|
Boost : |
From: Gennaro Prota (gennaro_prota_at_[hidden])
Date: 2006-06-24 15:58:44
On Sat, 24 Jun 2006 15:03:28 -0400, "Sean Huang"
<huangsean_at_[hidden]> wrote:
>Gennaro,
>Sorry for the late reply.
No problem :) I just thought you were away for the week-end.
>Here is the update. I manually ran all failed tests and they all passed.
Yep.
>The
>only thing I can think of now is I used bjam option --j2. I've just turned
>it off and started another run. The results will show up sometime tomorrow
>(without --j2, the whole things takes about 12hrs). In the mean time, can
>you think of anything that could prevent these tests from running
>simultaneity or along with other tests as in the case when --j2 is
>specified?
You mean that the dynamic_bitset tests are run together with tests of
*other libraries* or that dyn_bitset_unit_tests{1,2,3,4}.exe run
together? The latter shouldn't be a problem.
The only problem I see is if you run together two instances of
dyn_bitset_unit_tests4 *from the same directory*. Of course that's
only possible if you rename one or more of them, or generate 'em with
different names.
>I'm sure if this is the problem of the tests or the regression
>tests do not work well when --j 2 is specified.
Here my science ends and begin Dave's :)
--Gennaro.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk