public inbox for gcc-bugs@sourceware.org help / color / mirror / Atom feed
From: "dushistov at mail dot ru" <gcc-bugzilla@gcc.gnu.org> To: gcc-bugs@gcc.gnu.org Subject: [Bug tree-optimization/14741] graphite with loop blocking and interchanging doesn't optimize a matrix multiplication loop Date: Sat, 27 Jul 2013 15:19:00 -0000 [thread overview] Message-ID: <bug-14741-4-xf3oI6nbAc@http.gcc.gnu.org/bugzilla/> (raw) In-Reply-To: <bug-14741-4@http.gcc.gnu.org/bugzilla/> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14741 --- Comment #26 from Evgeniy Dushistov <dushistov at mail dot ru> --- I try such simple C++ function, compiled in separate object file(-march=native -Ofast): void mult(const double * const __restrict__ A, const double * const __restrict__ B, double * const __restrict__ C, const size_t N) { for (size_t j = 0; j < N; ++j) for (size_t i = 0; i < N; ++i) for (size_t k = 0; k < N; ++k) C[i * N + j] += A[i * N + k] + B[k * N + j]; } $ time ./test_gcc 204.800000 real 0m9.628s user 0m9.620s sys 0m0.000s $ time ./test_icc 204.800000 real 0m0.637s user 0m0.630s sys 0m0.000s Difference 15.2 times Looks like the difference here: GCC: Analyzing loop at mult.cpp:5 Analyzing loop at mult.cpp:6 Analyzing loop at mult.cpp:7 mult.cpp:3: note: vectorized 0 loops in function. ICC: mult.cpp(5): (col. 2) remark: PERMUTED LOOP WAS VECTORIZED. mult.cpp(5): (col. 2) remark: PERMUTED LOOP WAS VECTORIZED. mult.cpp(5): (col. 2) remark: PERMUTED LOOP WAS VECTORIZED.
next prev parent reply other threads:[~2013-07-27 15:19 UTC|newest] Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top [not found] <bug-14741-4@http.gcc.gnu.org/bugzilla/> 2013-03-29 11:03 ` Joost.VandeVondele at mat dot ethz.ch 2013-07-14 7:11 ` spop at gcc dot gnu.org 2013-07-16 2:25 ` spop at gcc dot gnu.org 2013-07-16 2:29 ` spop at gcc dot gnu.org 2013-07-21 9:15 ` spop at gcc dot gnu.org 2013-07-21 9:59 ` spop at gcc dot gnu.org 2013-07-21 10:14 ` spop at gcc dot gnu.org 2013-07-21 10:34 ` spop at gcc dot gnu.org 2013-07-21 22:01 ` spop at gcc dot gnu.org 2013-07-27 15:19 ` dushistov at mail dot ru [this message] 2013-07-27 15:22 ` dushistov at mail dot ru 2013-07-29 22:16 ` spop at gcc dot gnu.org 2014-03-13 15:03 ` spop at gcc dot gnu.org 2014-04-29 12:29 ` dominiq at lps dot ens.fr 2015-05-16 8:48 ` tkoenig at gcc dot gnu.org 2015-05-18 6:28 ` Joost.VandeVondele at mat dot ethz.ch 2015-09-11 9:26 ` manfred99 at gmx dot ch 2015-09-11 19:27 ` spop at gcc dot gnu.org 2015-09-11 21:55 ` dominiq at lps dot ens.fr 2015-09-12 12:06 ` dominiq at lps dot ens.fr 2015-09-12 12:19 ` spop 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-14741-4-xf3oI6nbAc@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).