home/gnu/obj/gcc32/gcc/testsuite/repo3.o
/home/gnu/gcc32/gcc/testsuite/g++.old-deja/
g++.pt/repo3.C (timeout = 300)
Executing on host: /home/gnu/obj/gcc32/gcc/testsuite/../g++
-B/home/gnu/obj/gcc32/gcc/testsuite/../ /home/gnu/obj/gcc32/gcc/testsuite/re
po3.o -nostdinc++
-I/home/gnu/obj/gcc32/i686-redhat-linux/libstdc++-v3/include/i686-redhat-linux
-I/home/gnu/obj/gcc32/i686-redhat-linu
x/libstdc++-v3/include -I/home/gnu/gcc32/libstdc++-v3/libsupc++
-I/home/gnu/gcc32/libstdc++-v3/libio -I/home/gnu/gcc32/libstdc++-v3/incl
ude/backward -I/home/gnu/gcc32/libstdc++-v3/testsuite -fmessage-length=0
-frepo -L/home/gnu/obj/gcc32/i686-redhat-linux//libstdc++-v
3/src/.libs -L/home/gnu/obj/gcc32/i686-redhat-linux//libiberty -lstdc++ -lm
-O2 -o /home/gnu/obj/gcc32/gcc/testsuite/g++-pt-repo3-C.e
xe (timeout = 300)
collect: recompiling /home/gnu/gcc32/gcc/testsuite/g++.old-deja/g++.pt/repo3.C
collect: relinking
/home/gnu/obj/gcc32/gcc/testsuite/repo3.o(.gnu.linkonce.d._ZTC1DIiE4_1CIiE+0xc):
undefined reference to `typeinfo for C<int>'
/home/gnu/obj/gcc32/gcc/testsuite/repo3.o(.gnu.linkonce.d._ZTC1DIiE4_1CIiE+0x20):
undefined reference to `typeinfo for C<int>'
/home/gnu/obj/gcc32/gcc/testsuite/repo3.o(.gnu.linkonce.d._ZTC1DIiE0_1BIiE+0xc):
undefined reference to `typeinfo for B<int>'
/home/gnu/obj/gcc32/gcc/testsuite/repo3.o(.gnu.linkonce.d._ZTI1DIiE+0x10):
undefined reference to `typeinfo for B<int>'
/home/gnu/obj/gcc32/gcc/testsuite/repo3.o(.gnu.linkonce.d._ZTI1DIiE+0x18):
undefined reference to `typeinfo for C<int>'
collect2: ld returned 1 exit status
compiler exited with status 1
output is:
collect: recompiling /home/gnu/gcc32/gcc/testsuite/g++.old-deja/g++.pt/repo3.C
collect: relinking
/home/gnu/obj/gcc32/gcc/testsuite/repo3.o(.gnu.linkonce.d._ZTC1DIiE4_1CIiE+0xc):
undefined reference to `typeinfo for C<int>'
/home/gnu/obj/gcc32/gcc/testsuite/repo3.o(.gnu.linkonce.d._ZTC1DIiE4_1CIiE+0x20):
undefined reference to `typeinfo for C<int>'
/home/gnu/obj/gcc32/gcc/testsuite/repo3.o(.gnu.linkonce.d._ZTC1DIiE0_1BIiE+0xc):
undefined reference to `typeinfo for B<int>'
/home/gnu/obj/gcc32/gcc/testsuite/repo3.o(.gnu.linkonce.d._ZTI1DIiE+0x10):
undefined reference to `typeinfo for B<int>'
/home/gnu/obj/gcc32/gcc/testsuite/repo3.o(.gnu.linkonce.d._ZTI1DIiE+0x18):
undefined reference to `typeinfo for C<int>'
collect2: ld returned 1 exit status
FAIL: g++.pt/repo3.C (test for excess errors)
Franz.
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: g++.pt/repo3.C regression between 3.1.1 and 3.2?
2002-08-06 12:48 ` Franz Sirl
@ 2002-08-06 12:54 ` Mark Mitchell
2002-08-06 13:13 ` Franz Sirl
2002-08-06 16:11 ` Nathan Sidwell
0 siblings, 2 replies; 10+ messages in thread
From: Mark Mitchell @ 2002-08-06 12:54 UTC (permalink / raw)
To: Franz Sirl, gcc
I believe so. I thought Jason reviewed them.
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: g++.pt/repo3.C regression between 3.1.1 and 3.2?
2002-08-06 12:54 ` Mark Mitchell
@ 2002-08-06 13:13 ` Franz Sirl
2002-08-06 13:24 ` Mark Mitchell
2002-08-06 16:11 ` Nathan Sidwell
1 sibling, 1 reply; 10+ messages in thread
From: Franz Sirl @ 2002-08-06 13:13 UTC (permalink / raw)
To: Mark Mitchell, gcc
On Dienstag, 6. August 2002 21:52, Mark Mitchell wrote:
> >> I believe so. I thought Jason reviewed them.
> >
> > There were no replies that reached me or the archives.
>
> Do you have your original message handy for this one too?
Uhm, for which one? I don't understand what you are referring too.
> >> I will take a look too -- can you point me at the original message?
> >
> > I quoted all of it. Just compile g++.pt/repo3.C with -O2.
> >
> > I get this in g++.log:
> >
> > Executing on host: /home/gnu/o>
>
> /home/gnu/obj/gcc32/gcc/testsuite/repo3.o(.gnu.linkonce.d._ZTI1DIiE+0x18)
>
> > : undefined reference to `typeinfo for C<int>'
> >
> > collect2: ld returned 1 exit status
>
> This may or may not be a bug -- but it's not a backwards-incompatible
> bug. In other words, fixing it won't result in an ABI change. So,
> it needn't worry us at the moment.
OK, so it's just a simple regression between 3.1.1 and 3.2 due to the ABI
patches.
Franz.
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: g++.pt/repo3.C regression between 3.1.1 and 3.2?
2002-08-06 13:13 ` Franz Sirl
@ 2002-08-06 13:24 ` Mark Mitchell
2002-08-06 13:49 ` Franz Sirl
0 siblings, 1 reply; 10+ messages in thread
From: Mark Mitchell @ 2002-08-06 13:24 UTC (permalink / raw)
To: Franz Sirl, gcc; +Cc: jason, nathan
--On Tuesday, August 06, 2002 10:13:11 PM +0200 Franz Sirl
<Franz.Sirl-kernel@lauterbach.com> wrote:
On Dienstag, 6. August 2002 21:52, Mark Mitchell wrote:
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: g++.pt/repo3.C regression between 3.1.1 and 3.2?
2002-08-06 13:24 ` Mark Mitchell
@ 2002-08-06 13:49 ` Franz Sirl
2002-08-06 15:10 ` Mark Mitchell
0 siblings, 1 reply; 10+ messages in thread
From: Franz Sirl @ 2002-08-06 13:49 UTC (permalink / raw)
To: Mark Mitchell, gcc; +Cc: jason, nathan
On Dienstag, 6. August 2002 22:23, Mark Mitchell wrote:
> --On Tuesday, August 06, 2002 10:13:11 PM +0200 Franz Sirl
>
> <Franz.Sirl-kernel@lauterbach.com> wrote:
> > On Dienstag, 6. August 2002 21:52, Mark Mitchell wrote:
> >> >> I believe so. I thought Jason reviewed them.
> >> >
> >> > There were no replies that reached me or the archives.
> >>
> >> Do you have your original message handy for this one too?
> >
> > Uhm, for which one? I don't understand what you are referring too.
>
> The bitfields.C bug.
Ah, silly me.
This probably is just a simple testcase coding problem, I shouldn't have
mentioned it in the 3.2 context.
Anyway, here is the output:
FAIL: g++.abi/bitfields.C (test for excess errors)
/home/gnu/gcc32/gcc/testsuite/g++.old-deja/g++.abi/bitfields.C:140: warning:
assignment of negative value `-1' to `char'
/home/gnu/gcc32/gcc/testsuite/g++.old-deja/g++.abi/bitfields.C:140: warning:
argument of negative value `-1' to `char'
/home/gnu/gcc32/gcc/testsuite/g++.old-deja/g++.abi/bitfields.C:147: warning:
assignment of negative value `-1' to `char'
/home/gnu/gcc32/gcc/testsuite/g++.old-deja/g++.abi/bitfields.C:147: warning:
argument of negative value `-1' to `char'
/home/gnu/gcc32/gcc/testsuite/g++.old-deja/g++.abi/bitfields.C:154: warning:
assignment of negative value `-1' to `char'
/home/gnu/gcc32/gcc/testsuite/g++.old-deja/g++.abi/bitfields.C:154: warning:
argument of negative value `-1' to `char'
/home/gnu/gcc32/gcc/testsuite/g++.old-deja/g++.abi/bitfields.C:155: warning:
assignment of negative value `-1' to `char'
/home/gnu/gcc32/gcc/testsuite/g++.old-deja/g++.abi/bitfields.C:155: warning:
argument of negative value `-1' to `char'
/home/gnu/gcc32/gcc/testsuite/g++.old-deja/g++.abi/bitfields.C:156: warning:
assignment of negative value `-1' to `char'
/home/gnu/gcc32/gcc/testsuite/g++.old-deja/g++.abi/bitfields.C:156: warning:
argument of negative value `-1' to `char'
Franz.
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: g++.pt/repo3.C regression between 3.1.1 and 3.2?
2002-08-06 13:49 ` Franz Sirl
@ 2002-08-06 15:10 ` Mark Mitchell
0 siblings, 0 replies; 10+ messages in thread
From: Mark Mitchell @ 2002-08-06 15:10 UTC (permalink / raw)
To: Franz Sirl, gcc; +Cc: jason, nathan
--On Tuesday, August 06, 2002 10:49:08 PM +0200 Franz Sirl
<Franz.Sirl-kernel@lauterbach.com> wrote:
On Dienstag, 6. August 2002 22:23, Mark Mitchell wrote:
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: g++.pt/repo3.C regression between 3.1.1 and 3.2?
2002-08-06 12:54 ` Mark Mitchell
2002-08-06 13:13 ` Franz Sirl
@ 2002-08-06 16:11 ` Nathan Sidwell
1 sibling, 0 replies; 10+ messages in thread
From: Nathan Sidwell @ 2002-08-06 16:11 UTC (permalink / raw)
To: Mark Mitchell; +Cc: Franz Sirl, gcc, jason
Mark Mitchell wrote:
> > I quoted all of it. Just compile g++.pt/repo3.C with -O2.
> >
> > I get this in g++.log:
> >
> > Executing on host: /home/gnu/o>
> /home/gnu/obj/gcc32/gcc/testsuite/repo3.o(.gnu.linkonce.d._ZTI1DIiE+0x18)
> > : undefined reference to `typeinfo for C<int>'
> > collect2: ld returned 1 exit status
Yes, this must be something Jason or I broke. The fix won't
change the ABI (i.e. it can wait til 3.2.1)
nathan
--
Dr Nathan Sidwell :: http://www.codesourcery.com :: CodeSourcery LLC
'But that's a lie.' - 'Yes it is. What's your point?'
nathan@codesourcery.com : http://www.cs.bris.ac.uk/~nathan/ : nathan@acm.org
^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2002-08-06 16:11 UTC | newest]
Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2002-07-31 10:36 g++.pt/repo3.C regression between 3.1.1 and 3.2? Franz Sirl
2002-08-06 12:30 ` Franz Sirl
2002-08-06 12:34 ` Mark Mitchell
2002-08-06 12:48 ` Franz Sirl
2002-08-06 12:54 ` Mark Mitchell
2002-08-06 13:13 ` Franz Sirl
2002-08-06 13:24 ` Mark Mitchell
2002-08-06 13:49 ` Franz Sirl
2002-08-06 15:10 ` Mark Mitchell
2002-08-06 16:11 ` Nathan Sidwell
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).