public inbox for gcc-bugs@sourceware.org help / color / mirror / Atom feed
From: "juzhe.zhong at rivai dot ai" <gcc-bugzilla@gcc.gnu.org> To: gcc-bugs@gcc.gnu.org Subject: [Bug tree-optimization/109088] GCC does not always vectorize conditional reduction Date: Wed, 27 Sep 2023 14:11:09 +0000 [thread overview] Message-ID: <bug-109088-4-o4Tns2gwuJ@http.gcc.gnu.org/bugzilla/> (raw) In-Reply-To: <bug-109088-4@http.gcc.gnu.org/bugzilla/> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=109088 --- Comment #13 from JuzheZhong <juzhe.zhong at rivai dot ai> --- Hi, Richi. This is my draft approach to enhance the finding more potential condtional reduction. diff --git a/gcc/tree-if-conv.cc b/gcc/tree-if-conv.cc index a8c915913ae..c25d2038f16 100644 --- a/gcc/tree-if-conv.cc +++ b/gcc/tree-if-conv.cc @@ -1790,8 +1790,72 @@ is_cond_scalar_reduction (gimple *phi, gimple **reduc, tree arg_0, tree arg_1, std::swap (r_op1, r_op2); std::swap (r_nop1, r_nop2); } - else if (r_nop1 != PHI_RESULT (header_phi)) - return false; + else if (r_nop1 == PHI_RESULT (header_phi)) + ; + else + { + /* Analyze the statement chain of STMT so that we could teach generate + better if-converison code sequence. We are trying to catch this + following situation: + + loop-header: + reduc_1 = PHI <..., reduc_2> + ... + if (...) + tmp1 = reduc_1 + rhs1; + tmp2 = tmp1 + rhs2; + tmp3 = tmp2 + rhs3; + ... + reduc_3 = tmpN-1 + rhsN-1; + + reduc_2 = PHI <reduc_1, reduc_3> + + and convert to + + reduc_2 = PHI <0, reduc_1> + tmp1 = rhs1 + rhs2; + tmp2 = tmp1 + rhs3; + tmp3 = tmp2 + rhs4; + ... + tmpN-1 = tmpN-2 + rhsN; + ifcvt = cond_expr ? tmpN-1 : 0 + reduc_1 = tmpN-1 +/- ifcvt; */ + if (num_imm_uses (PHI_RESULT (header_phi)) != 2) + return false; + FOR_EACH_IMM_USE_FAST (use_p, imm_iter, PHI_RESULT (header_phi)) + { + gimple *use_stmt = USE_STMT (use_p); + if (is_gimple_assign (use_stmt)) + { + if (gimple_assign_rhs_code (use_stmt) != reduction_op) + return false; + if (TREE_CODE (gimple_assign_lhs (use_stmt)) != SSA_NAME) + return false; + + bool visited_p = false; + while (!visited_p) + { + use_operand_p use; + if (!single_imm_use (gimple_assign_lhs (use_stmt), &use, + &use_stmt) + || gimple_bb (use_stmt) != gimple_bb (stmt) + || !is_gimple_assign (use_stmt) + || TREE_CODE (gimple_assign_lhs (use_stmt)) != SSA_NAME + || gimple_assign_rhs_code (use_stmt) != reduction_op) + return false; + + if (gimple_assign_lhs (use_stmt) == gimple_assign_lhs (stmt)) + { + r_op2 = r_op1; + r_op1 = PHI_RESULT (header_phi); + visited_p = true; + } + } + } + else if (use_stmt != phi) + return false; + } + } My approach is doing the check as follows: tmp1 = reduc_1 + rhs1; tmp2 = tmp1 + rhs2; tmp3 = tmp2 + rhs3; ... reduc_3 = tmpN-1 + rhsN-1; Start the iteration check from "tmp1 = reduc_1 + rhs1;" until "reduc_3 = tmpN-1 + rhsN-1;" Make sure each statement are PLUS_EXPR for reduction sum. Does it look reasonable ? It succeed on vectorization.
next prev parent reply other threads:[~2023-09-27 14:11 UTC|newest] Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top 2023-03-10 9:24 [Bug c/109088] New: GCC fail auto-vectorization juzhe.zhong at rivai dot ai 2023-03-10 10:39 ` [Bug c/109088] " ubizjak at gmail dot com 2023-03-10 12:39 ` [Bug tree-optimization/109088] " rguenth at gcc dot gnu.org 2023-03-10 13:16 ` juzhe.zhong at rivai dot ai 2023-03-10 14:04 ` pinskia at gcc dot gnu.org 2023-03-10 14:09 ` [Bug tree-optimization/109088] GCC does not always vectorize conditional reduction pinskia at gcc dot gnu.org 2023-09-26 12:14 ` juzhe.zhong at rivai dot ai 2023-09-27 2:45 ` juzhe.zhong at rivai dot ai 2023-09-27 2:58 ` juzhe.zhong at rivai dot ai 2023-09-27 7:15 ` rguenth at gcc dot gnu.org 2023-09-27 7:34 ` juzhe.zhong at rivai dot ai 2023-09-27 9:06 ` rguenth at gcc dot gnu.org 2023-09-27 9:27 ` juzhe.zhong at rivai dot ai 2023-09-27 14:11 ` juzhe.zhong at rivai dot ai [this message] 2023-10-06 9:44 ` rguenth at gcc dot gnu.org 2023-11-10 12:02 ` juzhe.zhong at rivai dot ai 2023-11-10 13:10 ` rguenth at gcc dot gnu.org 2023-11-10 13:42 ` juzhe.zhong at rivai dot ai 2023-11-15 14:09 ` rguenth at gcc dot gnu.org 2023-11-15 14:38 ` juzhe.zhong at rivai dot ai 2023-11-15 14:42 ` rguenther at suse dot de 2023-11-16 1:06 ` juzhe.zhong at rivai dot ai 2023-11-16 6:50 ` rguenther at suse dot de
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-109088-4-o4Tns2gwuJ@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).