Hello, Firstly, Kenneth Graunke's patch fixes the autoconf issue and supercedes this patch, so we should focus the review on that patch.
On Tue, Oct 26, 2010 at 06:23:00AM -0700, Dan Nicholson wrote: > However, removing the second -O from PYTHON_FLAGS is really a separate > change. Can you explain that problem in more detail? The second -O flag triggers the same behaviour as passing -OO to the python interpreter. A single -O enables optimizations, but -OO causes doc strings to be stripped from the bytecode in addition to enabling optimization. There's no real reason to do this, except perhaps to save a small amount of memory. This breaks u_format_srgb.py because that script uses its docstring as part of its output, specifically, the copyright notice of the module is copied to the output. If the docstring is stripped, this causes an exception as the __doc__ variable is undefined. > It might also be more reasonable to just drop $(PYTHON_FLAGS) when > using src/gallium/auxiliary/util/u_format_srgb.py so this isn't a > global change. There's really no global effect to dropping the second -O, and it's more consistent to use the same flags for everything. Why introduce a special case? Matt
pgpqxOrM3HD7y.pgp
Description: PGP signature
_______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
