backport plugin fixes to 2.21

backport plugin fixes to 2.21

[Alan Modra]

I've backported my plugin fixes to the branch, along with some from
HJ and Dave.  I also put Mark Wielaard's PR 10549 fix on the branch,
with the exception of the ARM patch.  I left ARM out since the branch
lacks ifunc support for ARM, and I didn't want to make the call that
ifunc was safe/desirable for ARM on the branch.  That's not to say
that I think it would be unsafe, just that I don't know.

As far as I'm aware, this means both mainline and 2.21 now have full
support for LTO, and neither needs gcc pass-through-libs.

-- 
Alan Modra
Australia Development Lab, IBM

Re: backport plugin fixes to 2.21

[Tristan Gingold]

On Apr 27, 2011, at 11:20 AM, Alan Modra wrote:

> I've backported my plugin fixes to the branch, along with some from
> HJ and Dave.  I also put Mark Wielaard's PR 10549 fix on the branch,
> with the exception of the ARM patch.  I left ARM out since the branch
> lacks ifunc support for ARM, and I didn't want to make the call that
> ifunc was safe/desirable for ARM on the branch.  That's not to say
> that I think it would be unsafe, just that I don't know.
> 
> As far as I'm aware, this means both mainline and 2.21 now have full
> support for LTO, and neither needs gcc pass-through-libs.

Dave,

according to your latest message about 2.21.1 and plugins:

> 
>  There's still one outstanding LTO issue that needs solving on HEAD and then
> a backport of all the plugin-api updates to the branch.  The one outstanding
> issue is the pr12365/2-stage link discussion which I'm about to revive, having
> spent the past little while trying a few different approaches.
> 
>    cheers,
>      DaveK

some of this work has been completed by Alan (pr12365 is closed).

Do you expect more backports to the release branch ?

Tristan.

Re: backport plugin fixes to 2.21

[Richard Sandiford]

Alan Modra <amodra@gmail.com> writes:
> I've backported my plugin fixes to the branch, along with some from
> HJ and Dave.  I also put Mark Wielaard's PR 10549 fix on the branch,
> with the exception of the ARM patch.  I left ARM out since the branch
> lacks ifunc support for ARM, and I didn't want to make the call that
> ifunc was safe/desirable for ARM on the branch.  That's not to say
> that I think it would be unsafe, just that I don't know.

Yeah, the ifunc patches were quite invasive, so I think it'd be
better to leave them out.

Richard