This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: RFC: Automatically test IFUNC implementations
On Tuesday, September 25, 2012 15:06:14 H.J. Lu wrote:
> On Tue, Sep 25, 2012 at 12:17 PM, H.J. Lu <hjl.tools@gmail.com> wrote:
> > On Tue, Sep 25, 2012 at 12:05 PM, Carlos O'Donell
> >
> > <carlos@systemhalted.org> wrote:
> >> On Tue, Sep 25, 2012 at 2:58 PM, Andreas Jaeger <aj@suse.com>
wrote:
> >>> So, you're building two libcs and one has this just extra symbol?
> >>> In that case I suggest to build only one libc and have the symbol
> >>> exported for everybody with a version in the namespace
> >>> GLIBC_PRIVATE. My fear is that the duplicate building cases us
> >>> more harm than good...
> >>
> >> I agree. We need to ship what we test. I don't want to see two
> >> libc's built.>
> > That was my original patch:
> >
> > http://sourceware.org/ml/libc-alpha/2010-11/msg00039.html
> >
> > But it was rejected. I can go back to the old scheme, but always
> > add __libc_func@GLIBC_PRIVATE when multi-arch is enabled.
>
> I updated hjl/ifunc/test branch to put __libc_func in libc.so
> with version GLIBC_PRIVATE.
Could you send a clean patch, please? You added a patch that removes
changes and thus makes reviewing hard.
I agree with Ulrich's comment from 2010-11 that your patch is the right
way forward, so let's clean this up properly...
I suggest you send one patch that includes the framework and a single
test using it. Once that's in, let's tackle the others...
Andreas
--
Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn,Jennifer Guild,Felix Imendörffer,HRB16746 (AG Nürnberg)
GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126