From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx0b-001b2d01.pphosted.com (mx0b-001b2d01.pphosted.com [148.163.158.5]) by sourceware.org (Postfix) with ESMTPS id 9A3323858D28 for ; Tue, 27 Jun 2023 18:52:09 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org 9A3323858D28 Authentication-Results: sourceware.org; dmarc=none (p=none dis=none) header.from=linux.ibm.com Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=linux.ibm.com Received: from pps.filterd (m0353725.ppops.net [127.0.0.1]) by mx0a-001b2d01.pphosted.com (8.17.1.19/8.17.1.19) with ESMTP id 35RIlkmc011356; Tue, 27 Jun 2023 18:52:09 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ibm.com; h=message-id : date : mime-version : to : from : subject : content-type : content-transfer-encoding; s=pp1; bh=QIbENRYD9eAc5fG1c26lYCXRl5IWWrvlVvgn1TIasWA=; b=q1c/1FqKwPEi+Omq8pggmhO+SxviuMpm/lJcDE/kfuBEVNsKuBA5tDsEZRg96NIFMr6Z LAIuK0y3MwmUJMBZQ3CwSgbr0znx4CwdeXl/WSrR3GW/fixR8FgGfLk8mHtuqr61WEPW RlIukj9Xz+7zIIqt/NWCxUHKB7+PRXK/JvGniC59kBmBl2DRPyrlr1CKEPDuBemhOKfb CxGN6dkOT6+neXmcbxzfUjP0ZOLmBmAwe1z+toJvQBWBleytM32+rcrEpy5u5a3axa8I B6q5i4bM0S/lrE6gE/hYpeV9oyY0wBXmMcvv34hIjfupDj/sDKY3pbrCTVZ5/X6GcMg3 tw== Received: from pps.reinject (localhost [127.0.0.1]) by mx0a-001b2d01.pphosted.com (PPS) with ESMTPS id 3rg5bv82pu-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Tue, 27 Jun 2023 18:52:08 +0000 Received: from m0353725.ppops.net (m0353725.ppops.net [127.0.0.1]) by pps.reinject (8.17.1.5/8.17.1.5) with ESMTP id 35RIn0iJ015084; Tue, 27 Jun 2023 18:52:07 GMT Received: from ppma03dal.us.ibm.com (b.bd.3ea9.ip4.static.sl-reverse.com [169.62.189.11]) by mx0a-001b2d01.pphosted.com (PPS) with ESMTPS id 3rg5bv82pj-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Tue, 27 Jun 2023 18:52:07 +0000 Received: from pps.filterd (ppma03dal.us.ibm.com [127.0.0.1]) by ppma03dal.us.ibm.com (8.17.1.19/8.17.1.19) with ESMTP id 35RH3Igu015204; Tue, 27 Jun 2023 18:52:07 GMT Received: from smtprelay05.dal12v.mail.ibm.com ([9.208.130.101]) by ppma03dal.us.ibm.com (PPS) with ESMTPS id 3rdr45ejxm-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Tue, 27 Jun 2023 18:52:07 +0000 Received: from smtpav04.wdc07v.mail.ibm.com (smtpav04.wdc07v.mail.ibm.com [10.39.53.231]) by smtprelay05.dal12v.mail.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id 35RIq5mE8323710 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Tue, 27 Jun 2023 18:52:05 GMT Received: from smtpav04.wdc07v.mail.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id BAF5258061; Tue, 27 Jun 2023 18:52:02 +0000 (GMT) Received: from smtpav04.wdc07v.mail.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 0F85E5805E; Tue, 27 Jun 2023 18:52:02 +0000 (GMT) Received: from [9.67.139.18] (unknown [9.67.139.18]) by smtpav04.wdc07v.mail.ibm.com (Postfix) with ESMTPS; Tue, 27 Jun 2023 18:52:01 +0000 (GMT) Message-ID: <68d7fbb3-59b3-6a59-a8ac-773d5d9c6817@linux.ibm.com> Date: Tue, 27 Jun 2023 13:52:01 -0500 MIME-Version: 1.0 User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:102.0) Gecko/20100101 Thunderbird/102.12.0 Content-Language: en-US To: GCC Patches , Segher Boessenkool , "Kewen.Lin" , Peter Bergner , David Edelsohn From: Pat Haugen Subject: [PATCH V3, rs6000] Disable generation of scalar modulo instructions Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit X-TM-AS-GCONF: 00 X-Proofpoint-GUID: pV5tPuiXIgcJlR5B1t4bBq2QLjcBbwId X-Proofpoint-ORIG-GUID: BqbMqdYsaNcKWhRgaS9gCfxxP778DIJB X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.254,Aquarius:18.0.957,Hydra:6.0.591,FMLib:17.11.176.26 definitions=2023-06-27_12,2023-06-27_01,2023-05-22_02 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 lowpriorityscore=0 suspectscore=0 phishscore=0 spamscore=0 impostorscore=0 mlxscore=0 bulkscore=0 priorityscore=1501 malwarescore=0 adultscore=0 clxscore=1015 mlxlogscore=999 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2305260000 definitions=main-2306270168 X-Spam-Status: No, score=-11.9 required=5.0 tests=BAYES_00,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_EF,GIT_PATCH_0,KAM_SHORT,RCVD_IN_MSPIKE_H5,RCVD_IN_MSPIKE_WL,SPF_HELO_NONE,SPF_PASS,TXREP,T_SCC_BODY_TEXT_LINE autolearn=ham autolearn_force=no version=3.4.6 X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on server2.sourceware.org List-Id: Updated from prior version to address review comments (update rs6000_rtx_cost, update scan strings of mod-1.c/mod-2.c)l. Disable generation of scalar modulo instructions. It was recently discovered that the scalar modulo instructions can suffer noticeable performance issues for certain input values. This patch disables their generation since the equivalent div/mul/sub sequence does not suffer the same problem. Bootstrapped and regression tested on powerpc64/powerpc64le. Ok for master and backports after burn in? -Pat 2023-06-27 Pat Haugen gcc/ * config/rs6000/rs6000.cc (rs6000_rtx_costs): Check if disabling scalar modulo. * config/rs6000/rs6000.h (RS6000_DISABLE_SCALAR_MODULO): New. * config/rs6000/rs6000.md (mod3, *mod3): Disable. (define_expand umod3): New. (define_insn umod3): Rename to *umod3 and disable. (umodti3, modti3): Disable. gcc/testsuite/ * gcc.target/powerpc/clone1.c: Add xfails. * gcc.target/powerpc/clone3.c: Likewise. * gcc.target/powerpc/mod-1.c: Update scan strings and add xfails. * gcc.target/powerpc/mod-2.c: Likewise. * gcc.target/powerpc/p10-vdivq-vmodq.c: Add xfails. diff --git a/gcc/config/rs6000/rs6000.cc b/gcc/config/rs6000/rs6000.cc index 546c353029b..2dae217bf64 100644 --- a/gcc/config/rs6000/rs6000.cc +++ b/gcc/config/rs6000/rs6000.cc @@ -22127,7 +22127,9 @@ rs6000_rtx_costs (rtx x, machine_mode mode, int outer_code, *total = rs6000_cost->divsi; } /* Add in shift and subtract for MOD unless we have a mod instruction. */ - if (!TARGET_MODULO && (code == MOD || code == UMOD)) + if ((!TARGET_MODULO + || (RS6000_DISABLE_SCALAR_MODULO && SCALAR_INT_MODE_P (mode))) + && (code == MOD || code == UMOD)) *total += COSTS_N_INSNS (2); return false; diff --git a/gcc/config/rs6000/rs6000.h b/gcc/config/rs6000/rs6000.h index 3503614efbd..22595f6ebd7 100644 --- a/gcc/config/rs6000/rs6000.h +++ b/gcc/config/rs6000/rs6000.h @@ -2492,3 +2492,9 @@ while (0) rs6000_asm_output_opcode (STREAM); \ } \ while (0) + +/* Disable generation of scalar modulo instructions due to performance issues + with certain input values. This can be removed in the future when the + issues have been resolved. */ +#define RS6000_DISABLE_SCALAR_MODULO 1 + diff --git a/gcc/config/rs6000/rs6000.md b/gcc/config/rs6000/rs6000.md index b0db8ae508d..6c2f237a539 100644 --- a/gcc/config/rs6000/rs6000.md +++ b/gcc/config/rs6000/rs6000.md @@ -3421,6 +3421,17 @@ (define_expand "mod3" FAIL; operands[2] = force_reg (mode, operands[2]); + + if (RS6000_DISABLE_SCALAR_MODULO) + { + temp1 = gen_reg_rtx (mode); + temp2 = gen_reg_rtx (mode); + + emit_insn (gen_div3 (temp1, operands[1], operands[2])); + emit_insn (gen_mul3 (temp2, temp1, operands[2])); + emit_insn (gen_sub3 (operands[0], operands[1], temp2)); + DONE; + } } else { @@ -3440,17 +3451,42 @@ (define_insn "*mod3" [(set (match_operand:GPR 0 "gpc_reg_operand" "=&r,r") (mod:GPR (match_operand:GPR 1 "gpc_reg_operand" "r,r") (match_operand:GPR 2 "gpc_reg_operand" "r,r")))] - "TARGET_MODULO" + "TARGET_MODULO && !RS6000_DISABLE_SCALAR_MODULO" "mods %0,%1,%2" [(set_attr "type" "div") (set_attr "size" "")]) +;; This define_expand can be removed when RS6000_DISABLE_SCALAR_MODULO is +;; removed. +(define_expand "umod3" + [(set (match_operand:GPR 0 "gpc_reg_operand") + (umod:GPR (match_operand:GPR 1 "gpc_reg_operand") + (match_operand:GPR 2 "gpc_reg_operand")))] + "" +{ + rtx temp1; + rtx temp2; + + if (!TARGET_MODULO) + FAIL; -(define_insn "umod3" + if (RS6000_DISABLE_SCALAR_MODULO) + { + temp1 = gen_reg_rtx (mode); + temp2 = gen_reg_rtx (mode); + + emit_insn (gen_udiv3 (temp1, operands[1], operands[2])); + emit_insn (gen_mul3 (temp2, temp1, operands[2])); + emit_insn (gen_sub3 (operands[0], operands[1], temp2)); + DONE; + } +}) + +(define_insn "*umod3" [(set (match_operand:GPR 0 "gpc_reg_operand" "=&r,r") (umod:GPR (match_operand:GPR 1 "gpc_reg_operand" "r,r") (match_operand:GPR 2 "gpc_reg_operand" "r,r")))] - "TARGET_MODULO" + "TARGET_MODULO && !RS6000_DISABLE_SCALAR_MODULO" "modu %0,%1,%2" [(set_attr "type" "div") (set_attr "size" "")]) @@ -3507,7 +3543,7 @@ (define_insn "umodti3" [(set (match_operand:TI 0 "altivec_register_operand" "=v") (umod:TI (match_operand:TI 1 "altivec_register_operand" "v") (match_operand:TI 2 "altivec_register_operand" "v")))] - "TARGET_POWER10 && TARGET_POWERPC64" + "TARGET_POWER10 && TARGET_POWERPC64 && !RS6000_DISABLE_SCALAR_MODULO" "vmoduq %0,%1,%2" [(set_attr "type" "vecdiv") (set_attr "size" "128")]) @@ -3516,7 +3552,7 @@ (define_insn "modti3" [(set (match_operand:TI 0 "altivec_register_operand" "=v") (mod:TI (match_operand:TI 1 "altivec_register_operand" "v") (match_operand:TI 2 "altivec_register_operand" "v")))] - "TARGET_POWER10 && TARGET_POWERPC64" + "TARGET_POWER10 && TARGET_POWERPC64 && !RS6000_DISABLE_SCALAR_MODULO" "vmodsq %0,%1,%2" [(set_attr "type" "vecdiv") (set_attr "size" "128")]) diff --git a/gcc/testsuite/gcc.target/powerpc/clone1.c b/gcc/testsuite/gcc.target/powerpc/clone1.c index c69fd2aa1b8..74323ca0e8c 100644 --- a/gcc/testsuite/gcc.target/powerpc/clone1.c +++ b/gcc/testsuite/gcc.target/powerpc/clone1.c @@ -21,6 +21,7 @@ long mod_func_or (long a, long b, long c) return mod_func (a, b) | c; } -/* { dg-final { scan-assembler-times {\mdivd\M} 1 } } */ -/* { dg-final { scan-assembler-times {\mmulld\M} 1 } } */ -/* { dg-final { scan-assembler-times {\mmodsd\M} 1 } } */ +/* { Fail due to RS6000_DISABLE_SCALAR_MODULO. */ +/* { dg-final { scan-assembler-times {\mdivd\M} 1 { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-times {\mmulld\M} 1 { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-times {\mmodsd\M} 1 { xfail *-*-* } } } */ diff --git a/gcc/testsuite/gcc.target/powerpc/clone3.c b/gcc/testsuite/gcc.target/powerpc/clone3.c index 911b88b781d..d3eb4dd2378 100644 --- a/gcc/testsuite/gcc.target/powerpc/clone3.c +++ b/gcc/testsuite/gcc.target/powerpc/clone3.c @@ -27,7 +27,8 @@ long mod_func_or (long a, long b, long c) return mod_func (a, b) | c; } -/* { dg-final { scan-assembler-times {\mdivd\M} 1 } } */ -/* { dg-final { scan-assembler-times {\mmulld\M} 1 } } */ -/* { dg-final { scan-assembler-times {\mmodsd\M} 2 } } */ +/* { Fail due to RS6000_DISABLE_SCALAR_MODULO. */ +/* { dg-final { scan-assembler-times {\mdivd\M} 1 { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-times {\mmulld\M} 1 { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-times {\mmodsd\M} 2 { xfail *-*-* } } } */ /* { dg-final { scan-assembler-times {\mpld\M} 1 } } */ diff --git a/gcc/testsuite/gcc.target/powerpc/mod-1.c b/gcc/testsuite/gcc.target/powerpc/mod-1.c index 861ba670af4..8720ffb3346 100644 --- a/gcc/testsuite/gcc.target/powerpc/mod-1.c +++ b/gcc/testsuite/gcc.target/powerpc/mod-1.c @@ -7,13 +7,14 @@ long lsmod (long a, long b) { return a%b; } unsigned int iumod (unsigned int a, unsigned int b) { return a%b; } unsigned long lumod (unsigned long a, unsigned long b) { return a%b; } -/* { dg-final { scan-assembler-times "modsw " 1 } } */ -/* { dg-final { scan-assembler-times "modsd " 1 } } */ -/* { dg-final { scan-assembler-times "moduw " 1 } } */ -/* { dg-final { scan-assembler-times "modud " 1 } } */ -/* { dg-final { scan-assembler-not "mullw " } } */ -/* { dg-final { scan-assembler-not "mulld " } } */ -/* { dg-final { scan-assembler-not "divw " } } */ -/* { dg-final { scan-assembler-not "divd " } } */ -/* { dg-final { scan-assembler-not "divwu " } } */ -/* { dg-final { scan-assembler-not "divdu " } } */ +/* { Fail due to RS6000_DISABLE_SCALAR_MODULO. */ +/* { dg-final { scan-assembler-times {\mmodsw\M} 1 { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-times {\mmodsd\M} 1 { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-times {\mmoduw\M} 1 { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-times {\mmodud\M} 1 { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-not {\mmullw\M} { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-not {\mmulld\M} { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-not {\mdivw\M} { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-not {\mdivd\M} { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-not {\mdivwu\M} { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-not {\mdivdu\M} { xfail *-*-* } } } */ diff --git a/gcc/testsuite/gcc.target/powerpc/mod-2.c b/gcc/testsuite/gcc.target/powerpc/mod-2.c index 441ec5878f1..54bdca88607 100644 --- a/gcc/testsuite/gcc.target/powerpc/mod-2.c +++ b/gcc/testsuite/gcc.target/powerpc/mod-2.c @@ -5,8 +5,9 @@ int ismod (int a, int b) { return a%b; } unsigned int iumod (unsigned int a, unsigned int b) { return a%b; } -/* { dg-final { scan-assembler-times "modsw " 1 } } */ -/* { dg-final { scan-assembler-times "moduw " 1 } } */ -/* { dg-final { scan-assembler-not "mullw " } } */ -/* { dg-final { scan-assembler-not "divw " } } */ -/* { dg-final { scan-assembler-not "divwu " } } */ +/* { Fail due to RS6000_DISABLE_SCALAR_MODULO. */ +/* { dg-final { scan-assembler-times {\mmodsw\M} 1 { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-times {\mmoduw\M} 1 { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-not {\mmullw\M} { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-not {\mdivw\M} { xfail *-*-* } } } */ +/* { dg-final { scan-assembler-not {\mdivwu\M} { xfail *-*-* } } } */ diff --git a/gcc/testsuite/gcc.target/powerpc/p10-vdivq-vmodq.c b/gcc/testsuite/gcc.target/powerpc/p10-vdivq-vmodq.c index 84685e5ff43..148998c8c9d 100644 --- a/gcc/testsuite/gcc.target/powerpc/p10-vdivq-vmodq.c +++ b/gcc/testsuite/gcc.target/powerpc/p10-vdivq-vmodq.c @@ -23,5 +23,6 @@ __int128 s_mod(__int128 a, __int128 b) /* { dg-final { scan-assembler {\mvdivsq\M} } } */ /* { dg-final { scan-assembler {\mvdivuq\M} } } */ -/* { dg-final { scan-assembler {\mvmodsq\M} } } */ -/* { dg-final { scan-assembler {\mvmoduq\M} } } */ +/* { Fail due to RS6000_DISABLE_SCALAR_MODULO. */ +/* { dg-final { scan-assembler {\mvmodsq\M} { xfail *-*-* } } } */ +/* { dg-final { scan-assembler {\mvmoduq\M} { xfail *-*-* } } } */