Bug#178596: {s390} operand out of range at assembly time when using -O2
Package: gcc-3.2 Version: 1:3.2.2-0pre3 Severity: important There is a build regression of wmcoincoin on s390 between GCC 2.95 and GCC 3.0, which is caused by the optimisation flags. See http://buildd.debian.org/fetch.php?&pkg=wmcoincoin&ver=2.4.3c-1&arch=s390&stamp=1043265213&file=log&as=raw Using -O0 or -O1 gets the file to build correctly, while -O2 or -O3 fails. The preprocessed source can be found at http://people.debian.org/~joss/wmcoincoin_failure.i Regards, -- .''`. Josselin Mouette/\./\ : :' : [EMAIL PROTECTED] `. `'[EMAIL PROTECTED] `- Debian GNU/Linux -- The power of freedom
Bug#178602: gcc-3.2_1:3.2.2ds5-0pre6(hppa/unstable): FTBFS: bad build-depends
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Package: gcc-3.2 Version: 1:3.2.2ds5-0pre6 Severity: serious There was an error while trying to autobuild your package: The autobuilder takes the first package of a list of alternatives, so the current source will only auto-build on machines that have libc6.1-dev. Those need to have architectures added to them in order for the autobuild to work. (That or remove the dependency altogether. Given the versioned b-d, that's probably not an option.) lamont > Automatic build of gcc-3.2_1:3.2.2ds5-0pre6 on sarti by sbuild/hppa 1.169 > Build started at 20030124-1415 [...] > ** Using build dependencies supplied by package: > Build-Depends: libc6.1-dev (>= 2.2.5-8) | libc6-dev (>= 2.2.5-8) | > libc0.3-dev | libc1-dev, m4, autoconf2.13, automake1.4, libtool, gawk, > dejagnu (>= 1.4), expect (>= 5.38.0), bzip2, binutils (>= 2.13.90.0.10) > [!s390], binutils (>= 2.13.90.0.4) [s390], debhelper (>= 3.0.25), gperf (>= > 2.7-3), bison-1.35, flex, gettext, texinfo, zlib1g-dev, libgc6-dev [!avr > !freebsd-i386], xlibs-dev, gnat-3.2 [!arm !hurd-i386 !m68k !freebsd-i386 > !netbsd-i386] | gnat (>= 3.14p-3) [i386 powerpc sparc], libncurses5-dev, > libgmp3-dev, libc6-dev-s390x [s390], tetex-bin, locales, gcc-2.95 [m68k] [...] > The following central src deps are (probably) missing: > gobjc (>= 1:3.2.2-0pre5) [...] A full build log can be found at: http://buildd.debian.org/build.php?arch=hppa&pkg=gcc-3.2&ver=1:3.2.2ds5-0pre6 -BEGIN PGP SIGNATURE- Version: GnuPG v1.0.7 (GNU/Linux) iD8DBQE+NUJTzN/kmwoKyScRAv04AKCEtVqcba45y+oRzAE6qU9F/hyFDgCfYRkB H95VV6UnUnC+9n2icwvip0w= =OKN+ -END PGP SIGNATURE-
Re: Bug#164766: Problem with VIA C3 chip and libcrypto
> >GCC 3.2 still uses CMOVE instructions on -march=i686.
> >
> >On the other hand:
> > {"c3", PROCESSOR_I486, PTA_MMX | PTA_3DNOW},
> >GCC disagrees with you that the C3 is an i686.
gcc uses i486 scheduling because that gives best performance
The situation is as follows
gcc "i686" definition is wrong. The gcc people wont fix it because the 686
definition without cmov is mostly useless anyway.
The C3 is a model 6 cpu, the kernel reports it correctly and wont change
Non Debian packaging systems know how to handle this and regard cmov as
a mandatory CPU attribute for 686 package install. Ugly but works.
Bug#178602: gcc-3.2_1:3.2.2ds5-0pre6(hppa/unstable): FTBFS: bad build-depends
[EMAIL PROTECTED] writes: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > Package: gcc-3.2 > Version: 1:3.2.2ds5-0pre6 > Severity: serious > > There was an error while trying to autobuild your package: already fixed in the debian cvs. > The autobuilder takes the first package of a list of alternatives, so > the current source will only auto-build on machines that have libc6.1-dev. > Those need to have architectures added to them in order for the autobuild > to work. (That or remove the dependency altogether. Given the versioned > b-d, that's probably not an option.) yes, some people still want to build for woody/testing.
Re: Bug#164766: Problem with VIA C3 chip and libcrypto
At Mon, 27 Jan 2003 11:39:43 -0500 (EST),
Alan Cox wrote:
> > >GCC 3.2 still uses CMOVE instructions on -march=i686.
> > >
> > >On the other hand:
> > > {"c3", PROCESSOR_I486, PTA_MMX | PTA_3DNOW},
> > >GCC disagrees with you that the C3 is an i686.
>
> gcc uses i486 scheduling because that gives best performance
>
> The situation is as follows
>
> gcc "i686" definition is wrong. The gcc people wont fix it because the 686
> definition without cmov is mostly useless anyway.
Thanks for your explanation. Hmm. "cmov" is really key instruction...
I think it may be needed that we add -mcpu=c3 for gcc, which generates
i686 without cmov instruction.
> The C3 is a model 6 cpu, the kernel reports it correctly and wont change
Yup.
We finally conclude that we fix this problem using glibc dynamic
libraries harware capabilities. It fixes at least for dynamic libraries,
not static libraries or single binaries (such software needs to check
hardware capabilities with itself, I think).
I've made a patch, and wait testers result... (See the last mail sent
by gotom: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=164766 )
I want VIA C3 user to test the latest my libc6 build at
http://people.debian.org/~gotom/
and test with the latest ssh and libssl0.9.7. "ldd /usr/bin/ssh" tells
us whether it uses cmov libs or not.
> Non Debian packaging systems know how to handle this and regard cmov as
> a mandatory CPU attribute for 686 package install. Ugly but works.
Interesting, I guess RPM has such capability. I dunno dpkg has such
functions, but I think it's useful on debian too.
Regards,
-- gotom
Re: Bug#164766: Problem with VIA C3 chip and libcrypto
On Tue, Jan 28, 2003 at 02:39:06AM +0900, GOTO Masanori wrote:
> At Mon, 27 Jan 2003 11:39:43 -0500 (EST),
> Alan Cox wrote:
> > > >GCC 3.2 still uses CMOVE instructions on -march=i686.
> > > >
> > > >On the other hand:
> > > > {"c3", PROCESSOR_I486, PTA_MMX | PTA_3DNOW},
> > > >GCC disagrees with you that the C3 is an i686.
> >
> > gcc uses i486 scheduling because that gives best performance
> >
> > The situation is as follows
> >
> > gcc "i686" definition is wrong. The gcc people wont fix it because the 686
> > definition without cmov is mostly useless anyway.
>
> Thanks for your explanation. Hmm. "cmov" is really key instruction...
>
> I think it may be needed that we add -mcpu=c3 for gcc, which generates
> i686 without cmov instruction.
It's already there in development GCC versions. I think Andi Kleen did
it.
--
Daniel Jacobowitz
MontaVista Software Debian GNU/Linux Developer
Re: Bug#164766: Problem with VIA C3 chip and libcrypto
> I think it may be needed that we add -mcpu=c3 for gcc, which generates > i686 without cmov instruction. You want i486 scheduling for the C3 from testing so far. I've been talking to VIA about releasing all the optimisation info Alan
Bug#167014: gij/gcj wrapper scripts
Okay, sorry this has taken so long. I've put a gij-wrapper-3.2 and a gcj-wrapper-3.2 up on http://people.debian.org/~bab/java if you want to take a look. I've done some basic testing and everything seems to be in order, but I wouldn't mind doing a little more testing with some of the more complex options. Ben.
