This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH, pretty-ipa merge 3/4] The IPA-SRA itself.
- From: Jan Hubicka <hubicka at ucw dot cz>
- To: Martin Jambor <mjambor at suse dot cz>
- Cc: GCC Patches <gcc-patches at gcc dot gnu dot org>, Jan Hubicka <hubicka at ucw dot cz>, Richard Guenther <rguenther at suse dot de>
- Date: Sun, 28 Jun 2009 14:14:18 +0200
- Subject: Re: [PATCH, pretty-ipa merge 3/4] The IPA-SRA itself.
- References: <20090624112042.791734971@virgil.suse.cz> <20090624112116.001386402@virgil.suse.cz>
> @@ -128,6 +130,11 @@ struct access
> /* Type. */
> tree type;
>
> + /* The basic block of this access. */
> + basic_block bb;
> + /* The statement this access belongs to. */
> + gimple stmt;
Stmt pointer should be enough, since there is pointer to BB it is
attached to. Also with new aliasing, perhaps we don't need to do this
in two passes anymore? Once you see access, you can immediately validate
it.
> +static bool
> +find_param_candidates (void)
Perhaps some comments on why those are candidates on invididual
conditionals would help.
I am still bit confused by code checking whether BB must be executed via
dominators, I will look into it tomorrow.
Honza