Please merge the fix for upstream bug 45102

[Bernhard Rosenkraenzer]

Upstream bug 45102

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45102

affects building Android ICS - please backport the fix on the bug report (committed on head) to the Linaro 4.6 compilers.

[Bernhard Rosenkraenzer (berolinux)]

Attaching backported patch

[Ramana Radhakrishnan (ramana)]

On 15 November 2011 11:58, Bernhard Rosenkraenzer
<email address hidden> wrote:
> Attaching backported patch
>
> ** Patch added: "backported patch"
>   https://bugs.launchpad.net/linaro-android/+bug/890657/+attachment/2597638/+files/gcc-bug45102.patch

Can we merge this patch into upstream GCC 4.6 as well if this affects
that so that we don't carry these patches for too long ?

Ramana
>

[Bernhard Rosenkraenzer (berolinux)]

I'm applying the patch manually in the Android toolchain build process for now

[Andrew Stubbs (ams-codesourcery)]

Ramana -- you're the upstream maintainer; why couldn't this applied to upstream 4.6? As far as I know, there's no 'regression fixes only' rule, or anything like that, and this *is* a backport from mainline.

We'd need to confirm that there's a problem in upstream 4.6, but if it's in ours, it's most likely upstream also.

[Ramana Radhakrishnan (ramana)]

On 16 November 2011 14:51, Andrew Stubbs <email address hidden> wrote:
> Ramana -- you're the upstream maintainer; why couldn't this applied to
> upstream 4.6? As far as I know, there's no 'regression fixes only' rule,
> or anything like that, and this *is* a backport from mainline.

If this bug exists in 4.6 then it should be backported upstream after
due approval from the maintainers for that area and the release
managers -

>
> We'd need to confirm that there's a problem in upstream 4.6, but if it's
> in ours, it's most likely upstream also.

Ofcourse, that goes without saying - this particular case is likely to
be a problem upstream as well isn't something for which I can approve
a backport for but there is no reason why it shouldn't make it back
upstream as well. Someone has to do the backport, test it and propose
it upstream. It's just not been done and probably slipped through the
cracks. We can consider tightening our patch review process to be
better in this area but that's not a discussion to be had on this bug
report :) .

Having said that I'm not sure if the patch mentioned above is for the
PR under question.

cheers
Ramana

>
> --
> You received this bug notification because you are a member of Linaro
> Toolchain Developers, which is subscribed to Linaro GCC.
> https://bugs.launchpad.net/bugs/890657
>
> Title:
>  Please merge the fix for upstream bug 45102
>
> Status in Linaro GCC:
>  New
> Status in Linaro Android:
>  Fix Committed
>
> Bug description:
>  Upstream bug 45102
>
>  http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45102
>
>  affects building Android ICS - please backport the fix on the bug
>  report (committed on head) to the Linaro 4.6 compilers.
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/gcc-linaro/+bug/890657/+subscriptions
>

[Michael Hope (michaelh1)]

I've put it on Monday's agenda.

[Andrew Stubbs (ams-codesourcery)]

Bernhard, we've so far failed to reproduce the issue in bugzilla 45102.

Could you please post a reproducible testcase (preprocessed source) along with the compiler configuration and command lines you're using.

Thanks

Andrew

[Michael Collison (michael-collison)]

Unable to reproduce with linaro 4.8 or 4.9 using testcase from Bugzilla.