Re: glibc 2.1.92 make check results under Linux PPC

[Andreas Jaeger <aj at suse dot de>]

>>>>> hendricks  writes:

 > Hi,
 > I just finished building glibc-2.1.92 using a fixed cvs glibc-2.1.3,
 > Franz latest gcc, and HJ Lu's latest binutils (along with an upgraded
 > make-3.79.1) on a 2.2.15 kernel using PPC linux.

 > Here are the only two failures from make check:

 > [root@localhost rt]# GCONV_PATH=../iconvdata LC_ALL=C  ../elf/ld.so.1
 > --library-path
 > ..:../math:../elf:../dlfcn:../nss:../nis:.:../resolv:../crypt:../linuxthreads
 > ./tst-aio5
 > Sleeping...
 > Sleeping...
 > Didn't expect signal from child: got `Segmentation fault'

I hope that aio5 is fixed, Uli just made a change.  
 > [root@localhost elf]# GCONV_PATH=../iconvdata LC_ALL=C
 > LD_PRELOAD=vismod3.so ../elf/ld.so.1 --library-path
 > ..:../math:.:../dlfcn:../nss:../nis:../rt:../resolv:../crypt:../linuxthreads
 > ./vismain
 > Didn't expect signal from child: got `Illegal instruction'

This is new to me.  vismain might have problems with buggy binutils,
on ia32 it fails if I use gcc 2.95.2 with a certain binutils version,
but the error message is different.

 > If these are known problems or recently fixed, please let me know and I
 > will upgrade to the latest cvs and try again.

But I'm running daily build of glibc on ia32 and never have seen such
a problem.  It looks like this is powerpc specific - and neither Uli
or myself have a powerpc machine to test this.  Could you investigate
this yourself, please?

Thanks,
Andreas
-- 
 Andreas Jaeger
  SuSE Labs aj@suse.de
   private aj@arthur.inka.de
    http://www.suse.de/~aj