This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: [PATCH] Handle degenerated PHI nodes for entry edge in expansion (PR rtl-optimization/47028)

From: Paolo Bonzini <bonzini at gnu dot org>
To: Jakub Jelinek <jakub at redhat dot com>
Cc: gcc-patches at gcc dot gnu dot org
Date: Mon, 03 Jan 2011 13:14:12 +0100
Subject: Re: [PATCH] Handle degenerated PHI nodes for entry edge in expansion (PR rtl-optimization/47028)
References: <20110101183057.GY16156@tyan-ft48-01.lab.bos.redhat.com>

On 01/01/2011 07:30 PM, Jakub Jelinek wrote:

Hi!

On the following testcase we end up with a degenerate PHI node for the entry
edge, which results in the insns being inserted on that edge, which is
wrong, because the insns depend on parameter setup.
The following patch fixes it by inserting those after parm_birth_insn.
Another alternative, mentioned by richi in the PR, would be to get rid of
degenerate PHI nodes during cfg cleanups, haven't tried that yet.

Are you sure this cannot happen in other commit_one_edge_insertion callers?

Paolo

Follow-Ups:
- Re: [PATCH] Handle degenerated PHI nodes for entry edge in expansion (PR rtl-optimization/47028)
  - From: Jakub Jelinek

References:
- [PATCH] Handle degenerated PHI nodes for entry edge in expansion (PR rtl-optimization/47028)
  - From: Jakub Jelinek

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]