This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [patch] Fix PR34017: ICE in lambda_loopnest_to_gcc_loopnest
- From: "Richard Guenther" <richard dot guenther at gmail dot com>
- To: "Sebastian Pop" <sebpop at gmail dot com>
- Cc: "Daniel Berlin" <dberlin at dberlin dot org>, "GCC Patches" <gcc-patches at gcc dot gnu dot org>
- Date: Wed, 9 Jan 2008 10:57:36 +0100
- Subject: Re: [patch] Fix PR34017: ICE in lambda_loopnest_to_gcc_loopnest
- References: <cb9d34b20801082217k3c344f17l2c87ae4360320c33@mail.gmail.com>
On Jan 9, 2008 7:17 AM, Sebastian Pop <sebpop@gmail.com> wrote:
> Hi,
>
> We used to abort when the expression to be rewritten was a PHI_NODE,
> but as Jakub remarked in the PR comments, there is no reason we cannot
> transform the PHI_NODE expressions. The attached patch is fixing this
> bug by implementing the translation of PHI_NODEs in function of the
> new induction variables.
>
> The patch passed tree-ssa.exp, and I'm going to regstrap it tonight.
> Okay for trunk after regtest?
This looks good, though the two functions share lots of things and their
interface is not exactly clear (the documentation doesn't mention any
of the params). Maybe folding them back into their caller makes the
control flow easier to read.
Thanks,
Richard.
> Sebastian
> --
> AMD - GNU Tools
>