Den 2010-11-16 12:15 skrev Stefano Lattarini: > On Tuesday 16 November 2010, Peter Rosin wrote: >> Den 2010-11-15 23:23 skrev Ralf Wildenhues: >>> Hi Peter, >>> >>> * Peter Rosin wrote on Mon, Nov 15, 2010 at 09:19:44AM CET: >>>> This is based off of master (but I guess it also applies to tests-init) >>>> but the backport to the msvc branch is trivial. I "stole" the xsi-test >>>> from libtool. >>>> >>>> Would be good if someone tested on that Alpha before pushing I suppose. >>> >>> Patch is ok if we also have test coverage of 'compile' in non-MSVC mode >>> on this system (so we can be sure that the shell doesn't barf skipping >>> over the XSI code bits). You can omit dropping the output as you like. >>> Thanks for the quick feedback! >> >> compile.test and compile2.test still covers the original non-msvc stuff, >> and they PASS according to your log. >> >> Now I wonder where I should commit this and how it should find its way >> to master. If I commit to tests-init, we need to backport to msvc and >> maint before 1.11.2 or we'll face failing tests there. >> >> Should I first commit to msvc, then merge and fixup the defs.in->defs >> change when going to tests-init and then merge to master? > I think you won't need to fixup the defs.in->defs renaming; git should > be smart enough to handle this automatically (it does so in the manual > checks I've performed). So IMHO you should apply the patch to msvc, > and then merge msvc to tests-init.
I committed to msvc, merged into tests-init and merged tests-init into master. And pushed. The rename was handled just fine as you said, but there was one trivial conflict caused by the churn in the tests-init branch. Cheers, Peter
