This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [patch] MIN/MAX_EXPR transform in phiopts
- From: Andrew Pinski <pinskia at physics dot uc dot edu>
- To: Zdenek Dvorak <rakdver at atrey dot karlin dot mff dot cuni dot cz>
- Cc: gcc-patches at gcc dot gnu dot org, stevenb at suse dot de
- Date: Fri, 11 Mar 2005 09:20:59 -0500
- Subject: Re: [patch] MIN/MAX_EXPR transform in phiopts
- References: <20050311085213.GA7476@atrey.karlin.mff.cuni.cz>
On Mar 11, 2005, at 3:52 AM, Zdenek Dvorak wrote:
1) To handle nested conditionals (like in the later transforms),
phiopts
pass the blocks using FOR_EACH_BB_REVERSE, so that the inner
conditions are tested and collapsed first. This is not reliable,
since the order of the blocks in the list may be arbitrary.
The patch forces a correct ordering -- the one in that a basic
block with a single predecessor is guaranteed to be processed
before the predecessor.
Zdenek thanks for doing this part. When I wrote that part of the code
I had forgot about the blocks can be in arbitrary order.
-- Pinski