How to test multiple versions of a library?
Currently the Boost.Threads regression tests only test the dynamically linked version of the library. On Win32, at least, there are fairly significant differences between the dynamically linked and statically linked versions (due to thread-specific-storage cleanup), so it would be good to run at least some of the tests for both versions. How do I do this? Mike
Michael Glassford wrote:
Currently the Boost.Threads regression tests only test the dynamically linked version of the library. On Win32, at least, there are fairly significant differences between the dynamically linked and statically linked versions (due to thread-specific-storage cleanup), so it would be good to run at least some of the tests for both versions. How do I do this?
You need to create a second set of tests that link to the <lib> instead of the <dll>. Look at the program_options/test/Jamfile for an example. -- -- Grafik - Don't Assume Anything -- Redshift Software, Inc. - http://redshift-software.com -- rrivera/acm.org - grafik/redshift-software.com - 102708583/icq
Rene Rivera wrote:
Michael Glassford wrote:
Currently the Boost.Threads regression tests only test the dynamically linked version of the library. On Win32, at least, there are fairly significant differences between the dynamically linked and statically linked versions (due to thread-specific-storage cleanup), so it would be good to run at least some of the tests for both versions. How do I do this?
You need to create a second set of tests that link to the <lib> instead of the <dll>. Look at the program_options/test/Jamfile for an example.
I'd appreciate it much if such tests would not be run in two versions on systems for which there are no significant differences between static and dynamically linked versions (the existing tests need >17 hours time on my box (CPU bound). Doubling that is not desirable.) Regards, m
Martin Wille wrote:
I'd appreciate it much if such tests would not be run in two versions on systems for which there are no significant differences between static and dynamically linked versions (the existing tests need >17 hours time on my box (CPU bound). Doubling that is not desirable.)
See use of <build> requirement for how to skip tests: http://article.gmane.org/gmane.comp.lib.boost.devel/106440 Re: Regression Tests Boost.Serialization -- -- Grafik - Don't Assume Anything -- Redshift Software, Inc. - http://redshift-software.com -- rrivera/acm.org - grafik/redshift-software.com - 102708583/icq
On Mon, 02 Aug 2004 09:46:38 -0400, Michael Glassford wrote
Currently the Boost.Threads regression tests only test the dynamically linked version of the library. On Win32, at least, there are fairly significant differences between the dynamically linked and statically linked versions (due to thread-specific-storage cleanup), so it would be good to run at least some of the tests for both versions. How do I do this?
You have to add new entries in the Jamfile with the appropriate compile flags and a new name. Take a look at date_time or regex -- they both do this. Jeff
participants (4)
-
Jeff Garland -
Martin Wille -
Michael Glassford -
Rene Rivera