This is the mail archive of the
mailing list for the binutils project.
Re: [gold commit] PR 19823: Don't allow COPY relocs to protected symbols
- From: Cary Coutant <ccoutant at gmail dot com>
- To: Alan Modra <amodra at gmail dot com>
- Cc: Binutils <binutils at sourceware dot org>
- Date: Tue, 28 Jun 2016 12:48:01 -0700
- Subject: Re: [gold commit] PR 19823: Don't allow COPY relocs to protected symbols
- Authentication-results: sourceware.org; auth=none
- References: <CAJimCsFDm4e4cXdTdJpMdu1KnM3CwD42Mj5wwgJ8ig6=CjKUVw at mail dot gmail dot com> <20160627145549 dot GA3665 at bubble dot grove dot modra dot org> <CAJimCsHto872Y1eJx8y=OBqYBaTbo0d=FtHzvfd7q77L0TAcrQ at mail dot gmail dot com> <20160628005737 dot GB3665 at bubble dot grove dot modra dot org>
>> Is there any reason we shouldn't just skip that test case for targets
>> that don't normally use copy relocs?
> I'd be happy with simply skipping the test. As it is, it is
> particularly nasty since failing that one test stops the majority of
> the other tests running, even with "make -k".
> Like this then?
> * testsuite/Makefile.am (copy_test_protected): Disable for powerpc.
> * testsuite/Makefile.in: Regenerate.
This is OK. Thanks!
After the branch, I think I'll look at making this test intelligently
skip itself if the expected non-pic code sequences aren't there.