This is the mail archive of the
libc-help@sourceware.org
mailing list for the glibc project.
Re: TEST_DIRECT flag and nptl eintr/stdio tests
Just FYI: I pulled the latest code and tested. Your patch, Florian,
seems to fix the problem that I was seeing. I am going to send along
my patch to test-skeleton.c to see if you think that would be
complementary to the change that you made.
Thanks again for your reply!
Will
On Sun, Aug 28, 2016 at 10:27 AM, Will Hawkins <whh8b@virginia.edu> wrote:
> Florian,
>
> What timing! This looks like exactly the issue I was dealing with. I
> corrected the problem in a slightly different manner. I will post the
> patch tomorrow to see if it's something that you'd like to incorporate
> in case there are future tests that use SIGALRM. I don't necessarily
> know if my fix is more general or too specific. I will let experts
> decide!
>
> Thanks, as ever, for the work that you do and for your response to my message!
>
> Will
>
> On Fri, Aug 26, 2016 at 12:37 PM, Florian Weimer <fweimer@redhat.com> wrote:
>> On 08/25/2016 08:23 PM, Will Hawkins wrote:
>>
>>> Second, I have a question about the interaction between TEST_DIRECT
>>> flag and certain tests in nptl. In particular, I am having trouble
>>> with eintr* and stdio1. When I run those tests without the TEST_DIRECT
>>> flag set, everything seems to run fine. However, when I run those
>>> tests with the TEST_DIRECT flag set, they "do not complete".
>>>
>>> I am testing against 6f9d4f595e4073807ad0e844cbb3b3d7158b76d5 from
>>> August 17. I am more than happy to investigate the problem on my own
>>> (this is what I love to do) but I wanted to first reach out and see if
>>> there was some known problem.
>>
>>
>> Last week, I posted a patch which should fix this:
>>
>> <https://sourceware.org/ml/libc-alpha/2016-08/msg00500.html>
>>
>> Florian