public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Dimitrios Apostolou <jimis@gmx.net>
To: gcc-patches@gcc.gnu.org
Cc: Paolo Bonzini <bonzini@gnu.org>, Steven Bosscher <stevenb.gcc@gmail.com>
Subject: Re: [df-scan.c] Optimise DF_REFs ordering in collection_rec, use HARD_REG_SETs instead of bitmaps
Date: Fri, 08 Jul 2011 03:52:00 -0000	[thread overview]
Message-ID: <alpine.LNX.2.02.1107080631090.1237@localhost.localdomain> (raw)
In-Reply-To: <alpine.LNX.2.02.1107080549070.1237@localhost.localdomain>

To document the gains from the bitmaps, here is (part of) the annotated 
source from callgrind profiler, showing instruction count. Before:


  1,154,400      if (bitmap_bit_p(regs_invalidated_by_call_regset, i)
  8,080,800  => bitmap.c:bitmap_bit_p (192400x)
  1,021,200          && !bitmap_bit_p (&defs_generated, i)
  5,106,000  => bitmap.c:bitmap_bit_p (170200x)
    340,400          && (!is_sibling_call
          .              || !bitmap_bit_p (df->exit_block_uses, i)
          .              || refers_to_regno_p (i, i+1,
          .                                    crtl->return_rtx, NULL)))
  2,053,500        df_ref_record (DF_REF_BASE, collection_rec, regno_reg_rtx[i
35,279,934  => df-scan.c:df_ref_record (170200x)
          .                       NULL, bb, insn_info, DF_REF_REG_DEF,
          .                       DF_REF_MAY_CLOBBER | flags);
          .      }


After:


  1,346,800      if (TEST_HARD_REG_BIT(regs_invalidated_by_call, i)
    510,600          && !TEST_HARD_REG_BIT (defs_generated, i)
    340,400          && (!is_sibling_call
          .              || !bitmap_bit_p (df->exit_block_uses, i)
          .              || refers_to_regno_p (i, i+1,
          .                                    crtl->return_rtx, NULL)))
  2,057,200        df_ref_record (DF_REF_BASE, collection_rec, regno_reg_rtx[i
35,279,934  => df-scan.c:df_ref_record (170200x)
          .                       NULL, bb, insn_info, DF_REF_REG_DEF,
          .                       DF_REF_MAY_CLOBBER | flags);
          .      }



Dimitris

  reply	other threads:[~2011-07-08  3:33 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-07-08  3:33 Dimitrios Apostolou
2011-07-08  3:52 ` Dimitrios Apostolou [this message]
2011-07-08  6:36 ` Dimitrios Apostolou
2011-07-08  7:12   ` Paolo Bonzini
2011-07-09  1:29     ` Dimitrios Apostolou
2011-07-08  6:37 ` Steven Bosscher
2011-07-08  7:04   ` Dimitrios Apostolou
2011-07-08  6:59 ` Jakub Jelinek
2011-07-08  9:09   ` Dimitrios Apostolou
2011-07-08 14:22     ` Paolo Bonzini
2011-07-08 14:25       ` Paolo Bonzini
2011-07-08  8:42 ` Richard Guenther
2011-07-08 10:20   ` Dimitrios Apostolou

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=alpine.LNX.2.02.1107080631090.1237@localhost.localdomain \
    --to=jimis@gmx.net \
    --cc=bonzini@gnu.org \
    --cc=gcc-patches@gcc.gnu.org \
    --cc=stevenb.gcc@gmail.com \
    /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).