public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
From: "sebastian dot pop at cri dot ensmp dot fr" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/18595] [4.0 Regression] IV-OPTS is O(N^3)
Date: Thu, 27 Jan 2005 15:12:00 -0000	[thread overview]
Message-ID: <20050127151236.3190.qmail@sourceware.org> (raw)
In-Reply-To: <20041121145949.18595.pinskia@gcc.gnu.org>


------- Additional Comments From sebastian dot pop at cri dot ensmp dot fr  2005-01-27 15:12 -------
Subject: Re:  [4.0 Regression] IV-OPTS is O(N^3)

rakdver at atrey dot karlin dot mff dot cuni dot cz wrote:
> the patch is below (in stronger form -- only removing entries that
> contain invalidated symbols), and indeed helps with this testcase.

Many thanks.

> It however causes measurable slow down on normal code (see analysis
> in one of the previous mails).
> 

I see. 

Another idea: would it be possible to insert the invalidated names
during the optimization pass instead of invalidating all the symbols?
This would avoid the first pass over the scev database that search for
symbols.

> Note that your patch is not entirely correct -- in case loop is
> unrolled, its number of iterations is changed, so in this case
> scev_reset should clear its number of iterations unconditionally

or the unroller should do that work on the unrolled loops?



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18595


  parent reply	other threads:[~2005-01-27 15:12 UTC|newest]

Thread overview: 59+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-11-21 14:59 [Bug tree-optimization/18595] New: IV-OPTS is slow (and does not scale) pinskia at gcc dot gnu dot org
2004-11-24 11:15 ` [Bug tree-optimization/18595] " steven at gcc dot gnu dot org
2004-11-25 11:16 ` [Bug tree-optimization/18595] [4.0 Regression] IV-OPTS is O(N^3) belyshev at lubercy dot com
2004-11-27  1:11 ` steven at gcc dot gnu dot org
2004-11-27  1:45 ` belyshev at lubercy dot com
2004-11-27 21:17 ` neroden at gcc dot gnu dot org
2004-11-28 18:43 ` rakdver at gcc dot gnu dot org
2004-12-22 23:40 ` belyshev at depni dot sinp dot msu dot ru
2004-12-23  1:26 ` steven at gcc dot gnu dot org
2004-12-23  1:29 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-23 14:18 ` steven at gcc dot gnu dot org
2005-01-23 15:01   ` Daniel Berlin
2005-01-23 14:22 ` steven at gcc dot gnu dot org
2005-01-23 14:23 ` steven at gcc dot gnu dot org
2005-01-23 14:24 ` steven at gcc dot gnu dot org
2005-01-23 15:02 ` dberlin at dberlin dot org
2005-01-23 15:06 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-23 16:38 ` rakdver at gcc dot gnu dot org
2005-01-23 17:04 ` rakdver at gcc dot gnu dot org
2005-01-23 17:14 ` rakdver at gcc dot gnu dot org
2005-01-24  1:40 ` rakdver at gcc dot gnu dot org
2005-01-24  1:46 ` rakdver at gcc dot gnu dot org
2005-01-24  1:49   ` Daniel Berlin
2005-01-24  1:49 ` dberlin at dberlin dot org
2005-01-24  1:58 ` dberlin at dberlin dot org
2005-01-24 21:36 ` sebastian dot pop at cri dot ensmp dot fr
2005-01-24 22:17 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-24 22:25 ` dberlin at dberlin dot org
2005-01-24 23:09 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-24 23:12 ` dberlin at dberlin dot org
2005-01-25  0:27 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-25  0:39 ` dberlin at dberlin dot org
2005-01-25  0:50 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-25  0:50 ` dberlin at dberlin dot org
2005-01-25  0:52 ` dberlin at dberlin dot org
2005-01-25  1:11 ` rakdver at gcc dot gnu dot org
2005-01-25  1:15 ` dberlin at dberlin dot org
2005-01-25  1:24 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-25  1:28 ` steven at gcc dot gnu dot org
2005-01-25  1:42 ` dberlin at dberlin dot org
2005-01-25  1:52 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-25  1:57 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-25 10:32 ` sebastian dot pop at cri dot ensmp dot fr
2005-01-25 10:41 ` sebastian dot pop at cri dot ensmp dot fr
2005-01-25 11:02 ` sebastian dot pop at cri dot ensmp dot fr
2005-01-25 11:15 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-25 11:16 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-25 11:29 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-25 12:03 ` sebastian dot pop at cri dot ensmp dot fr
2005-01-25 12:44 ` sebastian dot pop at cri dot ensmp dot fr
2005-01-25 15:54 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-25 16:27 ` sebastian dot pop at cri dot ensmp dot fr
2005-01-25 19:16 ` pinskia at gcc dot gnu dot org
2005-01-27 13:18 ` sebastian dot pop at cri dot ensmp dot fr
2005-01-27 15:00 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-27 15:12 ` sebastian dot pop at cri dot ensmp dot fr [this message]
2005-01-27 19:11 ` rakdver at atrey dot karlin dot mff dot cuni dot cz
2005-01-27 20:40 ` sebastian dot pop at cri dot ensmp dot fr
2005-02-14 11:12 ` pinskia at gcc dot gnu dot org

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20050127151236.3190.qmail@sourceware.org \
    --to=gcc-bugzilla@gcc.gnu.org \
    --cc=gcc-bugs@gcc.gnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).