This is the mail archive of the gcc@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: [lno] copy header and vectorization

From: Daniel Berlin <dberlin at dberlin dot org>
To: Devang Patel <dpatel at apple dot com>
Cc: "gcc at gcc dot gnu dot org list" <gcc at gcc dot gnu dot org>
Date: Tue, 17 Feb 2004 23:32:20 -0500
Subject: Re: [lno] copy header and vectorization
References: <14410F46-61C9-11D8-B714-000393A91CAA@apple.com>

On Feb 17, 2004, at 11:15 PM, Devang Patel wrote:

After copy header phase, scev determines the initial condition for this loop is i=1 and that causes vectorizer to reject this loop. I am including debug dump after copy header at the end of email.
Is it a known issue ?
--
Devang

Probably. I've sent sebastian several simple loops that are not properly analyzed by analyzer.

He has been working on a new version.

In the meantime, why not clean up the DDG GTY'ing mess, or use loop versioning somewhere?

Follow-Ups:
- Re: [lno] copy header and vectorization
  - From: Devang Patel
- Re: [lno] copy header and vectorization
  - From: Devang Patel

References:
- [lno] copy header and vectorization
  - From: Devang Patel

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