Openembedded Devel Discussions
 help / color / mirror / Atom feed
* Bad gcc packaging (libgcc-s1 instead of libgcc1)
@ 2007-03-15 11:00 Rod Whitby
  2007-03-15 11:52 ` Koen Kooi
  0 siblings, 1 reply; 3+ messages in thread
From: Rod Whitby @ 2007-03-15 11:00 UTC (permalink / raw)
  To: openembedded-devel

I'm getting a libgcc-s1 package emitted by gcc_4.1.1.bb instead of the
normal libgcc1.  This is poisoning all builds that depend on libgcc1,
and is polluting the slugos feeds with bad dependencies.

I believe it is caused by the patch in
http://lists.linuxtogo.org/pipermail/openembedded-commits/2007-March/004007.html

I have tested reverting that patch locally, and the libgcc-s1 package
was not emitted.

So I wish to revert that patch.

I'm looking for confirmation that this problem is happening to others.
I will revert the patch locally on our autobuilder, and wait for
confirmation before committing that reversion.

-- Rod



^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: Bad gcc packaging (libgcc-s1 instead of libgcc1)
  2007-03-15 11:00 Bad gcc packaging (libgcc-s1 instead of libgcc1) Rod Whitby
@ 2007-03-15 11:52 ` Koen Kooi
  2007-03-15 16:57   ` Rod Whitby
  0 siblings, 1 reply; 3+ messages in thread
From: Koen Kooi @ 2007-03-15 11:52 UTC (permalink / raw)
  To: openembedded-devel

Quoting Rod Whitby <rod@whitby.id.au>:

> I'm getting a libgcc-s1 package emitted by gcc_4.1.1.bb instead of the
> normal libgcc1.

The situation is a bit more nuanced: gcc (non-cross) never emitted libgcc1,
gcc-cross did. So gcc_4.bb was not packaging its libgcc as it should have.


>  This is poisoning all builds that depend on libgcc1,
> and is polluting the slugos feeds with bad dependencies.

That is correct and indeed very annoying.

> I believe it is caused by the patch in
>
http://lists.linuxtogo.org/pipermail/openembedded-commits/2007-March/004007.html
>
> I have tested reverting that patch locally, and the libgcc-s1 package
> was not emitted.
>
> So I wish to revert that patch.

Please properly fix it by having gcc emit libgcc1 instead, otherwise your native
gcc package won't work properly if gcc-cross was a different version (which is
not a common case, but still). I think putting a construction like
gcc-package-cross.inc has:

python do_package() {
        if bb.data.getVar('DEBIAN_NAMES', d, 1):
                bb.data.setVar('PKG_libgcc', 'libgcc1', d)
        bb.build.exec_func('package_do_package', d)
}

in gcc-package.inc should work. Could you test that?

Sorry for breaking everyones packages.

regards,

Koen


----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.



^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: Bad gcc packaging (libgcc-s1 instead of libgcc1)
  2007-03-15 11:52 ` Koen Kooi
@ 2007-03-15 16:57   ` Rod Whitby
  0 siblings, 0 replies; 3+ messages in thread
From: Rod Whitby @ 2007-03-15 16:57 UTC (permalink / raw)
  To: openembedded-devel

Koen Kooi wrote:
> I think putting a construction like
> gcc-package-cross.inc has:
> 
> python do_package() {
>         if bb.data.getVar('DEBIAN_NAMES', d, 1):
>                 bb.data.setVar('PKG_libgcc', 'libgcc1', d)
>         bb.build.exec_func('package_do_package', d)
> }
> 
> in gcc-package.inc should work. Could you test that?

That works here.

> Sorry for breaking everyones packages.

Apology accepted.

-- Rod



^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2007-03-15 16:58 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2007-03-15 11:00 Bad gcc packaging (libgcc-s1 instead of libgcc1) Rod Whitby
2007-03-15 11:52 ` Koen Kooi
2007-03-15 16:57   ` Rod Whitby

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox