Re: [Patch][BZ 17506] fix tst-strcoll-overflow returning before timeout

[Leonhard Holz <leonhard dot holz at web dot de>]

Am 06.11.2014 10:10, schrieb Siddhesh Poyarekar:
> On Thu, Nov 06, 2014 at 08:52:16AM +0100, Leonhard Holz wrote:
> > Well, the former implementation just relied on running into timeout because
> > of swapping and collating time, but it could not guarantee it. With enough
> > hardware resources (maybe 10GB memory and a high end cpu) the test would
> > have completeted within the five minutes boundary - and then counted as
> > failed.
>
> This test should have succeeded on alarm as well as normal completion,
> but may not be the case for a lot of EXPECTED_SIGNAL tests where a
> normal completion is a failure and the test may not have handled it
> correctly, i.e. returning non-zero finally.  If you can prove that all
> tests that use EXPECTED_SIGNAL handle completion correctly then your
> patch may be acceptable.

A "grep EXPECTED_SIGNAL * -R" reveals that all other tests that define 
EXPECTED_SIGNAL are in nptl and a "grep EXPECTED_SIGNAL * -R -l | xargs 
grep EXPECTED_STATUS" there shows that none of them defines 
EXPECTED_STATUS. So all theses test end up in line 394 of the patched 
test-skeleton.c if they return normaly which is:

      printf ("Expected signal '%s' from child, got none\n",
              strsignal (EXPECTED_SIGNAL));
      exit (1);

Leonhard