From patchwork Fri Jan 27 17:06:06 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Richard Sandiford X-Patchwork-Id: 63805 Return-Path: X-Original-To: patchwork@sourceware.org Delivered-To: patchwork@sourceware.org Received: from server2.sourceware.org (localhost [IPv6:::1]) by sourceware.org (Postfix) with ESMTP id 741CA385B517 for ; Fri, 27 Jan 2023 17:06:40 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 741CA385B517 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org; s=default; t=1674839200; bh=5mD3SS4AIx1gYyWykq8qV8VNMgQwRE/hQn2Ht+jcHmk=; h=To:Subject:Date:List-Id:List-Unsubscribe:List-Archive:List-Post: List-Help:List-Subscribe:From:Reply-To:From; b=MFgEDJOWEhC62R84wL65eg5G+PYcuU7W+v1g5rdh9xakg3LaMwTNRj5LNj8bLExLB rVoZN+BMWuBquYWevcT9+yjUcbT7Q+Vwcd/WaePjVyfbqmawInCBHnXcTKSz3/Bq4M mTExcieXv8Ec9g2NBojWkK8iuMBWH+0rz4QCEepo= X-Original-To: gcc-patches@gcc.gnu.org Delivered-To: gcc-patches@gcc.gnu.org Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by sourceware.org (Postfix) with ESMTP id 036853858C00 for ; Fri, 27 Jan 2023 17:06:09 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org 036853858C00 Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 86B7B2B for ; Fri, 27 Jan 2023 09:06:50 -0800 (PST) Received: from localhost (e121540-lin.manchester.arm.com [10.32.99.50]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPSA id 2B1D13F71E for ; Fri, 27 Jan 2023 09:06:08 -0800 (PST) To: gcc-patches@gcc.gnu.org Mail-Followup-To: gcc-patches@gcc.gnu.org, richard.sandiford@arm.com Subject: [pushed] aarch64: Prevent simd tests from being optimised away Date: Fri, 27 Jan 2023 17:06:06 +0000 Message-ID: User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/26.3 (gnu/linux) MIME-Version: 1.0 X-Spam-Status: No, score=-36.5 required=5.0 tests=BAYES_00, GIT_PATCH_0, KAM_DMARC_NONE, KAM_DMARC_STATUS, KAM_LAZY_DOMAIN_SECURITY, KAM_SHORT, SPF_HELO_NONE, SPF_NONE, 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: 2.1.29 Precedence: list List-Id: Gcc-patches mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-Patchwork-Original-From: Richard Sandiford via Gcc-patches From: Richard Sandiford Reply-To: Richard Sandiford Errors-To: gcc-patches-bounces+patchwork=sourceware.org@gcc.gnu.org Sender: "Gcc-patches" The vqdml[as]l[hs]_laneq_* tests were folded at compile time, meaning that we didn't have any Advanced SIMD instructions in the assembly. Kyrill's preference was to use wrapper functions, so this patch does that for the failing tests and for others that had scan-assemblers with inline intrinsics calls. (There were some tests that already used wrapper functions, some that used volatile, some that used inline asm barriers, and some that had no separation.) Doing that for vqdmulhs_lane_s32.c meant that we generated the scalar form of the instruction, rather than a vector instruction operating on lane 0. That seems fair enough, so the patch keeps that test but adds a second one for lane 1. Tested on aarch64-linux-gnu & pushed. Richard gcc/testsuite/ * gcc.target/aarch64/simd/vfma_f64.c: Use a wrapper function rather than an asm barrier. * gcc.target/aarch64/simd/vfms_f64.c: Likewise. * gcc.target/aarch64/simd/vmul_f64_1.c: Use a wrapper function rather than volatile. * gcc.target/aarch64/simd/vmul_n_f64_1.c: Likewise. * gcc.target/aarch64/simd/vqdmlalh_laneq_s16_1.c: Use a wrapper function. Remove -fno-inline. * gcc.target/aarch64/simd/vqdmlals_laneq_s32_1.c: Likewise. * gcc.target/aarch64/simd/vqdmlslh_laneq_s16_1.c: Likewise. * gcc.target/aarch64/simd/vqdmlsls_laneq_s32_1.c: Likewise. * gcc.target/aarch64/simd/vqdmulhh_lane_s16.c: Likewise. * gcc.target/aarch64/simd/vqdmulhh_laneq_s16_1.c: Likewise. * gcc.target/aarch64/simd/vqdmulhs_laneq_s32_1.c: Likewise. * gcc.target/aarch64/simd/vqrdmulhh_lane_s16.c: Likewise. * gcc.target/aarch64/simd/vqrdmulhh_laneq_s16_1.c: Likewise. * gcc.target/aarch64/simd/vqrdmulhs_lane_s32.c: Likewise. * gcc.target/aarch64/simd/vqrdmulhs_laneq_s32_1.c: Likewise. * gcc.target/aarch64/simd/vqdmulhs_lane_s32.c: Likewise. Allow the scalar form to be used when operating on lane 0. Add a test for lane 1. --- .../gcc.target/aarch64/simd/vfma_f64.c | 27 +++++---------- .../gcc.target/aarch64/simd/vfms_f64.c | 27 +++++---------- .../gcc.target/aarch64/simd/vmul_f64_1.c | 12 ++++--- .../gcc.target/aarch64/simd/vmul_n_f64_1.c | 12 ++++--- .../aarch64/simd/vqdmlalh_laneq_s16_1.c | 20 +++++------ .../aarch64/simd/vqdmlals_laneq_s32_1.c | 20 +++++------ .../aarch64/simd/vqdmlslh_laneq_s16_1.c | 20 +++++------ .../aarch64/simd/vqdmlsls_laneq_s32_1.c | 21 ++++++------ .../aarch64/simd/vqdmulhh_lane_s16.c | 15 +++++---- .../aarch64/simd/vqdmulhh_laneq_s16_1.c | 18 +++++----- .../aarch64/simd/vqdmulhs_lane_s32.c | 33 ++++++++++++++----- .../aarch64/simd/vqdmulhs_laneq_s32_1.c | 18 +++++----- .../aarch64/simd/vqrdmulhh_lane_s16.c | 15 +++++---- .../aarch64/simd/vqrdmulhh_laneq_s16_1.c | 18 +++++----- .../aarch64/simd/vqrdmulhs_lane_s32.c | 15 +++++---- .../aarch64/simd/vqrdmulhs_laneq_s32_1.c | 18 +++++----- 16 files changed, 163 insertions(+), 146 deletions(-) diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vfma_f64.c b/gcc/testsuite/gcc.target/aarch64/simd/vfma_f64.c index ef414f1b2fc..467c740ea12 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vfma_f64.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vfma_f64.c @@ -7,33 +7,24 @@ #define EPS 1.0e-15 -#define INHIB_OPT(x) asm volatile ("mov %d0, %1.d[0]" \ - : "=w"(x) \ - : "w"(x) \ - : /* No clobbers. */); - extern void abort (void); +float64_t __attribute__((noipa)) +test_vfma (float64x1_t arg1, float64x1_t arg2, float64x1_t arg3) +{ + return vget_lane_f64 (vfma_f64 (arg1, arg2, arg3), 0); +} + int main (void) { - float64x1_t arg1; - float64x1_t arg2; - float64x1_t arg3; - float64_t expected; float64_t actual; - arg1 = vcreate_f64 (0x3fe3955382d35b0eULL); - arg2 = vcreate_f64 (0x3fa88480812d6670ULL); - arg3 = vcreate_f64 (0x3fd5791ae2a92572ULL); - - INHIB_OPT (arg1); - INHIB_OPT (arg2); - INHIB_OPT (arg3); - expected = 0.6280448184360076; - actual = vget_lane_f64 (vfma_f64 (arg1, arg2, arg3), 0); + actual = test_vfma (vcreate_f64 (0x3fe3955382d35b0eULL), + vcreate_f64 (0x3fa88480812d6670ULL), + vcreate_f64 (0x3fd5791ae2a92572ULL)); if (__builtin_fabs (expected - actual) > EPS) abort (); diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vfms_f64.c b/gcc/testsuite/gcc.target/aarch64/simd/vfms_f64.c index afbb8a892c6..af6ca6ff11e 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vfms_f64.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vfms_f64.c @@ -7,33 +7,24 @@ #define EPS 1.0e-15 -#define INHIB_OPT(x) asm volatile ("mov %d0, %1.d[0]" \ - : "=w"(x) \ - : "w"(x) \ - : /* No clobbers. */); - extern void abort (void); +float64_t __attribute__((noipa)) +test_vfms (float64x1_t arg1, float64x1_t arg2, float64x1_t arg3) +{ + return vget_lane_f64 (vfms_f64 (arg1, arg2, arg3), 0); +} + int main (void) { - float64x1_t arg1; - float64x1_t arg2; - float64x1_t arg3; - float64_t expected; float64_t actual; - arg1 = vcreate_f64 (0x3fe730af8db9e6f7ULL); - arg2 = vcreate_f64 (0x3fe6b78680fa29ceULL); - arg3 = vcreate_f64 (0x3feea3cbf921fbe0ULL); - - INHIB_OPT (arg1); - INHIB_OPT (arg2); - INHIB_OPT (arg3); - expected = 4.4964705746355915e-2; - actual = vget_lane_f64 (vfms_f64 (arg1, arg2, arg3), 0); + actual = test_vfms (vcreate_f64 (0x3fe730af8db9e6f7ULL), + vcreate_f64 (0x3fe6b78680fa29ceULL), + vcreate_f64 (0x3feea3cbf921fbe0ULL)); if (__builtin_fabs (expected - actual) > EPS) abort (); diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vmul_f64_1.c b/gcc/testsuite/gcc.target/aarch64/simd/vmul_f64_1.c index c855c8cdbf4..dfa808cb9c2 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vmul_f64_1.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vmul_f64_1.c @@ -7,19 +7,23 @@ extern void abort (void); +float64_t __attribute__((noipa)) +test_vmul (float64x1_t arg1, float64x1_t arg2) +{ + return vget_lane_f64 (vmul_f64 (arg1, arg2), 0); +} + int main (void) { - volatile float64_t minus_e, pi; + float64_t minus_e, pi; float64_t expected, actual; pi = 3.14159265359; minus_e = -2.71828; expected = pi * minus_e; - - actual = vget_lane_f64 (vmul_f64 ((float64x1_t) { pi }, - (float64x1_t) { minus_e }), 0); + actual = test_vmul ((float64x1_t) { pi }, (float64x1_t) { minus_e }); if (expected != actual) abort (); diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vmul_n_f64_1.c b/gcc/testsuite/gcc.target/aarch64/simd/vmul_n_f64_1.c index f8f3cd2ed91..91c12bc9c82 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vmul_n_f64_1.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vmul_n_f64_1.c @@ -7,19 +7,23 @@ extern void abort (void); +float64_t __attribute__((noipa)) +test_vmul (float64x1_t arg1, float64_t arg2) +{ + return vget_lane_f64 (vmul_n_f64 (arg1, arg2), 0); +} + int main (void) { - volatile float64_t minus_e, pi; + float64_t minus_e, pi; float64_t expected, actual; pi = 3.14159265359; minus_e = -2.71828; expected = pi * minus_e; - - actual = vget_lane_f64 (vmul_n_f64 ((float64x1_t) { pi }, - minus_e), 0); + actual = test_vmul ((float64x1_t) { pi }, minus_e); if (expected != actual) abort (); diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vqdmlalh_laneq_s16_1.c b/gcc/testsuite/gcc.target/aarch64/simd/vqdmlalh_laneq_s16_1.c index 9a9bd0dcf05..aa6b89aaa5b 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vqdmlalh_laneq_s16_1.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vqdmlalh_laneq_s16_1.c @@ -1,27 +1,27 @@ /* Test the vqdmlalh_laneq_s16 AArch64 SIMD intrinsic. */ /* { dg-do run } */ -/* { dg-options "-save-temps -O3 -fno-inline" } */ +/* { dg-options "-save-temps -O3" } */ #include "arm_neon.h" extern void abort (void); +int32_t __attribute__((noipa)) +test_vqdmlalh (int32_t arg1, int16_t arg2, int16x8_t arg3) +{ + return vqdmlalh_laneq_s16 (arg1, arg2, arg3, 7); +} + int main (void) { - int32_t arg1; - int16_t arg2; - int16x8_t arg3; int32_t actual; int32_t expected; - arg1 = 0x80000000; - arg2 = -24497; - arg3 = vcombine_s16 (vcreate_s16 (0x008a80007fff7fffULL), - vcreate_s16 (0xfffffa797fff8000ULL)); - - actual = vqdmlalh_laneq_s16 (arg1, arg2, arg3, 7); + actual = test_vqdmlalh (0x80000000, -24497, + vcombine_s16 (vcreate_s16 (0x008a80007fff7fffULL), + vcreate_s16 (0xfffffa797fff8000ULL))); expected = -2147434654; if (expected != actual) diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vqdmlals_laneq_s32_1.c b/gcc/testsuite/gcc.target/aarch64/simd/vqdmlals_laneq_s32_1.c index 0dbe3392a6a..ea39595ae93 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vqdmlals_laneq_s32_1.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vqdmlals_laneq_s32_1.c @@ -1,27 +1,27 @@ /* Test the vqdmlals_laneq_s32 AArch64 SIMD intrinsic. */ /* { dg-do run } */ -/* { dg-options "-save-temps -O3 -fno-inline" } */ +/* { dg-options "-save-temps -O3" } */ #include "arm_neon.h" extern void abort (void); +int64_t __attribute__((noipa)) +test_vqdmlals (int64_t arg1, int32_t arg2, int32x4_t arg3) +{ + return vqdmlals_laneq_s32 (arg1, arg2, arg3, 3); +} + int main (void) { - int64_t arg1; - int32_t arg2; - int32x4_t arg3; int64_t actual; int64_t expected; - arg1 = -9223182289494545592LL; - arg2 = 32768; - arg3 = vcombine_s32 (vcreate_s32 (0xffff7fff8000ffffULL), - vcreate_s32 (0x80000000ffff0000ULL)); - - actual = vqdmlals_laneq_s32 (arg1, arg2, arg3, 3); + actual = test_vqdmlals (-9223182289494545592LL, 32768, + vcombine_s32 (vcreate_s32 (0xffff7fff8000ffffULL), + vcreate_s32 (0x80000000ffff0000ULL))); expected = -9223323026982900920LL; if (expected != actual) diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vqdmlslh_laneq_s16_1.c b/gcc/testsuite/gcc.target/aarch64/simd/vqdmlslh_laneq_s16_1.c index 2763e06127f..0f1babca38d 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vqdmlslh_laneq_s16_1.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vqdmlslh_laneq_s16_1.c @@ -1,27 +1,27 @@ /* Test the vqdmlslh_laneq_s16 AArch64 SIMD intrinsic. */ /* { dg-do run } */ -/* { dg-options "-save-temps -O3 -fno-inline" } */ +/* { dg-options "-save-temps -O3" } */ #include "arm_neon.h" extern void abort (void); +int32_t __attribute__((noipa)) +test_vqdmlslh (int32_t arg1, int16_t arg2, int16x8_t arg3) +{ + return vqdmlslh_laneq_s16 (arg1, arg2, arg3, 4); +} + int main (void) { - int32_t arg1; - int16_t arg2; - int16x8_t arg3; int32_t actual; int32_t expected; - arg1 = -2147450881; - arg2 = 32767; - arg3 = vcombine_s16 (vcreate_s16 (0x359d7fff00007fffULL), - vcreate_s16 (0xe678ffff00008000ULL)); - - actual = vqdmlslh_laneq_s16 (arg1, arg2, arg3, 4); + actual = test_vqdmlslh (-2147450881, 32767, + vcombine_s16 (vcreate_s16 (0x359d7fff00007fffULL), + vcreate_s16 (0xe678ffff00008000ULL))); expected = -32769; if (expected != actual) diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vqdmlsls_laneq_s32_1.c b/gcc/testsuite/gcc.target/aarch64/simd/vqdmlsls_laneq_s32_1.c index e003adb7c9e..ff922541f1a 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vqdmlsls_laneq_s32_1.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vqdmlsls_laneq_s32_1.c @@ -1,28 +1,27 @@ /* Test the vqdmlsls_laneq_s32 AArch64 SIMD intrinsic. */ /* { dg-do run } */ -/* { dg-options "-save-temps -O3 -fno-inline" } */ +/* { dg-options "-save-temps -O3" } */ #include "arm_neon.h" extern void abort (void); +int64_t __attribute__((noipa)) +test_vqdmlsls (int64_t arg1, int32_t arg2, int32x4_t arg3) +{ + return vqdmlsls_laneq_s32 (arg1, arg2, arg3, 3); +} + int main (void) { - int64_t arg1; - int32_t arg2; - int32x4_t arg3; int64_t actual; int64_t expected; - arg1 = 140733193453567LL; - arg2 = 25544; - arg3 = vcombine_s32 (vcreate_s32 (0x417b8000ffff8397LL), - vcreate_s32 (0x7fffffff58488000LL)); - - - actual = vqdmlsls_laneq_s32 (arg1, arg2, arg3, 3); + actual = test_vqdmlsls (140733193453567LL, 25544, + vcombine_s32 (vcreate_s32 (0x417b8000ffff8397LL), + vcreate_s32 (0x7fffffff58488000LL))); expected = 31022548895631LL; if (expected != actual) diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhh_lane_s16.c b/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhh_lane_s16.c index 75f67702493..12b79715b29 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhh_lane_s16.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhh_lane_s16.c @@ -1,25 +1,26 @@ /* Test the vqdmulhh_lane_s16 AArch64 SIMD intrinsic. */ /* { dg-do run } */ -/* { dg-options "-save-temps -O3 -fno-inline" } */ +/* { dg-options "-save-temps -O3" } */ #include "arm_neon.h" #include extern void abort (void); +int16_t __attribute__((noipa)) +test_vqdmulhh (int16_t arg1, int16x4_t arg2) +{ + return vqdmulhh_lane_s16 (arg1, arg2, 2); +} + int main (void) { - int16_t arg1; - int16x4_t arg2; - int16_t result; int16_t actual; int16_t expected; - arg1 = -32768; - arg2 = vcreate_s16 (0x0000ffff2489e398ULL); - actual = vqdmulhh_lane_s16 (arg1, arg2, 2); + actual = test_vqdmulhh (-32768, vcreate_s16 (0x0000ffff2489e398ULL)); expected = 1; if (expected != actual) diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhh_laneq_s16_1.c b/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhh_laneq_s16_1.c index b3ae37c67bd..1015c6878ed 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhh_laneq_s16_1.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhh_laneq_s16_1.c @@ -1,25 +1,27 @@ /* Test the vqdmulhh_laneq_s16 AArch64 SIMD intrinsic. */ /* { dg-do run } */ -/* { dg-options "-save-temps -O3 -fno-inline" } */ +/* { dg-options "-save-temps -O3" } */ #include "arm_neon.h" extern void abort (void); +int16_t __attribute__((noipa)) +test_vqdmulhh (int16_t arg1, int16x8_t arg2) +{ + return vqdmulhh_laneq_s16 (arg1, arg2, 7); +} + int main (void) { - int16_t arg1; - int16x8_t arg2; int16_t actual; int16_t expected; - arg1 = 268; - arg2 = vcombine_s16 (vcreate_s16 (0xffffffff00000000ULL), - vcreate_s16 (0x0000800018410000ULL)); - - actual = vqdmulhh_laneq_s16 (arg1, arg2, 7); + actual = test_vqdmulhh (268, + vcombine_s16 (vcreate_s16 (0xffffffff00000000ULL), + vcreate_s16 (0x0000800018410000ULL))); expected = 0; if (expected != actual) diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhs_lane_s32.c b/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhs_lane_s32.c index eef3ac0610a..f3b297ecc46 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhs_lane_s32.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhs_lane_s32.c @@ -1,27 +1,43 @@ /* Test the vqdmulhs_lane_s32 AArch64 SIMD intrinsic. */ /* { dg-do run } */ -/* { dg-options "-save-temps -O3 -fno-inline" } */ +/* { dg-options "-save-temps -O3" } */ #include "arm_neon.h" #include extern void abort (void); +int32_t __attribute__((noipa)) +test_vqdmulhs_0 (int32_t arg1, int32x2_t arg2) +{ + return vqdmulhs_lane_s32 (arg1, arg2, 0); +} + +int32_t __attribute__((noipa)) +test_vqdmulhs_1 (int32_t arg1, int32x2_t arg2) +{ + return vqdmulhs_lane_s32 (arg1, arg2, 1); +} + int main (void) { - int32_t arg1; - int32x2_t arg2; - int32_t result; int32_t actual; int32_t expected; - arg1 = 57336; - arg2 = vcreate_s32 (0x55897fff7fff0000ULL); - actual = vqdmulhs_lane_s32 (arg1, arg2, 0); + actual = test_vqdmulhs_0 (57336, vcreate_s32 (0x55897fff7fff0000ULL)); expected = 57334; + if (expected != actual) + { + fprintf (stderr, "Expected: %xd, got %xd\n", expected, actual); + abort (); + } + + actual = test_vqdmulhs_1 (57336, vcreate_s32 (0x55897fff7fff0000ULL)); + expected = 38315; + if (expected != actual) { fprintf (stderr, "Expected: %xd, got %xd\n", expected, actual); @@ -30,4 +46,5 @@ main (void) return 0; } -/* { dg-final { scan-assembler-times "sqdmulh\[ \t\]+\[sS\]\[0-9\]+, ?\[sS\]\[0-9\]+, ?\[vV\]\[0-9\]+\.\[sS\]\\\[0\\\]\n" 1 } } */ +/* { dg-final { scan-assembler-times "sqdmulh\[ \t\]+\[sS\]\[0-9\]+, ?\[sS\]\[0-9\]+, ?(?:\[sS\]\[0-9\]+|\[vV\]\[0-9\]+\.\[sS\]\\\[0\\\])\n" 1 } } */ +/* { dg-final { scan-assembler-times "sqdmulh\[ \t\]+\[sS\]\[0-9\]+, ?\[sS\]\[0-9\]+, ?\[vV\]\[0-9\]+\.\[sS\]\\\[1\\\]\n" 1 } } */ diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhs_laneq_s32_1.c b/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhs_laneq_s32_1.c index 71b260015ec..fd63bf9a4e9 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhs_laneq_s32_1.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vqdmulhs_laneq_s32_1.c @@ -1,25 +1,27 @@ /* Test the vqdmulhs_laneq_s32 AArch64 SIMD intrinsic. */ /* { dg-do run } */ -/* { dg-options "-save-temps -O3 -fno-inline" } */ +/* { dg-options "-save-temps -O3" } */ #include "arm_neon.h" extern void abort (void); +int32_t __attribute__((noipa)) +test_vqdmulhs (int32_t arg1, int32x4_t arg2) +{ + return vqdmulhs_laneq_s32 (arg1, arg2, 3); +} + int main (void) { - int32_t arg1; - int32x4_t arg2; int32_t actual; int32_t expected; - arg1 = 0x80000000; - arg2 = vcombine_s32 (vcreate_s32 (0x950dffffc4f40000ULL), - vcreate_s32 (0x7fff8000274a8000ULL)); - - actual = vqdmulhs_laneq_s32 (arg1, arg2, 3); + actual = test_vqdmulhs (0x80000000, + vcombine_s32 (vcreate_s32 (0x950dffffc4f40000ULL), + vcreate_s32 (0x7fff8000274a8000ULL))); expected = -2147450880; if (expected != actual) diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhh_lane_s16.c b/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhh_lane_s16.c index aca96d1fd6a..7dddb7550c4 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhh_lane_s16.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhh_lane_s16.c @@ -1,25 +1,26 @@ /* Test the vqrdmulhh_lane_s16 AArch64 SIMD intrinsic. */ /* { dg-do run } */ -/* { dg-options "-save-temps -O3 -fno-inline" } */ +/* { dg-options "-save-temps -O3" } */ #include "arm_neon.h" #include extern void abort (void); +int16_t __attribute__((noipa)) +test_vqrdmulhh (int16_t arg1, int16x4_t arg2) +{ + return vqrdmulhh_lane_s16 (arg1, arg2, 3); +} + int main (void) { - int16_t arg1; - int16x4_t arg2; - int16_t result; int16_t actual; int16_t expected; - arg1 = -32768; - arg2 = vcreate_s16 (0xd78e000005d78000ULL); - actual = vqrdmulhh_lane_s16 (arg1, arg2, 3); + actual = test_vqrdmulhh (-32768, vcreate_s16 (0xd78e000005d78000ULL)); expected = 10354; if (expected != actual) diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhh_laneq_s16_1.c b/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhh_laneq_s16_1.c index fd2c61deab8..78d6299d7d0 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhh_laneq_s16_1.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhh_laneq_s16_1.c @@ -1,25 +1,27 @@ /* Test the vqrdmulhh_laneq_s16 AArch64 SIMD intrinsic. */ /* { dg-do run } */ -/* { dg-options "-save-temps -O3 -fno-inline" } */ +/* { dg-options "-save-temps -O3" } */ #include "arm_neon.h" extern void abort (void); +int16_t __attribute__((noipa)) +test_vqrdmulhh (int16_t arg1, int16x8_t arg2) +{ + return vqrdmulhh_laneq_s16 (arg1, arg2, 7); +} + int main (void) { - int16_t arg1; - int16x8_t arg2; int16_t actual; int16_t expected; - arg1 = 0; - arg2 = vcombine_s16 (vcreate_s16 (0x7fffffffa7908000ULL), - vcreate_s16 (0x8000d2607fff0000ULL)); - - actual = vqrdmulhh_laneq_s16 (arg1, arg2, 7); + actual = test_vqrdmulhh (0, + vcombine_s16 (vcreate_s16 (0x7fffffffa7908000ULL), + vcreate_s16 (0x8000d2607fff0000ULL))); expected = 0; if (expected != actual) diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhs_lane_s32.c b/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhs_lane_s32.c index 30b21c9b4bc..827b52f33d1 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhs_lane_s32.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhs_lane_s32.c @@ -1,25 +1,26 @@ /* Test the vqrdmulhs_lane_s32 AArch64 SIMD intrinsic. */ /* { dg-do run } */ -/* { dg-options "-save-temps -O3 -fno-inline" } */ +/* { dg-options "-save-temps -O3" } */ #include "arm_neon.h" #include extern void abort (void); +int32_t __attribute__((noipa)) +test_vqrdmulhs (int32_t arg1, int32x2_t arg2) +{ + return vqrdmulhs_lane_s32 (arg1, arg2, 1); +} + int main (void) { - int32_t arg1; - int32x2_t arg2; - int32_t result; int32_t actual; int32_t expected; - arg1 = -2099281921; - arg2 = vcreate_s32 (0x000080007fff0000ULL); - actual = vqrdmulhs_lane_s32 (arg1, arg2, 1); + actual = test_vqrdmulhs (-2099281921, vcreate_s32 (0x000080007fff0000ULL)); expected = -32033; if (expected != actual) diff --git a/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhs_laneq_s32_1.c b/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhs_laneq_s32_1.c index 6d4e7648384..b06d16f20e9 100644 --- a/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhs_laneq_s32_1.c +++ b/gcc/testsuite/gcc.target/aarch64/simd/vqrdmulhs_laneq_s32_1.c @@ -1,25 +1,27 @@ /* Test the vqrdmulhs_laneq_s32 AArch64 SIMD intrinsic. */ /* { dg-do run } */ -/* { dg-options "-save-temps -O3 -fno-inline" } */ +/* { dg-options "-save-temps -O3" } */ #include "arm_neon.h" extern void abort (void); +int32_t __attribute__((noipa)) +test_vqrdmulhs (int32_t arg1, int32x4_t arg2) +{ + return vqrdmulhs_laneq_s32 (arg1, arg2, 3); +} + int main (void) { - int32_t arg1; - int32x4_t arg2; int32_t actual; int32_t expected; - arg1 = 32768; - arg2 = vcombine_s32 (vcreate_s32 (0x8000ffffffffcd5bULL), - vcreate_s32 (0x7fffffffffffffffULL)); - - actual = vqrdmulhs_laneq_s32 (arg1, arg2, 3); + actual = test_vqrdmulhs (32768, + vcombine_s32 (vcreate_s32 (0x8000ffffffffcd5bULL), + vcreate_s32 (0x7fffffffffffffffULL))); expected = 32768; if (expected != actual)