Re: Regressions for i686, yesterdays CVS

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: Regressions for i686, yesterdays CVS

avner (Bugzilla)
References: <[hidden email]>

Hi Carlos,

at Thu, 5 Feb 2004 11:29:15 -0500, you wrote:

>libc-alpha,
>
>I'm seeing the following failures on i686 as of yesterday, these are
>regressions
>since my current system *was* able to build and pass these tests before.
>Has anyone seen anything similar? I'm a bit unsettled by these errors, and
>I'm inclined to think it's toolchain related.
>
>/home/carlos/src/glibc-work/src/libc/configure
>--host=i686-linux
>--build=i686-linux
>--target=i686-linux
>--prefix=/usr
>--enable-add-ons=linuxthreads
>--enable-all-warnings
>
>make[2]: [/home/carlos/src/glibc-work/build-i686/posix/annexc.out] Error 1
>(ignored)
>make[2]: *** [/home/carlos/src/glibc-work/build-i686/resolv/tst-leaks.out]
>Error 1
>make[1]: *** [resolv/tests] Error 2
>make[2]: *** [/home/carlos/src/glibc-work/build-i686/elf/reldep3.out] Error 1
>make[2]: *** [/home/carlos/src/glibc-work/build-i686/elf/nodelete.out] Error
1
>make[2]: *** [/home/carlos/src/glibc-work/build-i686/elf/nodelete2.out] Error
139
>make[2]: *** [/home/carlos/src/glibc-work/build-i686/elf/neededtest.out]
Error 139
>make[2]: *** [/home/carlos/src/glibc-work/build-i686/elf/unload2.out] Error
139
>make[2]: *** [/home/carlos/src/glibc-work/build-i686/elf/dblload.out] Error
139
>make[2]: *** [/home/carlos/src/glibc-work/build-i686/elf/dblunload.out] Error
139
>make[2]: *** [/home/carlos/src/glibc-work/build-i686/elf/reldep6.out] Error
139

>make[1]: *** [elf/tests] Error 2
>make: *** [check] Error 2
>
>Reading specs from
>/home/carlos/src/toolchain-build/install/bin/../lib/gcc-lib/i686-pc-linux-gnu/3.3.2/specs
>Configured with: ../gcc-cvs/configure  :
>(reconfigured) ../gcc-cvs/configure
>--prefix=/home/carlos/src/toolchain-build/install/
>Thread model: posix
>gcc version 3.3.2
>
>GNU ld version 2.14.90 20031108
>GNU sed version 4.0.7
>grep (GNU grep) 2.4.2
>autoconf (GNU Autoconf) 2.58
>gettext (GNU gettext) 0.10.40
>
>c.

Well, that was a lot of time ago...
but this week I tried compiling glibc-2.3.4 and went into the exact same
problems, with very similar environment.

If you did not yet succeeded resolving the problem... what I did at last was
to rebuild gcc.
I came to the conclusion that, since my gcc was built when I still had older
version of binutils (2.12.something), it configured some of its defaults
according to that version (look at the man page of gcc, at the section
dealing with the option --static-libgcc for example).
after I rebuilt gcc, this time aware of the newly installed ld, I started libc
compilation all over again. this time it worked out of the box. all the tests
passed, and I only had one small problem installing the locales, but that was
easily fixed.

--
Best regard,
Avner Shapiro
______________________________________

Software Developer, Jungo - R&D
email: [hidden email]
Web:   http://www.jungo.com
Phone: 1-877-514-0537(USA) +972-9-8859365(Worldwide), ext. 286
Fax:   1-877-514-0538(USA) +972-9-8859366(Worldwide)
Reply | Threaded
Open this post in threaded view
|

Re: Regressions for i686, yesterdays CVS

Carlos O'Donell-2
On 12/4/06, Avner Shapiro <[hidden email]> wrote:
> Well, that was a lot of time ago...
> but this week I tried compiling glibc-2.3.4 and went into the exact same
> problems, with very similar environment.

Avner,

In two months this will have been three years ago, so quite a long
time ago. Thank you for mailing the list and explaining your
particular solution to the problem. To be honest I don't remember what
I did, but future i686-pc-linux-gnu glibc testing has been
successfull.

Cheers,
Carlos.