|
Boost Testing : |
From: Stefan Slapeta (stefan_at_[hidden])
Date: 2005-10-28 03:17:32
Stefan Slapeta wrote:
> The test compiles ok if I run it manually. Let's wait for the results of
> tomorrow. I've always seen some of these "backend signals" errors come
> and go without any obvious (deterministic) reason...
>
This is once again related to Intel's buggy debug info generation in the
8.x version. The failure does _only_ occur if you use the cmd line from
the regression test because the source/output directories are quite
different then.
Possible solutions:
- strip debug info for this test (like in some spirit/graph/etc. tests)
- mark as expected
Stefan