public inbox for gcc-bugs@sourceware.org help / color / mirror / Atom feed
From: "solar-gcc at openwall dot com" <gcc-bugzilla@gcc.gnu.org> To: gcc-bugs@gcc.gnu.org Subject: [Bug middle-end/51017] GCC 4.6 performance regression (vs. 4.4/4.5) Date: Mon, 16 Feb 2015 01:10:00 -0000 [thread overview] Message-ID: <bug-51017-4-ZrdgGIG58r@http.gcc.gnu.org/bugzilla/> (raw) In-Reply-To: <bug-51017-4@http.gcc.gnu.org/bugzilla/> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=51017 --- Comment #10 from Alexander Peslyak <solar-gcc at openwall dot com> --- I decided to take a look at the generated code. Compared to 4.6.2, GCC 4.9.2 started generating lots of xorps, orps, andps, andnps where it previously generated pxor, por, pand, pandn. Changing those with: sed -i 's/xorps/pxor/g; s/orps/por/g; s/andps/pand/g; s/andnps/pandn/g' made no difference for performance on this machine (still 4.9.2's poor performance). The next suspect were the varieties of MOV instructions. In 4.9.2's generated code, there were 1319 movaps, 721 movups. In 4.6.2's, there were 1258 movaps, 465 movups. Simply changing all movups to movaps in 4.9.2's original code with sed (thus, with no other changes except for this one), resulting in a total of 2040 movaps, brought the performance to levels similar to GCC 4.4 and 4.5's (and is better than 4.6's, but worse than 4.3's). So movups appear to be the main culprit. The same hack for 4.6.2's code brought its performance almost to 4.3's level (still 5% worse, though), and significantly above 4.9.2's (so there's still some other, smaller regression with 4.9.2). Here are my new results: 4.1.0o - 2960K c/s, 28182 bytes, 1758 movaps, 0 movups 4.3.6o - 2956K c/s, 28229 bytes, 1755 movaps, 0 movups 4.4.6o - 2694K c/s, 29316 bytes, 1709 movaps, 7 movups 4.4.6h - 2714K c/s, 29316 bytes, 1716 movaps, 0 movups 4.5.3o - 2709K c/s, 29203 bytes, 1669 movaps, 0 movups 4.6.2o - 2121K c/s, 29624 bytes, 1258 movaps, 465 movups 4.6.2h - 2817K c/s, 29624 bytes, 1723 movaps, 0 movups 4.9.2o - 1852K c/s, 28256 bytes, 1319 movaps, 721 movups 4.9.2h - 2688K c/s, 28256 bytes, 2040 movaps, 0 movups "o" means original, "h" means hacked generated assembly code (all movups changed to movaps). (BTW, there were no movdqa/movdqu in any of these code versions.) Now I am wondering to what extent this is a GCC issue and to what extent it might be my source code's, if GCC is somehow unsure it can assume alignment. What are the conditions when GCC should in fact use movups? Is it intentional that newer versions of GCC are being more careful at this, resulting in worse performance?
next prev parent reply other threads:[~2015-02-16 1:10 UTC|newest] Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top 2011-11-08 0:43 [Bug middle-end/51017] New: " solar-gcc at openwall dot com 2011-11-08 0:57 ` [Bug middle-end/51017] " solar-gcc at openwall dot com 2011-11-08 1:05 ` solar-gcc at openwall dot com 2011-12-15 0:34 ` pinskia at gcc dot gnu.org 2012-01-03 4:46 ` solar-gcc at openwall dot com 2012-01-04 19:39 ` solar-gcc at openwall dot com 2012-01-04 22:43 ` jakub at gcc dot gnu.org 2012-01-04 23:00 ` solar-gcc at openwall dot com 2015-02-09 0:12 ` pinskia at gcc dot gnu.org 2015-02-16 0:08 ` solar-gcc at openwall dot com 2015-02-16 1:10 ` solar-gcc at openwall dot com [this message] 2015-02-16 10:51 ` [Bug tree-optimization/51017] GCC 4.6 performance regression (vs. 4.4/4.5), PRE increases register pressure rguenth at gcc dot gnu.org 2015-02-17 2:21 ` solar-gcc at openwall dot com 2015-02-17 2:56 ` solar-gcc at openwall dot com 2015-02-17 3:11 ` solar-gcc at openwall dot com 2015-02-17 9:25 ` rguenth at gcc dot gnu.org 2015-02-17 9:27 ` rguenth at gcc dot gnu.org 2015-02-18 0:03 ` solar-gcc at openwall dot com 2015-02-18 1:25 ` solar-gcc at openwall dot com 2015-02-18 3:20 ` solar-gcc at openwall dot com 2015-02-18 10:32 ` [Bug tree-optimization/51017] [4.8/4.9/5 Regression] GCC performance regression (vs. 4.4/4.5), PRE increases register pressure too much rguenth at gcc dot gnu.org 2015-02-18 11:09 ` rguenth at gcc dot gnu.org 2015-02-25 14:26 ` law at redhat dot com 2015-06-23 8:14 ` [Bug tree-optimization/51017] [4.8/4.9/5/6 " rguenth at gcc dot gnu.org 2015-06-26 20:04 ` [Bug tree-optimization/51017] [4.9/5/6 " jakub at gcc dot gnu.org 2015-06-26 20:33 ` jakub at gcc dot gnu.org 2021-05-14 9:46 ` [Bug tree-optimization/51017] [9/10/11/12 Regression] GCC performance regression (vs. 4.4/4.5), PRE/LIM increase " jakub at gcc dot gnu.org 2021-06-01 8:05 ` rguenth at gcc dot gnu.org 2022-05-27 9:34 ` [Bug tree-optimization/51017] [10/11/12/13 " rguenth at gcc dot gnu.org 2022-06-28 10:30 ` jakub at gcc dot gnu.org 2023-07-07 10:29 ` [Bug tree-optimization/51017] [11/12/13/14 " rguenth at gcc dot gnu.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=bug-51017-4-ZrdgGIG58r@http.gcc.gnu.org/bugzilla/ \ --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: linkBe 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).