* Someone broke bootstrap
@ 2003-10-08 2:11 Carlo Wood
2003-10-08 13:02 ` Solved! (was: Re: Someone broke bootstrap) Carlo Wood
0 siblings, 1 reply; 10+ messages in thread
From: Carlo Wood @ 2003-10-08 2:11 UTC (permalink / raw)
To: gcc; +Cc: roger, zack
gcc -c -g -DIN_GCC -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedantic -Wno-long-long -fno-common -DHAVE_CONFIG_H -I. -I. -I../../../gcc/gcc-mainline/gcc -I../../../gcc/gcc-mainline/gcc/. -I../../../gcc/gcc-mainline/gcc/../include ../../../gcc/gcc-mainline/gcc/optabs.c -o optabs.o
../../../gcc/gcc-mainline/gcc/optabs.c: In function `prepare_float_lib_cmp':
../../../gcc/gcc-mainline/gcc/optabs.c:3886: `eqhf2_libfunc' undeclared (first use in this function)
../../../gcc/gcc-mainline/gcc/optabs.c:3886: (Each undeclared identifier is reported only once
../../../gcc/gcc-mainline/gcc/optabs.c:3886: for each function it appears in.)
../../../gcc/gcc-mainline/gcc/optabs.c:3890: `nehf2_libfunc' undeclared (first use in this function)
../../../gcc/gcc-mainline/gcc/optabs.c:3894: `gthf2_libfunc' undeclared (first use in this function)
../../../gcc/gcc-mainline/gcc/optabs.c:3899: `lthf2_libfunc' undeclared (first use in this function)
../../../gcc/gcc-mainline/gcc/optabs.c:3904: `gehf2_libfunc' undeclared (first use in this function)
../../../gcc/gcc-mainline/gcc/optabs.c:3909: `lehf2_libfunc' undeclared (first use in this function)
etc etc.
Guys, aren't we doing bootstraps anymore before a commit?
--
Carlo Wood <carlo@alinoe.com>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Solved! (was: Re: Someone broke bootstrap)
2003-10-08 2:11 Someone broke bootstrap Carlo Wood
@ 2003-10-08 13:02 ` Carlo Wood
0 siblings, 0 replies; 10+ messages in thread
From: Carlo Wood @ 2003-10-08 13:02 UTC (permalink / raw)
To: gcc, roger, zack
On Wed, Oct 08, 2003 at 04:11:54AM +0200, Carlo Wood wrote:
> gcc -c -g -DIN_GCC -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedantic -Wno-long-long -fno-common -DHAVE_CONFIG_H -I. -I. -I../../../gcc/gcc-mainline/gcc -I../../../gcc/gcc-mainline/gcc/. -I../../../gcc/gcc-mainline/gcc/../include ../../../gcc/gcc-mainline/gcc/optabs.c -o optabs.o
> ../../../gcc/gcc-mainline/gcc/optabs.c: In function `prepare_float_lib_cmp':
> ../../../gcc/gcc-mainline/gcc/optabs.c:3886: `eqhf2_libfunc' undeclared (first use in this function)
> ../../../gcc/gcc-mainline/gcc/optabs.c:3886: (Each undeclared identifier is reported only once
> ../../../gcc/gcc-mainline/gcc/optabs.c:3886: for each function it appears in.)
> ../../../gcc/gcc-mainline/gcc/optabs.c:3890: `nehf2_libfunc' undeclared (first use in this function)
> ../../../gcc/gcc-mainline/gcc/optabs.c:3894: `gthf2_libfunc' undeclared (first use in this function)
> ../../../gcc/gcc-mainline/gcc/optabs.c:3899: `lthf2_libfunc' undeclared (first use in this function)
> ../../../gcc/gcc-mainline/gcc/optabs.c:3904: `gehf2_libfunc' undeclared (first use in this function)
> ../../../gcc/gcc-mainline/gcc/optabs.c:3909: `lehf2_libfunc' undeclared (first use in this function)
> etc etc.
>
>
> Guys, aren't we doing bootstraps anymore before a commit?
Sorry - I find out why it was broken for me: I had a sticky version tag
on optabs.c, it wasn't updated correctly.
Bootstrap works fine!
--
Carlo Wood <carlo@alinoe.com>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Someone broke bootstrap
@ 2003-10-09 3:02 Roger Sayle
0 siblings, 0 replies; 10+ messages in thread
From: Roger Sayle @ 2003-10-09 3:02 UTC (permalink / raw)
To: John David Anglin, Eric Christopher; +Cc: Zack Weinberg, gcc
Hi David and Eric,
> > I'm seeing this on frv too. I blame Zack (at least in the little
> > bit of looking I did...), but that's ok. I'm sure he'll fix it.
>
> Yes, it looks as if this is caused by one of Zack's recent changes.
Mea culpa. Zack is completely innocent, the testsuite regressions
are my fault. Zack asked that I modify my patch to optabs.c's
prepare_float_lib_cmp to call emit_libcall_block directly, but I
screwed up the reorganization, not realizing that "equiv" mustn't
be a NULL_RTX, otherwise very bad things happen...
I've already posted a fix at:
http://gcc.gnu.org/ml/gcc-patches/2003-10/msg00635.html
Hopefully, it will be reviewed in the very near future.
The problem shouldn't affect bootstrap as it only shows up with
-O0 when using software floating point, but it is likely to be
the cause of FRV's/HPPA64's regressions.
Sorry again for any inconvenience. The problem doesn't show up
on x86, and even the -msoft-float tests are never run at -O0,
so it wasn't found during my retesting. The original patch to
prepare_float_lib_cmp was more widely tested without any problems,
but I underestimated the "obvious" tidy-up.
Roger
--
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Someone broke bootstrap
2003-10-08 7:25 ` Eric Christopher
@ 2003-10-08 17:26 ` John David Anglin
0 siblings, 0 replies; 10+ messages in thread
From: John David Anglin @ 2003-10-08 17:26 UTC (permalink / raw)
To: Eric Christopher; +Cc: gcc, zack
> I'm seeing this on frv too. I blame Zack (at least in the little bit of
> looking I did...), but that's ok. I'm sure he'll fix it.
Yes, it looks as if this is caused by one of Zack's recent changes.
We have:
#0 0x400000000035ee3c in copy_rtx (orig=0x0) at ../../gcc/gcc/rtl.c:377
#1 0x40000000002f869c in emit_libcall_block (insns=0x800003fffed1f550,
target=0x800003fffed20600, result=0x800003fffed1f2d0, equiv=0x0)
at ../../gcc/gcc/optabs.c:3397
#2 0x40000000002f9430 in emit_cmp_and_jump_insns (x=0x800003fffed3bfa0,
y=0x800003fffeff2768, comparison=NE, size=0x72, mode=TFmode, unsignedp=0,
label=0x800003fffed3bf50) at ../../gcc/gcc/optabs.c:3946
emit_libcall_block is called with equiv=0 when optimize is 0. We end
up trying to copy a NULL_RTX at optabs.c:3397.
Dave
--
J. David Anglin dave.anglin@nrc-cnrc.gc.ca
National Research Council of Canada (613) 990-0752 (FAX: 952-6602)
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Someone broke bootstrap
@ 2003-10-08 14:03 Kelley Cook
0 siblings, 0 replies; 10+ messages in thread
From: Kelley Cook @ 2003-10-08 14:03 UTC (permalink / raw)
To: gcc, gcc-patches; +Cc: Bernd Jendrissek, Carlo Wood
> Commit done as obvious.
> http://gcc.gnu.org/ml/gcc-cvs/2003-10/msg00283.html
>
> Tested with 2.5.4a and 2.5.31. Note that 2.5.31 accepts both '-ooutput'
> as well as '-o output', hence the mistake that Kelley made I assume.
It turns out the Debian's flex-old 2.5.4a-1 has the snippet below in its
debian/changeLog why is apparently why I never ran into the bug even
though I attempted to test with both 2.5.4a and a self-compiled 2.5.31.
flex (2.5.4a-15) unstable; urgency=low
* flex -o "ddd" doesn't write to the file "ddd", instead it tries to
write to the file "" (yes, blank) and read from "ddd". I changed the
main.c to use the next arg as the output file name if the next arg
does not start with a -.
closes: Bug#122581
Again, sorry to all and thanks again for Carlo and Bernd for fixing it.
Kelley Cook
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Someone broke bootstrap
2003-10-08 10:15 ` David Ayers
@ 2003-10-08 11:57 ` Carlo Wood
0 siblings, 0 replies; 10+ messages in thread
From: Carlo Wood @ 2003-10-08 11:57 UTC (permalink / raw)
To: David Ayers; +Cc: John David Anglin, gcc, kcook34
On Wed, Oct 08, 2003 at 12:10:14PM +0200, David Ayers wrote:
> John David Anglin wrote:
> >How about
> >
> >if flex -o g-$$ ../../gcc/gcc/gengtype-lex.l; then \
> > mv -f g-$$ ../../gcc/gcc/gengtype-lex.c ; \
> >else \
> > rm -f g-$$ ; \
> > false ; \
> >fi
> >flex: can't open g-16194
> >
> >on i686-pc-linux-gnu.
> >
> >There are also a bunch of testsuite regressions on hppa64-hp-hpux11.11.
> >
> >Dave
> >
> >
> This seems to be caused by:
>
> http://gcc.gnu.org/ml/gcc-patches/2003-10/msg00529.html
>
> Is flex 2.5.4 unsupported?
No, the patch is wrong.
I'll write a patch and commit it as obvious in a few minutes.
Flex doesn't take a space after the -o option. It must be -ooutput,
not -o output. Kelley, assuming you tested this - what version
of flex do you use that this worked for you??
--
Carlo Wood <carlo@alinoe.com>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Someone broke bootstrap
2003-10-08 2:47 Someone broke bootstrap John David Anglin
@ 2003-10-08 10:15 ` David Ayers
2003-10-08 11:57 ` Carlo Wood
0 siblings, 1 reply; 10+ messages in thread
From: David Ayers @ 2003-10-08 10:15 UTC (permalink / raw)
To: John David Anglin; +Cc: gcc, carlo, kcook34
John David Anglin wrote:
>How about
>
>if flex -o g-$$ ../../gcc/gcc/gengtype-lex.l; then \
> mv -f g-$$ ../../gcc/gcc/gengtype-lex.c ; \
>else \
> rm -f g-$$ ; \
> false ; \
>fi
>flex: can't open g-16194
>
>on i686-pc-linux-gnu.
>
>There are also a bunch of testsuite regressions on hppa64-hp-hpux11.11.
>
>Dave
>
>
This seems to be caused by:
http://gcc.gnu.org/ml/gcc-patches/2003-10/msg00529.html
Is flex 2.5.4 unsupported?
Cheers,
David Ayers
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Someone broke bootstrap
2003-10-08 3:11 ` John David Anglin
@ 2003-10-08 7:25 ` Eric Christopher
2003-10-08 17:26 ` John David Anglin
0 siblings, 1 reply; 10+ messages in thread
From: Eric Christopher @ 2003-10-08 7:25 UTC (permalink / raw)
To: John David Anglin; +Cc: gcc, zack
On Tue, 2003-10-07 at 20:11, John David Anglin wrote:
> > There are also a bunch of testsuite regressions on hppa64-hp-hpux11.11.
>
> See <http://gcc.gnu.org/ml/gcc-testresults/2003-10/msg00324.html>.
>
> The gcc.log has messages similar to the following for the new regressions:
>
> /xxx/gnu/gcc-3.3/gcc/gcc/testsuite/gcc.dg/compat/scalar-by-value-1_x.c:154: inte
> rnal compiler error: in copy_rtx, at rtl.c:377
> Please submit a full bug report,
I'm seeing this on frv too. I blame Zack (at least in the little bit of
looking I did...), but that's ok. I'm sure he'll fix it.
-eric
--
Eric Christopher <echristo@redhat.com>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Someone broke bootstrap
[not found] <no.id>
@ 2003-10-08 3:11 ` John David Anglin
2003-10-08 7:25 ` Eric Christopher
0 siblings, 1 reply; 10+ messages in thread
From: John David Anglin @ 2003-10-08 3:11 UTC (permalink / raw)
To: John David Anglin; +Cc: gcc
> There are also a bunch of testsuite regressions on hppa64-hp-hpux11.11.
See <http://gcc.gnu.org/ml/gcc-testresults/2003-10/msg00324.html>.
The gcc.log has messages similar to the following for the new regressions:
/xxx/gnu/gcc-3.3/gcc/gcc/testsuite/gcc.dg/compat/scalar-by-value-1_x.c:154: inte
rnal compiler error: in copy_rtx, at rtl.c:377
Please submit a full bug report,
Dave
--
J. David Anglin dave.anglin@nrc-cnrc.gc.ca
National Research Council of Canada (613) 990-0752 (FAX: 952-6602)
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: Someone broke bootstrap
@ 2003-10-08 2:47 John David Anglin
2003-10-08 10:15 ` David Ayers
0 siblings, 1 reply; 10+ messages in thread
From: John David Anglin @ 2003-10-08 2:47 UTC (permalink / raw)
To: gcc; +Cc: carlo
How about
if flex -o g-$$ ../../gcc/gcc/gengtype-lex.l; then \
mv -f g-$$ ../../gcc/gcc/gengtype-lex.c ; \
else \
rm -f g-$$ ; \
false ; \
fi
flex: can't open g-16194
on i686-pc-linux-gnu.
There are also a bunch of testsuite regressions on hppa64-hp-hpux11.11.
Dave
--
J. David Anglin dave.anglin@nrc-cnrc.gc.ca
National Research Council of Canada (613) 990-0752 (FAX: 952-6602)
^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2003-10-09 3:02 UTC | newest]
Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-10-08 2:11 Someone broke bootstrap Carlo Wood
2003-10-08 13:02 ` Solved! (was: Re: Someone broke bootstrap) Carlo Wood
2003-10-08 2:47 Someone broke bootstrap John David Anglin
2003-10-08 10:15 ` David Ayers
2003-10-08 11:57 ` Carlo Wood
[not found] <no.id>
2003-10-08 3:11 ` John David Anglin
2003-10-08 7:25 ` Eric Christopher
2003-10-08 17:26 ` John David Anglin
2003-10-08 14:03 Kelley Cook
2003-10-09 3:02 Roger Sayle
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).