[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#266598: Processed: severity 266598 serious

At Wed, 18 Aug 2004 21:18:05 -0700,
Debian Bug Tracking System wrote:
> This bug will prevent glibc from migrating to testing, and is easier
> to find with the appropriate severity.
> I duplicated it in a pbuilder sid chroot.

Hey, you didn't read Andreas opinion? :-D

At Wed, 18 Aug 2004 13:45:35 +0200,
Andreas Metzler wrote:
> Both 2.3.2.ds1-16 fails to build from source on sparc with:
> /lib/gcc-lib/sparc-linux/3.3.4/include -isystem /build/buildd/glibc-2.3.2.ds1/debian/include -D_LIBC_REENTRANT -include ../include/libc-symbols.h  -DPIC -DSHARED -DNOT_IN_libc=1 -DIS_IN_libm=1    -o /build/buildd/glibc-2.3.2.ds1/build-tree/sparc-sparcv9/math/e_asin.os -MD -MP -MF /build/buildd/glibc-2.3.2.ds1/build-tree/sparc-sparcv9/math/e_asin.os.dt
> {standard input}: Assembler messages:
> {standard input}:5961: Error: Illegal operands: There are only 32 single precision f registers; [0-31]
> make[3]: *** [/build/buildd/glibc-2.3.2.ds1/build-tree/sparc-sparcv9/math/e_asin.os] Error 1
> Submitting as important instead of serious because I am not sure this
> is glibc's fault. - The successful build of -15 used binutils
>, -16 failed with binutils 2.15.
> Complete >=20MB log:
> http://buildd.debian.org/fetch.php?&pkg=glibc&ver=2.3.2.ds1-16&arch=sparc&stamp=1092432565&file=log&as=raw

Already known, it's binutils problem.  I already tested binutils fix
patch.  I'll submit that patch to binutils before checking sparcv9 and
sparcv9b optimization libc sanity.

Upgrading binutils 2.15 breaks glibc compilation on 4 architectures
mips, mipsel, m68k, and sparc.  I suspect we're in the base/toolchain
freeze.  Though glibc has waited cvs update almost ten months.  Shrug.

-- gotom

Reply to: