Re: GCC 3.4.2 Release Candidate Available

Re: GCC 3.4.2 Release Candidate Available

[Harald Dunkel]

I tried to build it on Debian amd64(gcc-3.4), but there was
a problem at link time of gcc saying:

:
:
/usr/bin/ld: skipping incompatible /usr/lib/libc.so when searching for -lc
/usr/bin/ld: skipping incompatible /usr/lib/libc.a when searching for -lc
/usr/bin/ld: cannot find -lc
collect2: ld returned 1 exit status
make[3]: *** [32/libgcc_s_32.so] Error 1


Probably thats expected. I read something about "a bug in
glibcs older than 2004-07-16" in a Changelog file. But
this must be a pretty serious problem if you drop compatibility.
Do you have any more information about this, e.g. a bug
report number, or something like this?


Many thanx

Harri

Re: GCC 3.4.2 Release Candidate Available

[Daniel Jacobowitz]

On Tue, Aug 31, 2004 at 08:38:51PM +0200, Harald Dunkel wrote:
> I tried to build it on Debian amd64(gcc-3.4), but there was
> a problem at link time of gcc saying:
> 
> :
> :
> /usr/bin/ld: skipping incompatible /usr/lib/libc.so when searching for -lc
> /usr/bin/ld: skipping incompatible /usr/lib/libc.a when searching for -lc
> /usr/bin/ld: cannot find -lc
> collect2: ld returned 1 exit status
> make[3]: *** [32/libgcc_s_32.so] Error 1

You can't build a multilib enabled GCC on Debian's pure64 port.  It's
not supported.  Try --disable-multilib, or discuss with the amd64 port
maintainers (the reasonable reasons, certainly) why they lay out
libraries differently than the rest of the world.

-- 
Daniel Jacobowitz

Re: GCC 3.4.2 Release Candidate Available

[Harald Dunkel]

Daniel Jacobowitz wrote:
> 
> You can't build a multilib enabled GCC on Debian's pure64 port.  It's
> not supported.  Try --disable-multilib, or discuss with the amd64 port
> maintainers (the reasonable reasons, certainly) why they lay out
> libraries differently than the rest of the world.
> 

Thats historically grown, AFAIK. Currently Debian is in a very
late state of the release procedure. I am sure that Debian will
become LSB compliant for amd64, too, when Sarge is out of the
door.

The build using --disable-multilib succeeded, so thanx very
much for your hint.


Regards

Harri

Re: GCC 3.4.2 Release Candidate Available

[Michael Chastain]

[This is a weird time for me to be awake, but I'm under a bit of stress
 right now, and it soothes me to do this ...]

Mark Mitchell writes:
> Please perform testing by building directly from these sources,
> not from CVS, so that we can test our packaging.

Okay.  I am doing:

  foreach machine $(hp_test_drive_cluster)
    mkdir build install ; cd build
    /foo/bar/configure --prefix=/foo/bar/install
    make bootstrap
    make install
  done

I've run into several bootstrap errors, which I shall proceed to
file PR's for.

Michael

Re: GCC 3.4.2 Release Candidate Available

[Mark Mitchell]

Michael Chastain wrote:

>[This is a weird time for me to be awake, but I'm under a bit of stress
> right now, and it soothes me to do this ...]
>
>Mark Mitchell writes:
>  
>
>>Please perform testing by building directly from these sources,
>>not from CVS, so that we can test our packaging.
>>    
>>
>
>Okay.  I am doing:
>
>  foreach machine $(hp_test_drive_cluster)
>    mkdir build install ; cd build
>    /foo/bar/configure --prefix=/foo/bar/install
>    make bootstrap
>    make install
>  done
>
>I've run into several bootstrap errors, which I shall proceed to
>file PR's for.
>  
>
Thanks.  If these are funny old HP machines, we'll likely ignore the 
problems, but it will be good to have the reports. :-)

-- 
Mark Mitchell
CodeSourcery, LLC
(916) 791-8304
mark@codesourcery.com

Re: GCC 3.4.2 Release Candidate Available

[Michael Chastain]

> Thanks.  If these are funny old HP machines, we'll likely ignore the 
> problems, but it will be good to have the reports. :-)

Right.

Actually, some of the problems probably stem from my use of system
'make' rather than gnu make.  Gotta polish up my build script before
complaining!

Re: GCC 3.4.2 Release Candidate Available

[Joe Buck]

On Mon, Aug 30, 2004 at 09:22:03PM -0700, Mark Mitchell wrote:
> Michael Chastain wrote:
> >I've run into several bootstrap errors, which I shall proceed to
> >file PR's for.

> Thanks.  If these are funny old HP machines, we'll likely ignore the 
> problems, but it will be good to have the reports. :-)

I successfully built RC1 and ran tests on a funny old HPUX 11.0 machine.

http://gcc.gnu.org/ml/gcc-testresults/2004-08/msg01192.html

Likewise for i686-pc-linux-gnu on an old Red Hat 7.2 with updated binutils:
http://gcc.gnu.org/ml/gcc-testresults/2004-08/msg01188.html

I also have a sparc-sun-solaris2.8 build that has completed but has not
yet finished running tests.  The results will show up soon on
gcc-testresults.

Re: GCC 3.4.2 Release Candidate Available

[Christian Joensson]

On Tue, Aug 31, 2004 at 04:52:29PM -0700, Joe Buck wrote:
> On Tue, Aug 31, 2004 at 08:55:06AM -0700, Joe Buck wrote:
> > I successfully built RC1 and ran tests on a funny old HPUX 11.0 machine.
> > ...
> > I also have a sparc-sun-solaris2.8 build that has completed but has not
> > yet finished running tests.  The results will show up soon on
> > gcc-testresults.
> 
> Results now available:
> http://gcc.gnu.org/ml/gcc-testresults/2004-08/msg01228.html
> 
> There are large numbers of g77 failures.

The same holds for sparc-linux, see, e.g., 
http://gcc.gnu.org/ml/gcc-testresults/2004-08/msg01194.html

Due to bus error.. maybe glibc related?

Cheers,

/ChJ

Re: GCC 3.4.2 Release Candidate Available

[Andrew Pinski]

On Aug 31, 2004, at 10:17 PM, Christian Joensson wrote:

> On Tue, Aug 31, 2004 at 04:52:29PM -0700, Joe Buck wrote:
>> On Tue, Aug 31, 2004 at 08:55:06AM -0700, Joe Buck wrote:
>>> I successfully built RC1 and ran tests on a funny old HPUX 11.0 
>>> machine.
>>> ...
>>> I also have a sparc-sun-solaris2.8 build that has completed but has 
>>> not
>>> yet finished running tests.  The results will show up soon on
>>> gcc-testresults.
>>
>> Results now available:
>> http://gcc.gnu.org/ml/gcc-testresults/2004-08/msg01228.html
>>
>> There are large numbers of g77 failures.
>
> The same holds for sparc-linux, see, e.g.,
> http://gcc.gnu.org/ml/gcc-testresults/2004-08/msg01194.html
>
> Due to bus error.. maybe glibc related?

no a bug in the g77 front-end, see PR 17180.

Thanks,
Andrew Pinski

GCC 3.4.2 Release Candidate Available

[Mark Mitchell]

The first GCC 3.4.2 release candidate is available from:

  ftp://gcc.gnu.org/pub/gcc/prerelease-3.4.2-20040829/

for testing.  Please perform testing by building directly from these
sources, not from CVS, so that we can test our packaging.  

If you find problems that you think should hold up the release:

(1) Put them into Bugzilla in the usual way

(2) Add me to the CC: list for the PR

(3) Add a comment to the PR that indicates that you think this problem
    should hold up the release

In general, only problems that relate to generating incorrect code or
that prevent bootstrapping on popular platforms will result in holding
up the release.

Assuming no major problems are found, the final GCC 3.4.2 release will
be made in the next few days.

--
Mark Mitchell
CodeSourcery, LLC
mark@codesourcery.com

Re: GCC 3.4.2 Release Candidate Available

[Janis Johnson]

On Sun, Aug 29, 2004 at 02:46:39PM -0700, Mark Mitchell wrote:
> The first GCC 3.4.2 release candidate is available from:
> 
>   ftp://gcc.gnu.org/pub/gcc/prerelease-3.4.2-20040829/
> 
> for testing.  Please perform testing by building directly from these
> sources, not from CVS, so that we can test our packaging.  

Looks good for powerpc64-unkown-linux-gnu:

  http://gcc.gnu.org/ml/gcc-testresults/2004-08/msg01187.html

Janis