+ariel who wrote the original patch On Tue Sep 02 2014 at 1:41:37 AM Justin Bogner <[email protected]> wrote:
> Steffen Prohaska pointed out on cfe-dev that a condition is exactly > backwards in an error_code check in Replaceent::setFromSourceLocation. > > The obvious fix of reversing the condition causes about 30 tests to > fail, so it looks to me like we actually rely on the path *not* being > made absolute. Furthermore, this has been broken like this for at least > a year, and before that it didn't try making the path absolute. > > The attached patch simplifies the code such that it doesn't bother > calling make_absolute. This preserves the current behaviour, passes all > tests, and avoids the case where make_absolute fails and we use the > result anyway. > The original patch where this was introduced is: http://reviews.llvm.org/rL191776 Apparently the test in test/Tooling/clang-check-rel-path.cpp is incorrect (?) Ariel, did you actually see this test break at least once? Cheers, /Manuel > > _______________________________________________ > cfe-commits mailing list > [email protected] > http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits >
_______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
