This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: PATCH RFC: My version of the lower-subreg patch
- From: Rask Ingemann Lambertsen <rask at sygehus dot dk>
- To: Ian Lance Taylor <iant at google dot com>
- Cc: gcc-patches at gcc dot gnu dot org, rth at redhat dot com
- Date: Sun, 17 Dec 2006 22:02:09 +0100
- Subject: Re: PATCH RFC: My version of the lower-subreg patch
- References: <m3wt4segt6.fsf@localhost.localdomain>
On Fri, Dec 15, 2006 at 06:05:25PM -0800, Ian Lance Taylor wrote:
> Either way, we generate better code for C code using 'long long'
> values on 32-bit systems.
Not to mention 16-bit and 8-bit systems, which are suffering already at
'int' and 'long'.
> As I say, this pass is not ready to check in. The lower-subreg pass
> is run twice: once early in the RTL passes, and once after instruction
> splitting. This gives the best results: the first pass permits CSE on
> operands of & and | which are fully decomposable, and the second pass
> handles operands of + and - which we do not want to decompose before
> CSE and loop optimizations. But the pass is too slow to run twice.
Running it again after combine also means that the subregs which combine
reinserts into the insns can be lowered, doesn't it?
--
Rask Ingemann Lambertsen