Re: [Boost-testing] Question about "--no-warn" optionofcompiler_status tool

From: Beman Dawes (bdawes_at_[hidden])
Date: 2005-07-28 21:14:12

"Voronkov Konstantin" <beowinkle_at_[hidden]> wrote in message 
news:dca6ci$ai7$1_at_sea.gmane.org...
> Beman, thanks for you advice,
> I downloaded and build latest Boost Regression Tests tools,
>  but have problem. I developed 2 tests which run under control
> of Boost Regression Tests: first test should pass and second
> should fail.
>
> With latest build (boost_1_32_0) tests works as expected,
>  but with latest CVS tools version both two tests fail.
>
> Can you help me to detect what is wrong, please?

The way to debug this sort of problem is to inspect test_log.xml in the 
results directory.

The compile, link, and run results should all equal "succeed". For example,

   <run result="succeed" timestamp="2005-07-28 20:31:39 UTC">

If the contents of test_log.xml are wrong, the problem is with 
process_jam_log. If the contents of test_log.xml are correct, the problem is 
with compiler_status.

Please let me know what you find,

--Beman