From patchwork Mon Oct 16 11:31:08 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "juzhe.zhong@rivai.ai" X-Patchwork-Id: 153349 Return-Path: Delivered-To: ouuuleilei@gmail.com Received: by 2002:a05:612c:2908:b0:403:3b70:6f57 with SMTP id ib8csp3390889vqb; Mon, 16 Oct 2023 04:31:44 -0700 (PDT) X-Google-Smtp-Source: AGHT+IElE74LQakTUUmgTh1lXo2PXRbpeFwT/j95aA7rQAxzN6NVhQPy3RPRkv22scavqMEnyBTc X-Received: by 2002:a05:6808:1826:b0:3ae:e0d:1548 with SMTP id bh38-20020a056808182600b003ae0e0d1548mr39920786oib.15.1697455903793; Mon, 16 Oct 2023 04:31:43 -0700 (PDT) ARC-Seal: i=2; a=rsa-sha256; t=1697455903; cv=pass; d=google.com; s=arc-20160816; b=AG3ClZCkqr8gYC/ZQu0pK0/GIeoO25NtAeKkPn43gzKHE1MBCfxe1/Pri6esNrJ/AG 25Z9w/VLWv39dyrEVjiUdMhZtkXOHBwbblCl3SAZwgaJy5nUS/O6t8xxSJgrTdEeMRm3 m0ZVPE18WeiV7KvuhZbSiE3+LqX82EWIPpXYFKlWK8QIpHJ5ezCbr9yDSep5BWl25kr9 F5NDE6l0VHz9xyVvFtuugthpcsifD+TWw3I78n0tU8cLwuDhpcn29Qx37U7yD4UDQdmJ tnRO71RYViUGq59iAejqL/W+1486KdBVZ6OKXCHMANwdtU/vBXmBQt52PvPZTTZg6Tdv iX8A== ARC-Message-Signature: i=2; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=errors-to:list-subscribe:list-help:list-post:list-archive :list-unsubscribe:list-id:precedence:feedback-id :content-transfer-encoding:mime-version:message-id:date:subject:cc :to:from:arc-filter:dmarc-filter:delivered-to; bh=sauObYDYHZfz7kVTsB9Ka3/GhNOhEFmtU7LUzAG3IiY=; fh=12MRPJmZ1mgDpHqWoogMKqnaGRGM2b7lcuJroqfjJiw=; b=gq9FeThEXgScLe1ho6Wf8P5VPOl6uE6yBVu/9YH1hKh3N0F+k4Qb4BlPBBV8EAQ2NF LDzp/OAK/h8GjXdOEKNGnvuF2frr0ww3fqIvMA8GoYVFIVpt1mDo2mGaTQ2wIRmWvUUe 1VdyDgA6RaHJCIJAsRhWL8AGts10e3Z1Nx9BUtM2VKdcp6axoGqyRSO7o3lJfYOsg07j Ejo80oroIb8opcyaEehoNqzLwjHxEtStNlpgSkvQY7s+3nFi+jkav5dZTAVn+KL08Iik qBSdr21lGcopoJdbMWcTs6Q694bHTHxl2BogIl4NeA/ReonJV+HP1Oo9c5hW5j5v0tat QwsA== ARC-Authentication-Results: i=2; mx.google.com; arc=pass (i=1); spf=pass (google.com: domain of gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org designates 2620:52:3:1:0:246e:9693:128c as permitted sender) smtp.mailfrom="gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org" Received: from server2.sourceware.org (server2.sourceware.org. [2620:52:3:1:0:246e:9693:128c]) by mx.google.com with ESMTPS id p6-20020a05622a00c600b004069477817asi5065292qtw.273.2023.10.16.04.31.43 for (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Mon, 16 Oct 2023 04:31:43 -0700 (PDT) Received-SPF: pass (google.com: domain of gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org designates 2620:52:3:1:0:246e:9693:128c as permitted sender) client-ip=2620:52:3:1:0:246e:9693:128c; Authentication-Results: mx.google.com; arc=pass (i=1); spf=pass (google.com: domain of gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org designates 2620:52:3:1:0:246e:9693:128c as permitted sender) smtp.mailfrom="gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org" Received: from server2.sourceware.org (localhost [IPv6:::1]) by sourceware.org (Postfix) with ESMTP id 8A5EE38582AC for ; Mon, 16 Oct 2023 11:31:43 +0000 (GMT) X-Original-To: gcc-patches@gcc.gnu.org Delivered-To: gcc-patches@gcc.gnu.org Received: from smtpbgjp3.qq.com (smtpbgjp3.qq.com [54.92.39.34]) by sourceware.org (Postfix) with ESMTPS id DF9433858D32 for ; Mon, 16 Oct 2023 11:31:16 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org DF9433858D32 Authentication-Results: sourceware.org; dmarc=none (p=none dis=none) header.from=rivai.ai Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=rivai.ai ARC-Filter: OpenARC Filter v1.0.0 sourceware.org DF9433858D32 Authentication-Results: server2.sourceware.org; arc=none smtp.remote-ip=54.92.39.34 ARC-Seal: i=1; a=rsa-sha256; d=sourceware.org; s=key; t=1697455880; cv=none; b=KOYoy2c6GPDOqNLsTnPQQ7mJCipYlXGKZVMmpv4MRP0SIBLP8HnivhsJ3k8SqbAVfbiVtf0lpO3vjdF4skNk7XTWle+RNego2ozj1MoSMgAis5l5PycJJQ5wTSUQ5a4ei62eOyX5c64IbZI2E98BARWkrXL8IkDlKBgyeZS0mnw= ARC-Message-Signature: i=1; a=rsa-sha256; d=sourceware.org; s=key; t=1697455880; c=relaxed/simple; bh=Q7PQb4oaFrh+7KOcAceGnUy17xWEAf5GPlg1dMYztCo=; h=From:To:Subject:Date:Message-Id:MIME-Version; b=OZ0IrR5LMicmnz7aoAVUW0McHU66k4RSkuyTW/jenw/tPCFS5Sww26vFlK7fzTXprhm95OLXlgszGSCmy6YNX5o79NWAW2L0vW3jQri4GCiAM+NzOMOTQUQHs0FMZtBbmk2XH18hVjFhnSGOgZgt1s+uQjA492fAqnDHGbZlmaY= ARC-Authentication-Results: i=1; server2.sourceware.org X-QQ-mid: bizesmtp64t1697455871t0c8mmoh Received: from rios-cad121.hadoop.rioslab.org ( [58.60.1.9]) by bizesmtp.qq.com (ESMTP) with id ; Mon, 16 Oct 2023 19:31:09 +0800 (CST) X-QQ-SSF: 01400000000000G0V000000A0000000 X-QQ-FEAT: RrZlkntZBfn0W4OwdDADJNB6CLc0hWILSZzfyv+W+sGhYfokNc0VQQef+Oxo/ YSprQKkpOr3vD1h9744US85MrCF/3x7fwYE8dipaukKP7EIrJvaVaBE6poSMxkOC3PKyAl8 ZcwYWFar0QLGUpKFqWn3Hmxl8HPkFitUq8piet4793MyaI57VCUTLdLpYHNu3sDYQokcRsE VTQ3WcJFg5tIA3M8ukf9T2lLPrGScybnFS8zxIw0Ek+9dpS+cozTvzGiTsLWabPZvbt+7eW R6n/7VkOTYBtXf6wyAFBRSXoPGbYoJhRbFTBO5SygXZxGKut08pADsW6C9h+WkAGi1XRL2G HQwP8uJID69MufmLhaEHamRBfGWOLQ8rAeYLlyUi3Rdo3Yh+K9ioUUKKlQr7ez9zOLXYk74 sHbW1/DLRBIEGbeN5U/K1g== X-QQ-GoodBg: 2 X-BIZMAIL-ID: 11213932451443227117 From: Juzhe-Zhong To: gcc-patches@gcc.gnu.org Cc: kito.cheng@gmail.com, kito.cheng@sifive.com, jeffreyalaw@gmail.com, rdapp.gcc@gmail.com, Juzhe-Zhong Subject: [PATCH V2] RISC-V: Fix unexpected big LMUL choosing in dynamic LMUL model for non-adjacent load/store Date: Mon, 16 Oct 2023 19:31:08 +0800 Message-Id: <20231016113108.877163-1-juzhe.zhong@rivai.ai> X-Mailer: git-send-email 2.36.3 MIME-Version: 1.0 X-QQ-SENDSIZE: 520 Feedback-ID: bizesmtp:rivai.ai:qybglogicsvrgz:qybglogicsvrgz7a-one-0 X-Spam-Status: No, score=-11.7 required=5.0 tests=BAYES_00, GIT_PATCH_0, KAM_DMARC_STATUS, RCVD_IN_DNSWL_NONE, RCVD_IN_MSPIKE_H2, SCC_5_SHORT_WORD_LINES, SPF_HELO_PASS, SPF_PASS, TXREP autolearn=ham autolearn_force=no version=3.4.6 X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on server2.sourceware.org X-BeenThere: gcc-patches@gcc.gnu.org X-Mailman-Version: 2.1.30 Precedence: list List-Id: Gcc-patches mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org X-getmail-retrieved-from-mailbox: INBOX X-GMAIL-THRID: 1779882961446320659 X-GMAIL-MSGID: 1779911521740590034 Consider this following case: int bar (int *x, int a, int b, int n) { x = __builtin_assume_aligned (x, __BIGGEST_ALIGNMENT__); int sum1 = 0; int sum2 = 0; for (int i = 0; i < n; ++i) { sum1 += x[2*i] - a; sum1 += x[2*i+1] * b; sum2 += x[2*i] - b; sum2 += x[2*i+1] * a; } return sum1 + sum2; } Before this patch: bar: ble a3,zero,.L5 csrr t0,vlenb csrr a6,vlenb slli t1,t0,3 vsetvli a5,zero,e32,m4,ta,ma sub sp,sp,t1 vid.v v20 vmv.v.x v12,a1 vand.vi v4,v20,1 vmv.v.x v16,a2 vmseq.vi v4,v4,1 slli t3,a6,2 vsetvli zero,a5,e32,m4,ta,ma vmv1r.v v0,v4 viota.m v8,v4 add a7,t3,sp vsetvli a5,zero,e32,m4,ta,mu vand.vi v28,v20,-2 vadd.vi v4,v28,1 vs4r.v v20,0(a7) ----- spill vrgather.vv v24,v12,v8 vrgather.vv v20,v16,v8 vrgather.vv v24,v16,v8,v0.t vrgather.vv v20,v12,v8,v0.t vs4r.v v4,0(sp) ----- spill slli a3,a3,1 addi t4,a6,-1 neg t1,a6 vmv4r.v v0,v20 vmv.v.i v4,0 j .L4 .L13: vsetvli a5,zero,e32,m4,ta,ma .L4: mv a7,a3 mv a4,a3 bleu a3,a6,.L3 csrr a4,vlenb .L3: vmv.v.x v8,t4 vl4re32.v v12,0(sp) ---- spill vand.vv v20,v28,v8 vand.vv v8,v12,v8 vsetvli zero,a4,e32,m4,ta,ma vle32.v v16,0(a0) vsetvli a5,zero,e32,m4,ta,ma add a3,a3,t1 vrgather.vv v12,v16,v20 add a0,a0,t3 vrgather.vv v20,v16,v8 vsub.vv v12,v12,v0 vsetvli zero,a4,e32,m4,tu,ma vadd.vv v4,v4,v12 vmacc.vv v4,v24,v20 bgtu a7,a6,.L13 csrr a1,vlenb slli a1,a1,2 add a1,a1,sp li a4,-1 csrr t0,vlenb vsetvli a5,zero,e32,m4,ta,ma vl4re32.v v12,0(a1) ---- spill vmv.v.i v8,0 vmul.vx v0,v12,a4 li a2,0 slli t1,t0,3 vadd.vi v0,v0,-1 vand.vi v0,v0,1 vmseq.vv v0,v0,v8 vand.vi v12,v12,1 vmerge.vvm v16,v8,v4,v0 vmseq.vv v12,v12,v8 vmv.s.x v1,a2 vmv1r.v v0,v12 vredsum.vs v16,v16,v1 vmerge.vvm v8,v8,v4,v0 vmv.x.s a0,v16 vredsum.vs v8,v8,v1 vmv.x.s a5,v8 add sp,sp,t1 addw a0,a0,a5 jr ra .L5: li a0,0 ret We can there are multiple horrible register spillings. The root cause of this issue is for a scalar IR load: _5 = *_4; We didn't check whether it is a continguous load/store or gather/scatter load/store Since it will be translate into: 1. MASK_LEN_GATHER_LOAD (..., perm indice). 2. Continguous load/store + VEC_PERM (..., perm indice) It's obvious that no matter which situation, we will end up with consuming one vector register group (perm indice) that we didn't count it before. So this case we pick LMUL = 4 which is incorrect choice for dynamic LMUL cost model. The key of this patch is: if ((type == load_vec_info_type || type == store_vec_info_type) && !adjacent_dr_p (STMT_VINFO_DATA_REF (stmt_info))) { ... } Add one more register consumption if it is not an adjacent load/store. After this patch, it pick LMUL = 2 which is optimal: bar: ble a3,zero,.L4 csrr a6,vlenb vsetvli a5,zero,e32,m2,ta,ma vmv.v.x v6,a2 srli a2,a6,1 vmv.v.x v4,a1 vid.v v12 slli a3,a3,1 vand.vi v0,v12,1 addi t1,a2,-1 vmseq.vi v0,v0,1 slli a6,a6,1 vsetvli zero,a5,e32,m2,ta,ma neg a7,a2 viota.m v2,v0 vsetvli a5,zero,e32,m2,ta,mu vrgather.vv v16,v4,v2 vrgather.vv v14,v6,v2 vrgather.vv v16,v6,v2,v0.t vrgather.vv v14,v4,v2,v0.t vand.vi v18,v12,-2 vmv.v.i v2,0 vadd.vi v20,v18,1 .L3: minu a4,a3,a2 vsetvli zero,a4,e32,m2,ta,ma vle32.v v8,0(a0) vsetvli a5,zero,e32,m2,ta,ma vmv.v.x v4,t1 vand.vv v10,v18,v4 vrgather.vv v6,v8,v10 vsub.vv v6,v6,v14 vsetvli zero,a4,e32,m2,tu,ma vadd.vv v2,v2,v6 vsetvli a1,zero,e32,m2,ta,ma vand.vv v4,v20,v4 vrgather.vv v6,v8,v4 vsetvli zero,a4,e32,m2,tu,ma mv a4,a3 add a0,a0,a6 add a3,a3,a7 vmacc.vv v2,v16,v6 bgtu a4,a2,.L3 vsetvli a1,zero,e32,m2,ta,ma vand.vi v0,v12,1 vmv.v.i v4,0 li a3,-1 vmseq.vv v0,v0,v4 vmv.s.x v1,zero vmerge.vvm v6,v4,v2,v0 vredsum.vs v6,v6,v1 vmul.vx v0,v12,a3 vadd.vi v0,v0,-1 vand.vi v0,v0,1 vmv.x.s a4,v6 vmseq.vv v0,v0,v4 vmv.s.x v1,zero vmerge.vvm v4,v4,v2,v0 vredsum.vs v4,v4,v1 vmv.x.s a0,v4 addw a0,a0,a4 ret .L4: li a0,0 ret No spillings. gcc/ChangeLog: * config/riscv/riscv-vector-costs.cc (max_number_of_live_regs): Fix unexpected LMUL. (get_store_value): New function. gcc/testsuite/ChangeLog: * gcc.dg/vect/costmodel/riscv/rvv/dynamic-lmul2-7.c: New test. --- gcc/config/riscv/riscv-vector-costs.cc | 103 ++++++++++++++++-- .../costmodel/riscv/rvv/dynamic-lmul2-7.c | 25 +++++ 2 files changed, 120 insertions(+), 8 deletions(-) create mode 100644 gcc/testsuite/gcc.dg/vect/costmodel/riscv/rvv/dynamic-lmul2-7.c diff --git a/gcc/config/riscv/riscv-vector-costs.cc b/gcc/config/riscv/riscv-vector-costs.cc index 0b890396535..4e8842cd2a6 100644 --- a/gcc/config/riscv/riscv-vector-costs.cc +++ b/gcc/config/riscv/riscv-vector-costs.cc @@ -40,6 +40,7 @@ along with GCC; see the file COPYING3. If not see #include "bitmap.h" #include "ssa.h" #include "backend.h" +#include "tree-data-ref.h" /* This file should be included last. */ #include "riscv-vector-costs.h" @@ -135,8 +136,9 @@ compute_local_program_points ( || is_gimple_call (gsi_stmt (si)))) continue; stmt_vec_info stmt_info = vinfo->lookup_stmt (gsi_stmt (si)); - if (STMT_VINFO_TYPE (vect_stmt_to_vectorize (stmt_info)) - != undef_vec_info_type) + enum stmt_vec_info_type type + = STMT_VINFO_TYPE (vect_stmt_to_vectorize (stmt_info)); + if (type != undef_vec_info_type) { stmt_point info = {point, gsi_stmt (si)}; program_points.safe_push (info); @@ -289,9 +291,7 @@ max_number_of_live_regs (const basic_block bb, unsigned int i; unsigned int live_point = 0; auto_vec live_vars_vec; - live_vars_vec.safe_grow (max_point + 1, true); - for (i = 0; i < live_vars_vec.length (); ++i) - live_vars_vec[i] = 0; + live_vars_vec.safe_grow_cleared (max_point + 1, true); for (hash_map::iterator iter = live_ranges.begin (); iter != live_ranges.end (); ++iter) { @@ -360,6 +360,21 @@ get_current_lmul (class loop *loop) return loop_autovec_infos.get (loop)->current_lmul; } +/* Get STORE value. */ +static tree +get_store_value (gimple *stmt) +{ + if (is_gimple_call (stmt) && gimple_call_internal_p (stmt)) + { + if (gimple_call_internal_fn (stmt) == IFN_MASK_STORE) + return gimple_call_arg (stmt, 3); + else + gcc_unreachable (); + } + else + return gimple_assign_rhs1 (stmt); +} + /* Update the live ranges according PHI. Loop: @@ -395,12 +410,13 @@ update_local_live_ranges ( unsigned int nbbs = loop->num_nodes; unsigned int i, j; gphi_iterator psi; + gimple_stmt_iterator si; for (i = 0; i < nbbs; i++) { - basic_block bb = bbs[i]; if (dump_enabled_p ()) dump_printf_loc (MSG_NOTE, vect_location, - "Update local program points for bb %d:\n", bb->index); + "Update local program points for bb %d:\n", + bbs[i]->index); for (psi = gsi_start_phis (bbs[i]); !gsi_end_p (psi); gsi_next (&psi)) { gphi *phi = psi.phi (); @@ -419,9 +435,42 @@ update_local_live_ranges ( unsigned int max_point = (*program_points_per_bb.get (e->src)).length () - 1; auto *live_range = live_ranges->get (def); + basic_block bb = e->src; if (!live_range) - continue; + { + if (single_succ_p (e->src)) + { + /* + [local count: 850510900]: + goto ; [100.00%] + + Handle this case, we should extend live range of bb 3. + */ + bb = single_succ (e->src); + if (!program_points_per_bb.get (bb)) + continue; + live_ranges = live_ranges_per_bb.get (bb); + max_point + = (*program_points_per_bb.get (bb)).length () - 1; + live_range = live_ranges->get (def); + if (!live_range) + continue; + } + else + continue; + } + gcc_assert (live_range); + if (bbs[i] == bb) + { + gcc_assert (live_range); + unsigned int start = (*live_range).first; + (*live_range).first = 0; + if (dump_enabled_p ()) + dump_printf_loc (MSG_NOTE, vect_location, + "Update %T start point from %d to %d:\n", + def, start, (*live_range).first); + } unsigned int end = (*live_range).second; (*live_range).second = max_point; if (dump_enabled_p ()) @@ -430,6 +479,44 @@ update_local_live_ranges ( end, (*live_range).second); } } + for (si = gsi_start_bb (bbs[i]); !gsi_end_p (si); gsi_next (&si)) + { + if (!(is_gimple_assign (gsi_stmt (si)) + || is_gimple_call (gsi_stmt (si)))) + continue; + stmt_vec_info stmt_info = vinfo->lookup_stmt (gsi_stmt (si)); + enum stmt_vec_info_type type + = STMT_VINFO_TYPE (vect_stmt_to_vectorize (stmt_info)); + if ((type == load_vec_info_type || type == store_vec_info_type) + && !adjacent_dr_p (STMT_VINFO_DATA_REF (stmt_info))) + { + /* For non-adjacent load/store STMT, we will potentially + convert it into: + + 1. MASK_LEN_GATHER_LOAD (..., perm indice). + 2. Continguous load/store + VEC_PERM (..., perm indice) + + We will be likely using one more vector variable. */ + unsigned int max_point + = (*program_points_per_bb.get (bbs[i])).length () - 1; + auto *live_ranges = live_ranges_per_bb.get (bbs[i]); + bool existed_p = false; + tree var = type == load_vec_info_type + ? gimple_get_lhs (gsi_stmt (si)) + : get_store_value (gsi_stmt (si)); + tree sel_type = build_nonstandard_integer_type ( + TYPE_PRECISION (TREE_TYPE (var)), 1); + tree sel = build_decl (UNKNOWN_LOCATION, VAR_DECL, + get_identifier ("vect_perm"), sel_type); + pair &live_range = live_ranges->get_or_insert (sel, &existed_p); + gcc_assert (!existed_p); + live_range = pair (0, max_point); + if (dump_enabled_p ()) + dump_printf_loc (MSG_NOTE, vect_location, + "Add perm indice %T, start = 0, end = %d\n", + sel, max_point); + } + } } } diff --git a/gcc/testsuite/gcc.dg/vect/costmodel/riscv/rvv/dynamic-lmul2-7.c b/gcc/testsuite/gcc.dg/vect/costmodel/riscv/rvv/dynamic-lmul2-7.c new file mode 100644 index 00000000000..3dfc6f16a25 --- /dev/null +++ b/gcc/testsuite/gcc.dg/vect/costmodel/riscv/rvv/dynamic-lmul2-7.c @@ -0,0 +1,25 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64gcv -mabi=lp64d -fdump-tree-vect-details" } */ + +int +bar (int *x, int a, int b, int n) +{ + x = __builtin_assume_aligned (x, __BIGGEST_ALIGNMENT__); + int sum1 = 0; + int sum2 = 0; + for (int i = 0; i < n; ++i) + { + sum1 += x[2*i] - a; + sum1 += x[2*i+1] * b; + sum2 += x[2*i] - b; + sum2 += x[2*i+1] * a; + } + return sum1 + sum2; +} + +/* { dg-final { scan-assembler {e32,m2} } } */ +/* { dg-final { scan-assembler-times {csrr} 1 } } */ +/* { dg-final { scan-tree-dump-times "Maximum lmul = 8" 1 "vect" } } */ +/* { dg-final { scan-tree-dump-times "Maximum lmul = 4" 1 "vect" } } */ +/* { dg-final { scan-tree-dump-times "Maximum lmul = 2" 1 "vect" } } */ +/* { dg-final { scan-tree-dump-not "Maximum lmul = 1" "vect" } } */