Boost Testing :
Subject: Re: [Boost-testing] [random] Sandia-itanium intel-linux-11.1 failures
From: Jim Bell (Jim_at_[hidden])
Date: 2010-03-14 09:00:54
> I'm just seeing things like this with no information about what failed.
> Run [2010-03-12 09:03:53 UTC]: fail
> Running 1 test case...
> Basic tests for hellekalek1995
> ; seeding uses 1 words
Pure speculation: could it be that some tests aren't explicitly returning 0
from main() on success? If so, the test passes, but the process' return code
is uninitialized. Bjam takes what it finds there and labels it failed. So it
may work on some platforms and fail on others.