This is the mail archive of the
mailing list for the glibc project.
Re: Another ARM cancellation patch
- From: Philip Blundell <pb at nexus dot co dot uk>
- To: Daniel Jacobowitz <drow at mvista dot com>
- Cc: libc-alpha at sources dot redhat dot com
- Date: 24 Mar 2003 21:53:39 +0000
- Subject: Re: Another ARM cancellation patch
- References: <20030324213621.GA29418@nevyn.them.org>
On Mon, 2003-03-24 at 21:36, Daniel Jacobowitz wrote:
> make: *** [/home/testing/ipaq-glibc/glibc/math/test-fenv.out] Error 1
That one has always been there. I think it's a kernel bug. Same goes
> /tmp/ccBrmj5M.s:27: Error: selected processor does not support `rfs r4'
Don't think I've seen that one before. rfs is an fpa instruction, so it
might be caused by bad cflags or a binutils bug.
> make: *** [/home/testing/ipaq-glibc/glibc/math/test-float.out] Error 139
> make: *** [/home/testing/ipaq-glibc/glibc/math/test-double.out] Error 139
> make: *** [/home/testing/ipaq-glibc/glibc/math/test-ifloat.out] Error 1
> make: *** [/home/testing/ipaq-glibc/glibc/math/test-idouble.out] Error 1
Those tests have been failing for a while, but I think the segfaults are
a fairly recent development.
> make: *** [/home/testing/ipaq-glibc/glibc/elf/check-textrel.out] Error 1
That one seems likely to be a binutils bug as well. I think it's caused
by a spurious R_ARM_NONE reloc.
> 2003-03-24 Daniel Jacobowitz <drow at mvista dot com>
> * sysdeps/unix/sysv/linux/arm/sysdep-cancel.h (DOARGS_5,
> (DOARGS_7): Redefine.
The patch looks good to me. Thanks for fixing that.