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] Improve EDGE_ABNORMAL construction (PR middle-end/59917, tree-optimization/59920)

From: Jeff Law <law at redhat dot com>
To: Richard Biener <rguenther at suse dot de>, Jakub Jelinek <jakub at redhat dot com>
Cc: gcc-patches at gcc dot gnu dot org
Date: Wed, 29 Jan 2014 09:42:49 -0700
Subject: Re: [PATCH] Improve EDGE_ABNORMAL construction (PR middle-end/59917, tree-optimization/59920)
Authentication-results: sourceware.org; auth=none
References: <20140128224424 dot GQ892 at tucnak dot redhat dot com> <alpine dot LSU dot 2 dot 11 dot 1401291041380 dot 469 at zhemvz dot fhfr dot qr>

On 01/29/14 02:47, Richard Biener wrote:


I wonder if you can't simply prune the edges for the OpenMP regions
(similar as to how we could prune some of the edges from setjmp
to calls before that setjmp call)?  Eventually this asks for delaying
of abnormal edge creation (but that's what you do anyway) to be able
to compute dominators between "regular" CFG creation and abnormal
edge creation.

I'd been pondering this after I looked at the code we generated for59919. The question in my mind is how many edges can we reallyavoid/prune in real world code. I meant to open a PR for this so thatwe can track it in the next stage1 cycle -- I'll do that now ;-)

For 59917/59920 we really have to avoid generating the edges in thefirst place though.


jeff

References:
- [PATCH] Improve EDGE_ABNORMAL construction (PR middle-end/59917, tree-optimization/59920)
  - From: Jakub Jelinek
- Re: [PATCH] Improve EDGE_ABNORMAL construction (PR middle-end/59917, tree-optimization/59920)
  - From: Richard Biener

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