Sorry I didn't explain that well. The -fcolor-diagnostics needs to be able to 
override the simple automatic detection. And actually it uses a different test 
from has_color. See lib/Driver/Tools.cpp:2903. I agree that there is potential 
to clean this up some (eg by storing the ShowColors flag in the raw_ostream), 
but I think that should be separate from this patch.

http://llvm-reviews.chandlerc.com/D127
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits

Reply via email to