[libcxx] r220512 - [libcxx] XFAIL all currently failing libc++ tests for linux.

[Eric Fiselier]

This is a different issue. The XFAIL statements now work but there are two
pesky tests that only fail on some Linux distributions. I'll take a look
tonight but this is a different and known issue.

/Eric
On Oct 28, 2014 4:12 AM, "İsmail Dönmez" <ismail at donmez.ws> wrote:

Hi,

Still fails:

https://build.opensuse.org/build/home:namtrac:clang/openSUSE_Factory/x86_64/llvm-clang-head/_log

On Tue, Oct 28, 2014 at 12:25 AM, Eric Fiselier <eric at efcs.ca> wrote:

> This should be fixed in r220724. Let me know if you run into any other
> problems.
>
> /Eric
>
> On Mon, Oct 27, 2014 at 3:39 PM, Eric Fiselier <eric at efcs.ca> wrote:
>
>> For libc++'s test suite I think I'll just check for `redhat-linux` and
>> `suse-linux` and append the `-gnu` manually.
>> I imagine this is not correct in all cases, but it seems like the best
>> solution for the test suite.
>>
>> Another approach may be to probe for glibc but that would be more complex.
>>
>> /Eric
>>
>> On Mon, Oct 27, 2014 at 1:27 AM, İsmail Dönmez <ismail at donmez.ws> wrote:
>>
>>> Hi,
>>>
>>> On Mon, Oct 27, 2014 at 9:22 AM, Chandler Carruth <chandlerc at google.com>
>>> wrote:
>>>
>>>> The triple specified is 'x86_64-suse-linux'... If you're using glibc,
>>>> then it seems reasonable to use 'linux-gnu' as the last component of the
>>>> triple? It seems a real shame to add a vendor here just to catch the case
>>>> that your system uses glibc but deviates from every other linux triple
>>>> using glibc in this way.... =[
>>>>
>>>
>>> This matches the gcc's default target hence its like that for a long
>>> time :/ Btw. I just checked and Centos/RHEL uses x86_64-redhat-linux so we
>>> are not alone in this.
>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20141028/61aa3f8d/attachment.html>