This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Simple recompute_reg_usage cleanup
- To: Michael Hayes <m dot hayes at elec dot canterbury dot ac dot nz>
- Subject: Re: Simple recompute_reg_usage cleanup
- From: Jeffrey A Law <law at cygnus dot com>
- Date: Thu, 09 Dec 1999 00:09:47 -0700
- cc: Jan Hubicka <hubicka at atrey dot karlin dot mff dot cuni dot cz>, egcs-patches at egcs dot cygnus dot com, rth at cygnus dot com
- Reply-To: law at cygnus dot com
In message <14407.14547.794346.311149@ongaonga.elec.canterbury.ac.nz>you writ
e:
> FYI, I have been experimenting with a new field for each basic block
> which records the innermost loop that the block is part of. The loop
> depth can then be derived using bb->loop->depth.
>
> I have been trying to use this with the current loop optimiser but am
> having trouble with insns that are not included in any basic block.
Are these real insns or notes that are not included in a basic block?
Presumably this is happening because loop relies on just creating new insns
wherever it wants -- ie it doesn't try to readjust the cfg related info.
It might be worth looking into how hard it would be to revamp loop to use
the edge insertion framework provided by flow.
The patch itself seems fine to me.
jeff