From patchwork Tue Jul 28 11:26:50 2015 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alan Lawrence X-Patchwork-Id: 501175 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Received: from sourceware.org (server1.sourceware.org [209.132.180.131]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id CDCE51402CA for ; Tue, 28 Jul 2015 21:27:38 +1000 (AEST) Authentication-Results: ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=gcc.gnu.org header.i=@gcc.gnu.org header.b=iNM7qcx3; dkim-atps=neutral DomainKey-Signature: a=rsa-sha1; c=nofws; d=gcc.gnu.org; h=list-id :list-unsubscribe:list-archive:list-post:list-help:sender :message-id:date:from:mime-version:to:subject:in-reply-to :content-type; q=dns; s=default; b=iWkNrHpIT5jcbs4fserykwahHN/HX XN1WUrFOh7JJAtCxixiewC+gM15wNwt6tytITY+XR5XStVFq5FBXDM8W7g0f+Pcw Moggm7dz8/wzfLegh3/RXYMfbDwOl0DdkYf1h58Kq+u7Glp8DRU1/YxtajKsSb/v 03Bud0LYKxXYok= DKIM-Signature: v=1; a=rsa-sha1; c=relaxed; d=gcc.gnu.org; h=list-id :list-unsubscribe:list-archive:list-post:list-help:sender :message-id:date:from:mime-version:to:subject:in-reply-to :content-type; s=default; bh=Ff5xcsFRGsflKdJ3m0YZy1WCPz8=; b=iNM 7qcx3UorHj579VndV+ZauG6i4Hp9LrG4e5cGwa0xpFm4u2wyTOiUYtIQaPsws+4/ hwWNmo7IsVFTB1wQk3Az/pV+5Zv6OgQcjx/UK/kQp4yNIb04m2OVhFDhbtQ46GnO THhN9tmC/ruPall0PGZD+qGnnZonQGG29uXm2KnU= Received: (qmail 105320 invoked by alias); 28 Jul 2015 11:27:28 -0000 Mailing-List: contact gcc-patches-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Id: List-Unsubscribe: List-Archive: List-Post: List-Help: Sender: gcc-patches-owner@gcc.gnu.org Delivered-To: mailing list gcc-patches@gcc.gnu.org Received: (qmail 105302 invoked by uid 89); 28 Jul 2015 11:27:27 -0000 Authentication-Results: sourceware.org; auth=none X-Virus-Found: No X-Spam-SWARE-Status: No, score=-0.2 required=5.0 tests=AWL, BAYES_50, KAM_LOTSOFHASH, SPF_PASS autolearn=no version=3.3.2 X-HELO: eu-smtp-delivery-143.mimecast.com Received: from eu-smtp-delivery-143.mimecast.com (HELO eu-smtp-delivery-143.mimecast.com) (146.101.78.143) by sourceware.org (qpsmtpd/0.93/v0.84-503-g423c35a) with ESMTP; Tue, 28 Jul 2015 11:27:02 +0000 Received: from cam-owa1.Emea.Arm.com (fw-tnat.cambridge.arm.com [217.140.96.140]) by eu-smtp-1.mimecast.com with ESMTP id uk-mta-13-G7c6_CJfQJ63GpZYw_b2xA-1; Tue, 28 Jul 2015 12:26:51 +0100 Received: from [10.2.207.65] ([10.1.2.79]) by cam-owa1.Emea.Arm.com with Microsoft SMTPSVC(6.0.3790.3959); Tue, 28 Jul 2015 12:26:50 +0100 Message-ID: <55B766FA.6000608@arm.com> Date: Tue, 28 Jul 2015 12:26:50 +0100 From: Alan Lawrence User-Agent: Thunderbird 2.0.0.24 (X11/20101213) MIME-Version: 1.0 To: "gcc-patches@gcc.gnu.org" Subject: [PATCH 13/15][ARM/AArch64 Testsuite] Add float16 tests to advsimd-intrinsics testsuite In-Reply-To: <55B765DF.4040706@arm.com> X-MC-Unique: G7c6_CJfQJ63GpZYw_b2xA-1 X-IsSubscribed: yes This is a respin of https://gcc.gnu.org/ml/gcc-patches/2015-07/msg00488.html, fixing up the testsuite for float16 vectors. Relative to the previous version, most of the additions to the tests are now within #if..#endif such that they are only compiled if we have a scalar __fp16 type (the exception is hfloat16_t: since this is actually an integer type, we can define and use it without any compiler fp16 support). Also we try to use add_options_for_arm_neon_fp16 for all tests (on ARM targets), falling back to add_options_for_arm_neon if the previous fails. Cross-tested on many multilibs, including -march=armv6, -march=armv7-a{,-mfpu=neon-fp16}, -march=armv7-a/-mfpu=neon, -march=armv7-a/-mfp16-format=none{,/-mfpu=neon-fp16,/-mfpu=neon}, -march=armv7-a/-mfp16-format=alternative . Note that on bigendian, this requires path at https://gcc.gnu.org/ml/gcc-patches/2015-07/msg00696.html , which I will commit at the same time. gcc/testsuite/ChangeLog: * gcc.target/aarch64/advsimd-intrinsics/advsimd-intrinsics.exp: Set additional_flags for neon-fp16 if supported, else fallback to neon. * gcc.target/aarch64/advsimd-intrinsics/arm-neon-ref.h (hfloat16_t): New. (result, expected, clean_results, DECL_VARIABLE_64BITS_VARIANTS, DECL_VARIABLE_128BITS_VARIANTS): Add float16x4_t and float16x8_t cases if supported. (CHECK_RESULTS): Redefine using CHECK_RESULTS_NAMED. (CHECK_RESULTS_NAMED): Move body to CHECK_RESULTS_NAMED_NO_FP16; redefine in terms of CHECK_RESULTS_NAMED_NO_FP16 with float16 variants when those are supported. (CHECK_RESULTS_NAMED_NO_FP16, CHECK_RESULTS_NO_FP16): New. (vdup_n_f16): New. * gcc.target/aarch64/advsimd-intrinsics/compute-ref-data.h (buffer, buffer_pad, buffer_dup, buffer_dup_pad): Add float16x4 and float16x8_t cases if supported. * gcc.target/aarch64/advsimd-intrinsics/vbsl.c (exec_vbsl): Use CHECK_RESULTS_NO_FP16 in place of CHECK_RESULTS. * gcc.target/aarch64/advsimd-intrinsics/vdup-vmov.c (exec_vdup_vmov): Likewise. * gcc.target/aarch64/advsimd-intrinsics/vdup_lane.c (exec_vdup_lane): Likewise. * gcc.target/aarch64/advsimd-intrinsics/vext.c (exec_vext): Likewise. * gcc.target/aarch64/advsimd-intrinsics/vcombine.c (expected): Add float16x8_t case. (main, exec_vcombine): test float16x4_t -> float16x8_t, if supported. * gcc.target/aarch64/advsimd-intrinsics/vcreate.c (expected, main, exec_vcreate): Likewise. * gcc.target/aarch64/advsimd-intrinsics/vget_high (expected, exec_vget_high): Likewise. * gcc.target/aarch64/advsimd-intrinsics/vget_low.c (expected, exec_vget_low): Likewise. * gcc.target/aarch64/advsimd-intrinsics/vld1.c (expected, exec_vld1): Likewise. * gcc.target/aarch64/advsimd-intrinsics/vld1_dup.c (expected, exec_vld1_dup): Likewise. * gcc.target/aarch64/advsimd-intrinsics/vld1_lane.c (expected, exec_vld1_lane): Likewise. * gcc.target/aarch64/advsimd-intrinsics/vldX.c (expected, exec_vldX): Likewise. * gcc.target/aarch64/advsimd-intrinsics/vldX_dup.c (expected, exec_vldX_dup): Likewise. * gcc.target/aarch64/advsimd-intrinsics/vldX_lane.c (expected, exec_vldX_lane): Likewise. * gcc.target/aarch64/advsimd-intrinsics/vset_lane.c (expected, exec_vset_lane): Likewise. * gcc.target/aarch64/advsimd-intrinsics/vst1_lane.c (expected, exec_vst1_lane): Likewise. diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/advsimd-intrinsics.exp b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/advsimd-intrinsics.exp index ceada839d982d3b6a922d924cad910a1c860eed0..462696315e05ea220dff60c1a605160ae2b59a1c 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/advsimd-intrinsics.exp +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/advsimd-intrinsics.exp @@ -52,8 +52,12 @@ if {[istarget arm*-*-*]} then { torture-init set-torture-options $C_TORTURE_OPTIONS {{}} $LTO_TORTURE_OPTIONS -# Make sure Neon flags are provided, if necessary. -set additional_flags [add_options_for_arm_neon ""] +# Make sure Neon flags are provided, if necessary. Use fp16 if we can. +if {[check_effective_target_arm_neon_fp16_ok]} then { + set additional_flags [add_options_for_arm_neon_fp16 ""] +} else { + set additional_flags [add_options_for_arm_neon ""] +} # Main loop. gcc-dg-runtest [lsort [glob -nocomplain $srcdir/$subdir/*.c]] \ diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/arm-neon-ref.h b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/arm-neon-ref.h index 4e728d5572c8e669bf6e175a07b6575cb6baf66d..49fbd843e507ede8aa81d02c175a82a1221750a4 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/arm-neon-ref.h +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/arm-neon-ref.h @@ -7,6 +7,7 @@ #include /* helper type, to help write floating point results in integer form. */ +typedef uint16_t hfloat16_t; typedef uint32_t hfloat32_t; typedef uint64_t hfloat64_t; @@ -132,6 +133,9 @@ static ARRAY(result, uint, 32, 2); static ARRAY(result, uint, 64, 1); static ARRAY(result, poly, 8, 8); static ARRAY(result, poly, 16, 4); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +static ARRAY(result, float, 16, 4); +#endif static ARRAY(result, float, 32, 2); static ARRAY(result, int, 8, 16); static ARRAY(result, int, 16, 8); @@ -143,6 +147,9 @@ static ARRAY(result, uint, 32, 4); static ARRAY(result, uint, 64, 2); static ARRAY(result, poly, 8, 16); static ARRAY(result, poly, 16, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +static ARRAY(result, float, 16, 8); +#endif static ARRAY(result, float, 32, 4); #ifdef __aarch64__ static ARRAY(result, float, 64, 2); @@ -160,6 +167,7 @@ extern ARRAY(expected, uint, 32, 2); extern ARRAY(expected, uint, 64, 1); extern ARRAY(expected, poly, 8, 8); extern ARRAY(expected, poly, 16, 4); +extern ARRAY(expected, hfloat, 16, 4); extern ARRAY(expected, hfloat, 32, 2); extern ARRAY(expected, int, 8, 16); extern ARRAY(expected, int, 16, 8); @@ -171,38 +179,11 @@ extern ARRAY(expected, uint, 32, 4); extern ARRAY(expected, uint, 64, 2); extern ARRAY(expected, poly, 8, 16); extern ARRAY(expected, poly, 16, 8); +extern ARRAY(expected, hfloat, 16, 8); extern ARRAY(expected, hfloat, 32, 4); extern ARRAY(expected, hfloat, 64, 2); -/* Check results. Operates on all possible vector types. */ -#define CHECK_RESULTS(test_name,comment) \ - { \ - CHECK(test_name, int, 8, 8, PRIx8, expected, comment); \ - CHECK(test_name, int, 16, 4, PRIx16, expected, comment); \ - CHECK(test_name, int, 32, 2, PRIx32, expected, comment); \ - CHECK(test_name, int, 64, 1, PRIx64, expected, comment); \ - CHECK(test_name, uint, 8, 8, PRIx8, expected, comment); \ - CHECK(test_name, uint, 16, 4, PRIx16, expected, comment); \ - CHECK(test_name, uint, 32, 2, PRIx32, expected, comment); \ - CHECK(test_name, uint, 64, 1, PRIx64, expected, comment); \ - CHECK(test_name, poly, 8, 8, PRIx8, expected, comment); \ - CHECK(test_name, poly, 16, 4, PRIx16, expected, comment); \ - CHECK_FP(test_name, float, 32, 2, PRIx32, expected, comment); \ - \ - CHECK(test_name, int, 8, 16, PRIx8, expected, comment); \ - CHECK(test_name, int, 16, 8, PRIx16, expected, comment); \ - CHECK(test_name, int, 32, 4, PRIx32, expected, comment); \ - CHECK(test_name, int, 64, 2, PRIx64, expected, comment); \ - CHECK(test_name, uint, 8, 16, PRIx8, expected, comment); \ - CHECK(test_name, uint, 16, 8, PRIx16, expected, comment); \ - CHECK(test_name, uint, 32, 4, PRIx32, expected, comment); \ - CHECK(test_name, uint, 64, 2, PRIx64, expected, comment); \ - CHECK(test_name, poly, 8, 16, PRIx8, expected, comment); \ - CHECK(test_name, poly, 16, 8, PRIx16, expected, comment); \ - CHECK_FP(test_name, float, 32, 4, PRIx32, expected, comment); \ - } \ - -#define CHECK_RESULTS_NAMED(test_name,EXPECTED,comment) \ +#define CHECK_RESULTS_NAMED_NO_FP16(test_name,EXPECTED,comment) \ { \ CHECK(test_name, int, 8, 8, PRIx8, EXPECTED, comment); \ CHECK(test_name, int, 16, 4, PRIx16, EXPECTED, comment); \ @@ -229,6 +210,24 @@ extern ARRAY(expected, hfloat, 64, 2); CHECK_FP(test_name, float, 32, 4, PRIx32, EXPECTED, comment); \ } \ +/* Check results against EXPECTED. Operates on all possible vector types. */ +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define CHECK_RESULTS_NAMED(test_name,EXPECTED,comment) \ + { \ + CHECK_RESULTS_NAMED_NO_FP16(test_name, EXPECTED, comment) \ + CHECK_FP(test_name, float, 16, 4, PRIx16, EXPECTED, comment); \ + CHECK_FP(test_name, float, 16, 8, PRIx16, EXPECTED, comment); \ + } +#else +#define CHECK_RESULTS_NAMED(test_name,EXPECTED,comment) \ + CHECK_RESULTS_NAMED_NO_FP16(test_name, EXPECTED, comment) +#endif + +#define CHECK_RESULTS_NO_FP16(test_name,comment) \ + CHECK_RESULTS_NAMED_NO_FP16(test_name, expected, comment) + +#define CHECK_RESULTS(test_name,comment) \ + CHECK_RESULTS_NAMED(test_name, expected, comment) #if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__ @@ -380,6 +379,9 @@ static void clean_results (void) CLEAN(result, uint, 64, 1); CLEAN(result, poly, 8, 8); CLEAN(result, poly, 16, 4); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + CLEAN(result, float, 16, 4); +#endif CLEAN(result, float, 32, 2); CLEAN(result, int, 8, 16); @@ -392,6 +394,9 @@ static void clean_results (void) CLEAN(result, uint, 64, 2); CLEAN(result, poly, 8, 16); CLEAN(result, poly, 16, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + CLEAN(result, float, 16, 8); +#endif CLEAN(result, float, 32, 4); #if defined(__aarch64__) @@ -443,21 +448,40 @@ static void clean_results (void) DECL_VARIABLE(VAR, uint, 64, 2) /* Declare all 64 bits variants. */ +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define DECL_VARIABLE_64BITS_VARIANTS(VAR) \ + DECL_VARIABLE_64BITS_SIGNED_VARIANTS(VAR); \ + DECL_VARIABLE_64BITS_UNSIGNED_VARIANTS(VAR); \ + DECL_VARIABLE(VAR, poly, 8, 8); \ + DECL_VARIABLE(VAR, poly, 16, 4); \ + DECL_VARIABLE(VAR, float, 16, 4); \ + DECL_VARIABLE(VAR, float, 32, 2) +#else #define DECL_VARIABLE_64BITS_VARIANTS(VAR) \ DECL_VARIABLE_64BITS_SIGNED_VARIANTS(VAR); \ DECL_VARIABLE_64BITS_UNSIGNED_VARIANTS(VAR); \ DECL_VARIABLE(VAR, poly, 8, 8); \ DECL_VARIABLE(VAR, poly, 16, 4); \ DECL_VARIABLE(VAR, float, 32, 2) +#endif /* Declare all 128 bits variants. */ +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) #define DECL_VARIABLE_128BITS_VARIANTS(VAR) \ DECL_VARIABLE_128BITS_SIGNED_VARIANTS(VAR); \ DECL_VARIABLE_128BITS_UNSIGNED_VARIANTS(VAR); \ DECL_VARIABLE(VAR, poly, 8, 16); \ DECL_VARIABLE(VAR, poly, 16, 8); \ + DECL_VARIABLE(VAR, float, 16, 8); \ DECL_VARIABLE(VAR, float, 32, 4) - +#else +#define DECL_VARIABLE_128BITS_VARIANTS(VAR) \ + DECL_VARIABLE_128BITS_SIGNED_VARIANTS(VAR); \ + DECL_VARIABLE_128BITS_UNSIGNED_VARIANTS(VAR); \ + DECL_VARIABLE(VAR, poly, 8, 16); \ + DECL_VARIABLE(VAR, poly, 16, 8); \ + DECL_VARIABLE(VAR, float, 32, 4) +#endif /* Declare all variants. */ #define DECL_VARIABLE_ALL_VARIANTS(VAR) \ DECL_VARIABLE_64BITS_VARIANTS(VAR); \ @@ -476,6 +500,15 @@ static void clean_results (void) /* Helpers to initialize vectors. */ #define VDUP(VAR, Q, T1, T2, W, N, V) \ VECT_VAR(VAR, T1, W, N) = vdup##Q##_n_##T2##W(V) +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +/* Work around that there is no vdup_n_f16 intrinsic. */ +#define vdup_n_f16(VAL) \ + __extension__ \ + ({ \ + float16_t f = VAL; \ + vld1_dup_f16(&f); \ + }) +#endif #define VSET_LANE(VAR, Q, T1, T2, W, N, L, V) \ VECT_VAR(VAR, T1, W, N) = vset##Q##_lane_##T2##W(V, \ diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/compute-ref-data.h b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/compute-ref-data.h index 26203cc0a69372e6a5d826f3c2b6663a65686a37..c8d43367bef3845bedc5a3cc35255db58a987e70 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/compute-ref-data.h +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/compute-ref-data.h @@ -118,6 +118,10 @@ VECT_VAR_DECL_INIT(buffer, uint, 32, 2); PAD(buffer_pad, uint, 32, 2); VECT_VAR_DECL_INIT(buffer, uint, 64, 1); PAD(buffer_pad, uint, 64, 1); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +VECT_VAR_DECL_INIT(buffer, float, 16, 4); +PAD(buffer_pad, float, 16, 4); +#endif VECT_VAR_DECL_INIT(buffer, float, 32, 2); PAD(buffer_pad, float, 32, 2); VECT_VAR_DECL_INIT(buffer, int, 8, 16); @@ -140,6 +144,10 @@ VECT_VAR_DECL_INIT(buffer, poly, 8, 16); PAD(buffer_pad, poly, 8, 16); VECT_VAR_DECL_INIT(buffer, poly, 16, 8); PAD(buffer_pad, poly, 16, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +VECT_VAR_DECL_INIT(buffer, float, 16, 8); +PAD(buffer_pad, float, 16, 8); +#endif VECT_VAR_DECL_INIT(buffer, float, 32, 4); PAD(buffer_pad, float, 32, 4); #ifdef __aarch64__ @@ -170,6 +178,10 @@ VECT_VAR_DECL_INIT(buffer_dup, poly, 8, 8); VECT_VAR_DECL(buffer_dup_pad, poly, 8, 8); VECT_VAR_DECL_INIT(buffer_dup, poly, 16, 4); VECT_VAR_DECL(buffer_dup_pad, poly, 16, 4); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +VECT_VAR_DECL_INIT4(buffer_dup, float, 16, 4); +VECT_VAR_DECL(buffer_dup_pad, float, 16, 4); +#endif VECT_VAR_DECL_INIT4(buffer_dup, float, 32, 2); VECT_VAR_DECL(buffer_dup_pad, float, 32, 2); @@ -193,5 +205,9 @@ VECT_VAR_DECL_INIT(buffer_dup, poly, 8, 16); VECT_VAR_DECL(buffer_dup_pad, poly, 8, 16); VECT_VAR_DECL_INIT(buffer_dup, poly, 16, 8); VECT_VAR_DECL(buffer_dup_pad, poly, 16, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +VECT_VAR_DECL_INIT(buffer_dup, float, 16, 8); +VECT_VAR_DECL(buffer_dup_pad, float, 16, 8); +#endif VECT_VAR_DECL_INIT(buffer_dup, float, 32, 4); VECT_VAR_DECL(buffer_dup_pad, float, 32, 4); diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vbsl.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vbsl.c index bb17f0a9649beedaf0c054679b3efc1a86463c35..c4fdbb45102a5407d71cdc75043e74dd0ebb076e 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vbsl.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vbsl.c @@ -114,7 +114,7 @@ void exec_vbsl (void) TEST_VBSL(uint, , float, f, 32, 2); TEST_VBSL(uint, q, float, f, 32, 4); - CHECK_RESULTS (TEST_MSG, ""); + CHECK_RESULTS_NO_FP16 (TEST_MSG, ""); } int main (void) diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vcombine.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vcombine.c index 295768a0348185c663994c8361d063454cc8467a..5100375e5fe0c1f1f6b1e0cbff549990d73948e3 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vcombine.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vcombine.c @@ -27,6 +27,8 @@ VECT_VAR_DECL(expected,poly,16,8) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3, 0x66, 0x66, 0x66, 0x66 }; VECT_VAR_DECL(expected,hfloat,32,4) [] = { 0xc1800000, 0xc1700000, 0x40533333, 0x40533333 }; +VECT_VAR_DECL(expected,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80, + 0x4080, 0x4080, 0x4080, 0x4080 }; #define TEST_MSG "VCOMBINE" void exec_vcombine (void) @@ -44,6 +46,9 @@ void exec_vcombine (void) /* Initialize input "vector64_a" from "buffer". */ TEST_MACRO_64BITS_VARIANTS_2_5(VLOAD, vector64_a, buffer); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VLOAD(vector64_a, buffer, , float, f, 16, 4); +#endif VLOAD(vector64_a, buffer, , float, f, 32, 2); /* Choose init value arbitrarily. */ @@ -57,6 +62,9 @@ void exec_vcombine (void) VDUP(vector64_b, , uint, u, 64, 1, 0x88); VDUP(vector64_b, , poly, p, 8, 8, 0x55); VDUP(vector64_b, , poly, p, 16, 4, 0x66); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VDUP(vector64_b, , float, f, 16, 4, 2.25); +#endif VDUP(vector64_b, , float, f, 32, 2, 3.3f); clean_results (); @@ -72,6 +80,9 @@ void exec_vcombine (void) TEST_VCOMBINE(uint, u, 64, 1, 2); TEST_VCOMBINE(poly, p, 8, 8, 16); TEST_VCOMBINE(poly, p, 16, 4, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + TEST_VCOMBINE(float, f, 16, 4, 8); +#endif TEST_VCOMBINE(float, f, 32, 2, 4); CHECK(TEST_MSG, int, 8, 16, PRIx8, expected, ""); @@ -84,6 +95,9 @@ void exec_vcombine (void) CHECK(TEST_MSG, uint, 64, 2, PRIx64, expected, ""); CHECK(TEST_MSG, poly, 8, 16, PRIx8, expected, ""); CHECK(TEST_MSG, poly, 16, 8, PRIx16, expected, ""); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + CHECK_FP(TEST_MSG, float, 16, 8, PRIx16, expected, ""); +#endif CHECK_FP(TEST_MSG, float, 32, 4, PRIx32, expected, ""); } diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vcreate.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vcreate.c index b2289d3a628245835f8359c8770d83b15b13e931..b8b338ef3c06ff2489b525e22760cbaed1fda335 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vcreate.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vcreate.c @@ -16,6 +16,7 @@ VECT_VAR_DECL(expected,uint,64,1) [] = { 0x123456789abcdef0 }; VECT_VAR_DECL(expected,poly,8,8) [] = { 0xf0, 0xde, 0xbc, 0x9a, 0x78, 0x56, 0x34, 0x12 }; VECT_VAR_DECL(expected,poly,16,4) [] = { 0xdef0, 0x9abc, 0x5678, 0x1234 }; +VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xdef0, 0x9abc, 0x5678, 0x1234 }; VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0x9abcdef0, 0x12345678 }; #define INSN_NAME vcreate @@ -38,6 +39,9 @@ FNNAME (INSN_NAME) DECL_VAL(val, int, 16, 4); DECL_VAL(val, int, 32, 2); DECL_VAL(val, int, 64, 1); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + DECL_VAL(val, float, 16, 4); +#endif DECL_VAL(val, float, 32, 2); DECL_VAL(val, uint, 8, 8); DECL_VAL(val, uint, 16, 4); @@ -50,6 +54,9 @@ FNNAME (INSN_NAME) DECL_VARIABLE(vector_res, int, 16, 4); DECL_VARIABLE(vector_res, int, 32, 2); DECL_VARIABLE(vector_res, int, 64, 1); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + DECL_VARIABLE(vector_res, float, 16, 4); +#endif DECL_VARIABLE(vector_res, float, 32, 2); DECL_VARIABLE(vector_res, uint, 8, 8); DECL_VARIABLE(vector_res, uint, 16, 4); @@ -65,6 +72,9 @@ FNNAME (INSN_NAME) VECT_VAR(val, int, 16, 4) = 0x123456789abcdef0LL; VECT_VAR(val, int, 32, 2) = 0x123456789abcdef0LL; VECT_VAR(val, int, 64, 1) = 0x123456789abcdef0LL; +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VECT_VAR(val, float, 16, 4) = 0x123456789abcdef0LL; +#endif VECT_VAR(val, float, 32, 2) = 0x123456789abcdef0LL; VECT_VAR(val, uint, 8, 8) = 0x123456789abcdef0ULL; VECT_VAR(val, uint, 16, 4) = 0x123456789abcdef0ULL; @@ -76,6 +86,9 @@ FNNAME (INSN_NAME) TEST_VCREATE(int, s, 8, 8); TEST_VCREATE(int, s, 16, 4); TEST_VCREATE(int, s, 32, 2); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + TEST_VCREATE(float, f, 16, 4); +#endif TEST_VCREATE(float, f, 32, 2); TEST_VCREATE(int, s, 64, 1); TEST_VCREATE(uint, u, 8, 8); @@ -95,6 +108,9 @@ FNNAME (INSN_NAME) CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected, ""); CHECK(TEST_MSG, poly, 8, 8, PRIx8, expected, ""); CHECK(TEST_MSG, poly, 16, 4, PRIx16, expected, ""); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + CHECK_FP(TEST_MSG, float, 16, 4, PRIx16, expected, ""); +#endif CHECK_FP(TEST_MSG, float, 32, 2, PRIx32, expected, ""); } diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vdup-vmov.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vdup-vmov.c index b5132f41ac424fa1c9c59805efb4a43f2c3dab12..22d45d56c8e3c517da8cc595f767f8034aeabde8 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vdup-vmov.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vdup-vmov.c @@ -187,13 +187,13 @@ void exec_vdup_vmov (void) switch (i) { case 0: - CHECK_RESULTS_NAMED (TEST_MSG, expected0, ""); + CHECK_RESULTS_NAMED_NO_FP16 (TEST_MSG, expected0, ""); break; case 1: - CHECK_RESULTS_NAMED (TEST_MSG, expected1, ""); + CHECK_RESULTS_NAMED_NO_FP16 (TEST_MSG, expected1, ""); break; case 2: - CHECK_RESULTS_NAMED (TEST_MSG, expected2, ""); + CHECK_RESULTS_NAMED_NO_FP16 (TEST_MSG, expected2, ""); break; default: abort(); @@ -232,13 +232,13 @@ void exec_vdup_vmov (void) switch (i) { case 0: - CHECK_RESULTS_NAMED (TEST_MSG, expected0, ""); + CHECK_RESULTS_NAMED_NO_FP16 (TEST_MSG, expected0, ""); break; case 1: - CHECK_RESULTS_NAMED (TEST_MSG, expected1, ""); + CHECK_RESULTS_NAMED_NO_FP16 (TEST_MSG, expected1, ""); break; case 2: - CHECK_RESULTS_NAMED (TEST_MSG, expected2, ""); + CHECK_RESULTS_NAMED_NO_FP16 (TEST_MSG, expected2, ""); break; default: abort(); diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vdup_lane.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vdup_lane.c index c1ff6dd3007100966a488a33bc714802b987a084..ef708dcba17e68ccc0e3540c52bb6507d0562fad 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vdup_lane.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vdup_lane.c @@ -90,7 +90,7 @@ void exec_vdup_lane (void) TEST_VDUP_LANE(q, poly, p, 16, 8, 4, 1); TEST_VDUP_LANE(q, float, f, 32, 4, 2, 1); - CHECK_RESULTS (TEST_MSG, ""); + CHECK_RESULTS_NO_FP16 (TEST_MSG, ""); } int main (void) diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vext.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vext.c index 0b014ebda87e1486ea6005d597f6dc30b5dfc470..98f88a69898130dc922e205717d1dcd82c77aa11 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vext.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vext.c @@ -113,7 +113,7 @@ void exec_vext (void) TEST_VEXT(q, poly, p, 16, 8, 6); TEST_VEXT(q, float, f, 32, 4, 3); - CHECK_RESULTS (TEST_MSG, ""); + CHECK_RESULTS_NO_FP16 (TEST_MSG, ""); } int main (void) diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vget_high.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vget_high.c index d7581125edd133877cedbdf2bb7262d9f93337fd..9f0a1687f189bc2dfbe111c7f5c3b96c9acecd52 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vget_high.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vget_high.c @@ -16,6 +16,7 @@ VECT_VAR_DECL(expected,uint,64,1) [] = { 0xfffffffffffffff1 }; VECT_VAR_DECL(expected,poly,8,8) [] = { 0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xff }; VECT_VAR_DECL(expected,poly,16,4) [] = { 0xfff4, 0xfff5, 0xfff6, 0xfff7 }; +VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xca00, 0xc980, 0xc900, 0xc880 }; VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 }; #define TEST_MSG "VGET_HIGH" @@ -31,6 +32,9 @@ void exec_vget_high (void) DECL_VARIABLE_128BITS_VARIANTS(vector128); TEST_MACRO_128BITS_VARIANTS_2_5(VLOAD, vector128, buffer); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VLOAD(vector128, buffer, q, float, f, 16, 8); +#endif VLOAD(vector128, buffer, q, float, f, 32, 4); clean_results (); @@ -46,6 +50,9 @@ void exec_vget_high (void) TEST_VGET_HIGH(uint, u, 64, 1, 2); TEST_VGET_HIGH(poly, p, 8, 8, 16); TEST_VGET_HIGH(poly, p, 16, 4, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + TEST_VGET_HIGH(float, f, 16, 4, 8); +#endif TEST_VGET_HIGH(float, f, 32, 2, 4); CHECK(TEST_MSG, int, 8, 8, PRIx8, expected, ""); diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vget_low.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vget_low.c index 12ecfc21ba03225ddd9c9d88eec64c2f6774227f..2b875b9b7b86d850647cf977086f336b932cfb0b 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vget_low.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vget_low.c @@ -16,6 +16,7 @@ VECT_VAR_DECL(expected,uint,64,1) [] = { 0xfffffffffffffff0 }; VECT_VAR_DECL(expected,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7 }; VECT_VAR_DECL(expected,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 }; +VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 }; VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 }; #define TEST_MSG "VGET_LOW" @@ -31,6 +32,9 @@ void exec_vget_low (void) DECL_VARIABLE_128BITS_VARIANTS(vector128); TEST_MACRO_128BITS_VARIANTS_2_5(VLOAD, vector128, buffer); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VLOAD(vector128, buffer, q, float, f, 16, 8); +#endif VLOAD(vector128, buffer, q, float, f, 32, 4); clean_results (); @@ -46,6 +50,9 @@ void exec_vget_low (void) TEST_VGET_LOW(uint, u, 64, 1, 2); TEST_VGET_LOW(poly, p, 8, 8, 16); TEST_VGET_LOW(poly, p, 16, 4, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + TEST_VGET_LOW(float, f, 16, 4, 8); +#endif TEST_VGET_LOW(float, f, 32, 2, 4); CHECK(TEST_MSG, int, 8, 8, PRIx8, expected, ""); @@ -58,6 +65,9 @@ void exec_vget_low (void) CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected, ""); CHECK(TEST_MSG, poly, 8, 8, PRIx8, expected, ""); CHECK(TEST_MSG, poly, 16, 4, PRIx16, expected, ""); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + CHECK_FP(TEST_MSG, float, 16, 4, PRIx16, expected, ""); +#endif CHECK_FP(TEST_MSG, float, 32, 2, PRIx32, expected, ""); } diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vld1.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vld1.c index ced9d736d6d22cbd93352c1972fd9bab81fa747e..4ed0e464f9ce6b0f599a6a72d3f49db5ac9a0374 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vld1.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vld1.c @@ -16,6 +16,7 @@ VECT_VAR_DECL(expected,uint,64,1) [] = { 0xfffffffffffffff0 }; VECT_VAR_DECL(expected,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7 }; VECT_VAR_DECL(expected,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 }; +VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 }; VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 }; VECT_VAR_DECL(expected,int,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, @@ -44,6 +45,8 @@ VECT_VAR_DECL(expected,poly,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xfc, 0xfd, 0xfe, 0xff }; VECT_VAR_DECL(expected,poly,16,8) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3, 0xfff4, 0xfff5, 0xfff6, 0xfff7 }; +VECT_VAR_DECL(expected,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80, + 0xca00, 0xc980, 0xc900, 0xc880 }; VECT_VAR_DECL(expected,hfloat,32,4) [] = { 0xc1800000, 0xc1700000, 0xc1600000, 0xc1500000 }; @@ -62,6 +65,10 @@ void exec_vld1 (void) TEST_MACRO_ALL_VARIANTS_2_5(TEST_VLD1, vector, buffer); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + TEST_VLD1(vector, buffer, , float, f, 16, 4); + TEST_VLD1(vector, buffer, q, float, f, 16, 8); +#endif TEST_VLD1(vector, buffer, , float, f, 32, 2); TEST_VLD1(vector, buffer, q, float, f, 32, 4); diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vld1_dup.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vld1_dup.c index 0e052743926ca7cdc3108a4e6ef96b54e815de1d..34be214e9122c5701a341b09479443fdb5f2716b 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vld1_dup.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vld1_dup.c @@ -17,6 +17,7 @@ VECT_VAR_DECL(expected0,uint,64,1) [] = { 0xfffffffffffffff0 }; VECT_VAR_DECL(expected0,poly,8,8) [] = { 0xf0, 0xf0, 0xf0, 0xf0, 0xf0, 0xf0, 0xf0, 0xf0 }; VECT_VAR_DECL(expected0,poly,16,4) [] = { 0xfff0, 0xfff0, 0xfff0, 0xfff0 }; +VECT_VAR_DECL(expected0,hfloat,16,4) [] = { 0xcc00, 0xcc00, 0xcc00, 0xcc00 }; VECT_VAR_DECL(expected0,hfloat,32,2) [] = { 0xc1800000, 0xc1800000 }; VECT_VAR_DECL(expected0,int,8,16) [] = { 0xf0, 0xf0, 0xf0, 0xf0, 0xf0, 0xf0, 0xf0, 0xf0, @@ -44,6 +45,8 @@ VECT_VAR_DECL(expected0,poly,8,16) [] = { 0xf0, 0xf0, 0xf0, 0xf0, 0xf0, 0xf0, 0xf0, 0xf0 }; VECT_VAR_DECL(expected0,poly,16,8) [] = { 0xfff0, 0xfff0, 0xfff0, 0xfff0, 0xfff0, 0xfff0, 0xfff0, 0xfff0 }; +VECT_VAR_DECL(expected0,hfloat,16,8) [] = { 0xcc00, 0xcc00, 0xcc00, 0xcc00, + 0xcc00, 0xcc00, 0xcc00, 0xcc00 }; VECT_VAR_DECL(expected0,hfloat,32,4) [] = { 0xc1800000, 0xc1800000, 0xc1800000, 0xc1800000 }; @@ -61,6 +64,7 @@ VECT_VAR_DECL(expected1,uint,64,1) [] = { 0xfffffffffffffff1 }; VECT_VAR_DECL(expected1,poly,8,8) [] = { 0xf1, 0xf1, 0xf1, 0xf1, 0xf1, 0xf1, 0xf1, 0xf1 }; VECT_VAR_DECL(expected1,poly,16,4) [] = { 0xfff1, 0xfff1, 0xfff1, 0xfff1 }; +VECT_VAR_DECL(expected1,hfloat,16,4) [] = { 0xcb80, 0xcb80, 0xcb80, 0xcb80 }; VECT_VAR_DECL(expected1,hfloat,32,2) [] = { 0xc1700000, 0xc1700000 }; VECT_VAR_DECL(expected1,int,8,16) [] = { 0xf1, 0xf1, 0xf1, 0xf1, 0xf1, 0xf1, 0xf1, 0xf1, @@ -88,6 +92,8 @@ VECT_VAR_DECL(expected1,poly,8,16) [] = { 0xf1, 0xf1, 0xf1, 0xf1, 0xf1, 0xf1, 0xf1, 0xf1 }; VECT_VAR_DECL(expected1,poly,16,8) [] = { 0xfff1, 0xfff1, 0xfff1, 0xfff1, 0xfff1, 0xfff1, 0xfff1, 0xfff1 }; +VECT_VAR_DECL(expected1,hfloat,16,8) [] = { 0xcb80, 0xcb80, 0xcb80, 0xcb80, + 0xcb80, 0xcb80, 0xcb80, 0xcb80 }; VECT_VAR_DECL(expected1,hfloat,32,4) [] = { 0xc1700000, 0xc1700000, 0xc1700000, 0xc1700000 }; @@ -105,6 +111,7 @@ VECT_VAR_DECL(expected2,uint,64,1) [] = { 0xfffffffffffffff2 }; VECT_VAR_DECL(expected2,poly,8,8) [] = { 0xf2, 0xf2, 0xf2, 0xf2, 0xf2, 0xf2, 0xf2, 0xf2 }; VECT_VAR_DECL(expected2,poly,16,4) [] = { 0xfff2, 0xfff2, 0xfff2, 0xfff2 }; +VECT_VAR_DECL(expected2,hfloat,16,4) [] = { 0xcb00, 0xcb00, 0xcb00, 0xcb00 }; VECT_VAR_DECL(expected2,hfloat,32,2) [] = { 0xc1600000, 0xc1600000 }; VECT_VAR_DECL(expected2,int,8,16) [] = { 0xf2, 0xf2, 0xf2, 0xf2, 0xf2, 0xf2, 0xf2, 0xf2, @@ -132,6 +139,8 @@ VECT_VAR_DECL(expected2,poly,8,16) [] = { 0xf2, 0xf2, 0xf2, 0xf2, 0xf2, 0xf2, 0xf2, 0xf2 }; VECT_VAR_DECL(expected2,poly,16,8) [] = { 0xfff2, 0xfff2, 0xfff2, 0xfff2, 0xfff2, 0xfff2, 0xfff2, 0xfff2 }; +VECT_VAR_DECL(expected2,hfloat,16,8) [] = { 0xcb00, 0xcb00, 0xcb00, 0xcb00, + 0xcb00, 0xcb00, 0xcb00, 0xcb00 }; VECT_VAR_DECL(expected2,hfloat,32,4) [] = { 0xc1600000, 0xc1600000, 0xc1600000, 0xc1600000 }; @@ -154,6 +163,10 @@ void exec_vld1_dup (void) TEST_MACRO_ALL_VARIANTS_2_5(TEST_VLD1_DUP, vector, buffer_dup); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + TEST_VLD1_DUP(vector, buffer_dup, , float, f, 16, 4); + TEST_VLD1_DUP(vector, buffer_dup, q, float, f, 16, 8); +#endif TEST_VLD1_DUP(vector, buffer_dup, , float, f, 32, 2); TEST_VLD1_DUP(vector, buffer_dup, q, float, f, 32, 4); diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vld1_lane.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vld1_lane.c index d5c5d22a8ce80cd7e4e6e46750b6c5297845f366..1f39006498d16bada53c42d298c080c6ba2556d1 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vld1_lane.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vld1_lane.c @@ -16,6 +16,7 @@ VECT_VAR_DECL(expected,uint,64,1) [] = { 0xfffffffffffffff0 }; VECT_VAR_DECL(expected,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xf0 }; VECT_VAR_DECL(expected,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xfff0 }; +VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xcc00, 0xaaaa }; VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0xaaaaaaaa, 0xc1800000 }; VECT_VAR_DECL(expected,int,8,16) [] = { 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, @@ -43,6 +44,8 @@ VECT_VAR_DECL(expected,poly,8,16) [] = { 0xaa, 0xaa, 0xaa, 0xaa, 0xf0, 0xaa, 0xaa, 0xaa }; VECT_VAR_DECL(expected,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xfff0, 0xaaaa }; +VECT_VAR_DECL(expected,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, + 0xaaaa, 0xcc00, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa, 0xc1800000, 0xaaaaaaaa }; @@ -72,6 +75,9 @@ void exec_vld1_lane (void) ARRAY(buffer_src, uint, 64, 1); ARRAY(buffer_src, poly, 8, 8); ARRAY(buffer_src, poly, 16, 4); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + ARRAY(buffer_src, float, 16, 4); +#endif ARRAY(buffer_src, float, 32, 2); ARRAY(buffer_src, int, 8, 16); @@ -84,6 +90,9 @@ void exec_vld1_lane (void) ARRAY(buffer_src, uint, 64, 2); ARRAY(buffer_src, poly, 8, 16); ARRAY(buffer_src, poly, 16, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + ARRAY(buffer_src, float, 16, 8); +#endif ARRAY(buffer_src, float, 32, 4); clean_results (); @@ -99,6 +108,9 @@ void exec_vld1_lane (void) TEST_VLD1_LANE(, uint, u, 64, 1, 0); TEST_VLD1_LANE(, poly, p, 8, 8, 7); TEST_VLD1_LANE(, poly, p, 16, 4, 3); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + TEST_VLD1_LANE(, float, f, 16, 4, 2); +#endif TEST_VLD1_LANE(, float, f, 32, 2, 1); TEST_VLD1_LANE(q, int, s, 8, 16, 15); @@ -111,6 +123,9 @@ void exec_vld1_lane (void) TEST_VLD1_LANE(q, uint, u, 64, 2, 0); TEST_VLD1_LANE(q, poly, p, 8, 16, 12); TEST_VLD1_LANE(q, poly, p, 16, 8, 6); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + TEST_VLD1_LANE(q, float, f, 16, 8, 5); +#endif TEST_VLD1_LANE(q, float, f, 32, 4, 2); CHECK_RESULTS (TEST_MSG, ""); diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vldX.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vldX.c index f20aa03f51b770db68bc2e48181fcfb50f025f55..1e02dc3fa103ca6f71325f668eb04f0b88b594a3 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vldX.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vldX.c @@ -18,6 +18,7 @@ VECT_VAR_DECL(expected_vld2_0,uint,64,1) [] = { 0xfffffffffffffff0 }; VECT_VAR_DECL(expected_vld2_0,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7 }; VECT_VAR_DECL(expected_vld2_0,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 }; +VECT_VAR_DECL(expected_vld2_0,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 }; VECT_VAR_DECL(expected_vld2_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 }; VECT_VAR_DECL(expected_vld2_0,int,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, @@ -41,6 +42,8 @@ VECT_VAR_DECL(expected_vld2_0,poly,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xfc, 0xfd, 0xfe, 0xff }; VECT_VAR_DECL(expected_vld2_0,poly,16,8) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3, 0xfff4, 0xfff5, 0xfff6, 0xfff7 }; +VECT_VAR_DECL(expected_vld2_0,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80, + 0xca00, 0xc980, 0xc900, 0xc880 }; VECT_VAR_DECL(expected_vld2_0,hfloat,32,4) [] = { 0xc1800000, 0xc1700000, 0xc1600000, 0xc1500000 }; @@ -58,6 +61,7 @@ VECT_VAR_DECL(expected_vld2_1,uint,64,1) [] = { 0xfffffffffffffff1 }; VECT_VAR_DECL(expected_vld2_1,poly,8,8) [] = { 0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xff }; VECT_VAR_DECL(expected_vld2_1,poly,16,4) [] = { 0xfff4, 0xfff5, 0xfff6, 0xfff7 }; +VECT_VAR_DECL(expected_vld2_1,hfloat,16,4) [] = { 0xca00, 0xc980, 0xc900, 0xc880 }; VECT_VAR_DECL(expected_vld2_1,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 }; VECT_VAR_DECL(expected_vld2_1,int,8,16) [] = { 0x0, 0x1, 0x2, 0x3, 0x4, 0x5, 0x6, 0x7, @@ -81,6 +85,8 @@ VECT_VAR_DECL(expected_vld2_1,poly,8,16) [] = { 0x0, 0x1, 0x2, 0x3, 0xc, 0xd, 0xe, 0xf }; VECT_VAR_DECL(expected_vld2_1,poly,16,8) [] = { 0xfff8, 0xfff9, 0xfffa, 0xfffb, 0xfffc, 0xfffd, 0xfffe, 0xffff }; +VECT_VAR_DECL(expected_vld2_1,hfloat,16,8) [] = { 0xc800, 0xc700, 0xc600, 0xc500, + 0xc400, 0xc200, 0xc000, 0xbc00 }; VECT_VAR_DECL(expected_vld2_1,hfloat,32,4) [] = { 0xc1400000, 0xc1300000, 0xc1200000, 0xc1100000 }; @@ -98,6 +104,7 @@ VECT_VAR_DECL(expected_vld3_0,uint,64,1) [] = { 0xfffffffffffffff0 }; VECT_VAR_DECL(expected_vld3_0,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7 }; VECT_VAR_DECL(expected_vld3_0,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 }; +VECT_VAR_DECL(expected_vld3_0,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 }; VECT_VAR_DECL(expected_vld3_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 }; VECT_VAR_DECL(expected_vld3_0,int,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, @@ -121,6 +128,8 @@ VECT_VAR_DECL(expected_vld3_0,poly,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xfc, 0xfd, 0xfe, 0xff }; VECT_VAR_DECL(expected_vld3_0,poly,16,8) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3, 0xfff4, 0xfff5, 0xfff6, 0xfff7 }; +VECT_VAR_DECL(expected_vld3_0,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80, + 0xca00, 0xc980, 0xc900, 0xc880 }; VECT_VAR_DECL(expected_vld3_0,hfloat,32,4) [] = { 0xc1800000, 0xc1700000, 0xc1600000, 0xc1500000 }; @@ -138,6 +147,7 @@ VECT_VAR_DECL(expected_vld3_1,uint,64,1) [] = { 0xfffffffffffffff1 }; VECT_VAR_DECL(expected_vld3_1,poly,8,8) [] = { 0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xff }; VECT_VAR_DECL(expected_vld3_1,poly,16,4) [] = { 0xfff4, 0xfff5, 0xfff6, 0xfff7 }; +VECT_VAR_DECL(expected_vld3_1,hfloat,16,4) [] = { 0xca00, 0xc980, 0xc900, 0xc880 }; VECT_VAR_DECL(expected_vld3_1,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 }; VECT_VAR_DECL(expected_vld3_1,int,8,16) [] = { 0x0, 0x1, 0x2, 0x3, 0x4, 0x5, 0x6, 0x7, @@ -161,6 +171,8 @@ VECT_VAR_DECL(expected_vld3_1,poly,8,16) [] = { 0x0, 0x1, 0x2, 0x3, 0xc, 0xd, 0xe, 0xf }; VECT_VAR_DECL(expected_vld3_1,poly,16,8) [] = { 0xfff8, 0xfff9, 0xfffa, 0xfffb, 0xfffc, 0xfffd, 0xfffe, 0xffff }; +VECT_VAR_DECL(expected_vld3_1,hfloat,16,8) [] = { 0xc800, 0xc700, 0xc600, 0xc500, + 0xc400, 0xc200, 0xc000, 0xbc00 }; VECT_VAR_DECL(expected_vld3_1,hfloat,32,4) [] = { 0xc1400000, 0xc1300000, 0xc1200000, 0xc1100000 }; @@ -181,6 +193,7 @@ VECT_VAR_DECL(expected_vld3_2,poly,8,8) [] = { 0x0, 0x1, 0x2, 0x3, 0x4, 0x5, 0x6, 0x7 }; VECT_VAR_DECL(expected_vld3_2,poly,16,4) [] = { 0xfff8, 0xfff9, 0xfffa, 0xfffb }; +VECT_VAR_DECL(expected_vld3_2,hfloat,16,4) [] = { 0xc800, 0xc700, 0xc600, 0xc500 }; VECT_VAR_DECL(expected_vld3_2,hfloat,32,2) [] = { 0xc1400000, 0xc1300000 }; VECT_VAR_DECL(expected_vld3_2,int,8,16) [] = { 0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, @@ -204,6 +217,8 @@ VECT_VAR_DECL(expected_vld3_2,poly,8,16) [] = { 0x10, 0x11, 0x12, 0x13, 0x1c, 0x1d, 0x1e, 0x1f }; VECT_VAR_DECL(expected_vld3_2,poly,16,8) [] = { 0x0, 0x1, 0x2, 0x3, 0x4, 0x5, 0x6, 0x7 }; +VECT_VAR_DECL(expected_vld3_2,hfloat,16,8) [] = { 0x0000, 0x3c00, 0x4000, 0x4200, + 0x4400, 0x4500, 0x4600, 0x4700 }; VECT_VAR_DECL(expected_vld3_2,hfloat,32,4) [] = { 0xc1000000, 0xc0e00000, 0xc0c00000, 0xc0a00000 }; @@ -223,6 +238,7 @@ VECT_VAR_DECL(expected_vld4_0,uint,64,1) [] = { 0xfffffffffffffff0 }; VECT_VAR_DECL(expected_vld4_0,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7 }; VECT_VAR_DECL(expected_vld4_0,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 }; +VECT_VAR_DECL(expected_vld4_0,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 }; VECT_VAR_DECL(expected_vld4_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 }; VECT_VAR_DECL(expected_vld4_0,int,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, @@ -246,6 +262,8 @@ VECT_VAR_DECL(expected_vld4_0,poly,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xfc, 0xfd, 0xfe, 0xff }; VECT_VAR_DECL(expected_vld4_0,poly,16,8) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3, 0xfff4, 0xfff5, 0xfff6, 0xfff7 }; +VECT_VAR_DECL(expected_vld4_0,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80, + 0xca00, 0xc980, 0xc900, 0xc880 }; VECT_VAR_DECL(expected_vld4_0,hfloat,32,4) [] = { 0xc1800000, 0xc1700000, 0xc1600000, 0xc1500000 }; @@ -263,6 +281,7 @@ VECT_VAR_DECL(expected_vld4_1,uint,64,1) [] = { 0xfffffffffffffff1 }; VECT_VAR_DECL(expected_vld4_1,poly,8,8) [] = { 0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xff }; VECT_VAR_DECL(expected_vld4_1,poly,16,4) [] = { 0xfff4, 0xfff5, 0xfff6, 0xfff7 }; +VECT_VAR_DECL(expected_vld4_1,hfloat,16,4) [] = { 0xca00, 0xc980, 0xc900, 0xc880 }; VECT_VAR_DECL(expected_vld4_1,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 }; VECT_VAR_DECL(expected_vld4_1,int,8,16) [] = { 0x0, 0x1, 0x2, 0x3, 0x4, 0x5, 0x6, 0x7, @@ -286,6 +305,8 @@ VECT_VAR_DECL(expected_vld4_1,poly,8,16) [] = { 0x0, 0x1, 0x2, 0x3, 0xc, 0xd, 0xe, 0xf }; VECT_VAR_DECL(expected_vld4_1,poly,16,8) [] = { 0xfff8, 0xfff9, 0xfffa, 0xfffb, 0xfffc, 0xfffd, 0xfffe, 0xffff }; +VECT_VAR_DECL(expected_vld4_1,hfloat,16,8) [] = { 0xc800, 0xc700, 0xc600, 0xc500, + 0xc400, 0xc200, 0xc000, 0xbc00 }; VECT_VAR_DECL(expected_vld4_1,hfloat,32,4) [] = { 0xc1400000, 0xc1300000, 0xc1200000, 0xc1100000 }; @@ -303,6 +324,7 @@ VECT_VAR_DECL(expected_vld4_2,uint,64,1) [] = { 0xfffffffffffffff2 }; VECT_VAR_DECL(expected_vld4_2,poly,8,8) [] = { 0x0, 0x1, 0x2, 0x3, 0x4, 0x5, 0x6, 0x7 }; VECT_VAR_DECL(expected_vld4_2,poly,16,4) [] = { 0xfff8, 0xfff9, 0xfffa, 0xfffb }; +VECT_VAR_DECL(expected_vld4_2,hfloat,16,4) [] = { 0xc800, 0xc700, 0xc600, 0xc500 }; VECT_VAR_DECL(expected_vld4_2,hfloat,32,2) [] = { 0xc1400000, 0xc1300000 }; VECT_VAR_DECL(expected_vld4_2,int,8,16) [] = { 0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, @@ -326,6 +348,8 @@ VECT_VAR_DECL(expected_vld4_2,poly,8,16) [] = { 0x10, 0x11, 0x12, 0x13, 0x1c, 0x1d, 0x1e, 0x1f }; VECT_VAR_DECL(expected_vld4_2,poly,16,8) [] = { 0x0, 0x1, 0x2, 0x3, 0x4, 0x5, 0x6, 0x7 }; +VECT_VAR_DECL(expected_vld4_2,hfloat,16,8) [] = { 0x0000, 0x3c00, 0x4000, 0x4200, + 0x4400, 0x4500, 0x4600, 0x4700 }; VECT_VAR_DECL(expected_vld4_2,hfloat,32,4) [] = { 0xc1000000, 0xc0e00000, 0xc0c00000, 0xc0a00000 }; @@ -343,6 +367,7 @@ VECT_VAR_DECL(expected_vld4_3,uint,64,1) [] = { 0xfffffffffffffff3 }; VECT_VAR_DECL(expected_vld4_3,poly,8,8) [] = { 0x8, 0x9, 0xa, 0xb, 0xc, 0xd, 0xe, 0xf }; VECT_VAR_DECL(expected_vld4_3,poly,16,4) [] = { 0xfffc, 0xfffd, 0xfffe, 0xffff }; +VECT_VAR_DECL(expected_vld4_3,hfloat,16,4) [] = { 0xc400, 0xc200, 0xc000, 0xbc00 }; VECT_VAR_DECL(expected_vld4_3,hfloat,32,2) [] = { 0xc1200000, 0xc1100000 }; VECT_VAR_DECL(expected_vld4_3,int,8,16) [] = { 0x20, 0x21, 0x22, 0x23, 0x24, 0x25, 0x26, 0x27, @@ -366,6 +391,8 @@ VECT_VAR_DECL(expected_vld4_3,poly,8,16) [] = { 0x20, 0x21, 0x22, 0x23, 0x2c, 0x2d, 0x2e, 0x2f }; VECT_VAR_DECL(expected_vld4_3,poly,16,8) [] = { 0x8, 0x9, 0xa, 0xb, 0xc, 0xd, 0xe, 0xf }; +VECT_VAR_DECL(expected_vld4_3,hfloat,16,8) [] = { 0x4800, 0x4880, 0x4900, 0x4980, + 0x4a00, 0x4a80, 0x4b00, 0x4b80 }; VECT_VAR_DECL(expected_vld4_3,hfloat,32,4) [] = { 0xc0800000, 0xc0400000, 0xc0000000, 0xbf800000 }; @@ -398,7 +425,7 @@ void exec_vldX (void) sizeof(VECT_VAR(result, T1, W, N))); /* We need all variants in 64 bits, but there is no 64x2 variant. */ -#define DECL_ALL_VLDX(X) \ +#define DECL_ALL_VLDX_NO_FP16(X) \ DECL_VLDX(int, 8, 8, X); \ DECL_VLDX(int, 16, 4, X); \ DECL_VLDX(int, 32, 2, X); \ @@ -420,7 +447,16 @@ void exec_vldX (void) DECL_VLDX(poly, 16, 8, X); \ DECL_VLDX(float, 32, 4, X) -#define TEST_ALL_VLDX(X) \ +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define DECL_ALL_VLDX(X) \ + DECL_ALL_VLDX_NO_FP16(X); \ + DECL_VLDX(float, 16, 4, X); \ + DECL_VLDX(float, 16, 8, X) +#else +#define DECL_ALL_VLDX(X) DECL_ALL_VLDX_NO_FP16(X) +#endif + +#define TEST_ALL_VLDX_NO_FP16(X) \ TEST_VLDX(, int, s, 8, 8, X); \ TEST_VLDX(, int, s, 16, 4, X); \ TEST_VLDX(, int, s, 32, 2, X); \ @@ -442,7 +478,16 @@ void exec_vldX (void) TEST_VLDX(q, poly, p, 16, 8, X); \ TEST_VLDX(q, float, f, 32, 4, X) -#define TEST_ALL_EXTRA_CHUNKS(X, Y) \ +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define TEST_ALL_VLDX(X) \ + TEST_ALL_VLDX_NO_FP16(X); \ + TEST_VLDX(, float, f, 16, 4, X); \ + TEST_VLDX(q, float, f, 16, 8, X) +#else +#define TEST_ALL_VLDX(X) TEST_ALL_VLDX_NO_FP16(X) +#endif + +#define TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y) \ TEST_EXTRA_CHUNK(int, 8, 8, X, Y); \ TEST_EXTRA_CHUNK(int, 16, 4, X, Y); \ TEST_EXTRA_CHUNK(int, 32, 2, X, Y); \ @@ -464,9 +509,17 @@ void exec_vldX (void) TEST_EXTRA_CHUNK(poly, 16, 8, X, Y); \ TEST_EXTRA_CHUNK(float, 32, 4, X, Y) +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define TEST_ALL_EXTRA_CHUNKS(X, Y) \ + TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y) \ + TEST_EXTRA_CHUNK(float, 16, 4, X, Y); \ + TEST_EXTRA_CHUNK(float, 16, 8, X, Y); +#else +#define TEST_ALL_EXTRA_CHUNKS(X, Y) TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y) +#endif + /* vldX supports all vector types except [u]int64x2. */ -#define CHECK_RESULTS_VLDX(test_name,EXPECTED,comment) \ - { \ +#define CHECK_RESULTS_VLDX_NO_FP16(test_name,EXPECTED,comment) \ CHECK(test_name, int, 8, 8, PRIx8, EXPECTED, comment); \ CHECK(test_name, int, 16, 4, PRIx16, EXPECTED, comment); \ CHECK(test_name, int, 32, 2, PRIx32, EXPECTED, comment); \ @@ -487,8 +540,19 @@ void exec_vldX (void) CHECK(test_name, uint, 32, 4, PRIx32, EXPECTED, comment); \ CHECK(test_name, poly, 8, 16, PRIx8, EXPECTED, comment); \ CHECK(test_name, poly, 16, 8, PRIx16, EXPECTED, comment); \ - CHECK_FP(test_name, float, 32, 4, PRIx32, EXPECTED, comment); \ - } \ + CHECK_FP(test_name, float, 32, 4, PRIx32, EXPECTED, comment) + +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define CHECK_RESULTS_VLDX(test_name,EXPECTED,comment) \ + { \ + CHECK_RESULTS_VLDX_NO_FP16(test_name, EXPECTED, comment); \ + CHECK_FP(test_name, float, 16, 4, PRIx16, EXPECTED, comment); \ + CHECK_FP(test_name, float, 16, 8, PRIx16, EXPECTED, comment); \ + } +#else +#define CHECK_RESULTS_VLDX(test_name, EXPECTED, comment) \ + { CHECK_RESULTS_VLDX_NO_FP16(test_name, EXPECTED, comment); } +#endif DECL_ALL_VLDX(2); DECL_ALL_VLDX(3); @@ -516,6 +580,10 @@ void exec_vldX (void) PAD(buffer_vld2_pad, poly, 8, 8); VECT_ARRAY_INIT2(buffer_vld2, poly, 16, 4); PAD(buffer_vld2_pad, poly, 16, 4); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VECT_ARRAY_INIT2(buffer_vld2, float, 16, 4); + PAD(buffer_vld2_pad, float, 16, 4); +#endif VECT_ARRAY_INIT2(buffer_vld2, float, 32, 2); PAD(buffer_vld2_pad, float, 32, 2); @@ -539,6 +607,10 @@ void exec_vldX (void) PAD(buffer_vld2_pad, poly, 8, 16); VECT_ARRAY_INIT2(buffer_vld2, poly, 16, 8); PAD(buffer_vld2_pad, poly, 16, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VECT_ARRAY_INIT2(buffer_vld2, float, 16, 8); + PAD(buffer_vld2_pad, float, 16, 8); +#endif VECT_ARRAY_INIT2(buffer_vld2, float, 32, 4); PAD(buffer_vld2_pad, float, 32, 4); @@ -563,6 +635,10 @@ void exec_vldX (void) PAD(buffer_vld3_pad, poly, 8, 8); VECT_ARRAY_INIT3(buffer_vld3, poly, 16, 4); PAD(buffer_vld3_pad, poly, 16, 4); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VECT_ARRAY_INIT3(buffer_vld3, float, 16, 4); + PAD(buffer_vld3_pad, float, 16, 4); +#endif VECT_ARRAY_INIT3(buffer_vld3, float, 32, 2); PAD(buffer_vld3_pad, float, 32, 2); @@ -586,6 +662,10 @@ void exec_vldX (void) PAD(buffer_vld3_pad, poly, 8, 16); VECT_ARRAY_INIT3(buffer_vld3, poly, 16, 8); PAD(buffer_vld3_pad, poly, 16, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VECT_ARRAY_INIT3(buffer_vld3, float, 16, 8); + PAD(buffer_vld3_pad, float, 16, 8); +#endif VECT_ARRAY_INIT3(buffer_vld3, float, 32, 4); PAD(buffer_vld3_pad, float, 32, 4); @@ -610,6 +690,10 @@ void exec_vldX (void) PAD(buffer_vld4_pad, poly, 8, 8); VECT_ARRAY_INIT4(buffer_vld4, poly, 16, 4); PAD(buffer_vld4_pad, poly, 16, 4); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VECT_ARRAY_INIT4(buffer_vld4, float, 16, 4); + PAD(buffer_vld4_pad, float, 16, 4); +#endif VECT_ARRAY_INIT4(buffer_vld4, float, 32, 2); PAD(buffer_vld4_pad, float, 32, 2); @@ -633,6 +717,10 @@ void exec_vldX (void) PAD(buffer_vld4_pad, poly, 8, 16); VECT_ARRAY_INIT4(buffer_vld4, poly, 16, 8); PAD(buffer_vld4_pad, poly, 16, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VECT_ARRAY_INIT4(buffer_vld4, float, 16, 8); + PAD(buffer_vld4_pad, float, 16, 8); +#endif VECT_ARRAY_INIT4(buffer_vld4, float, 32, 4); PAD(buffer_vld4_pad, float, 32, 4); diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vldX_dup.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vldX_dup.c index c66dade8e45b9550612b7d736391f56622f510e3..e4cde46725fff839dc6e7b81ef60f045949a368a 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vldX_dup.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vldX_dup.c @@ -18,6 +18,7 @@ VECT_VAR_DECL(expected_vld2_0,uint,64,1) [] = { 0xfffffffffffffff0 }; VECT_VAR_DECL(expected_vld2_0,poly,8,8) [] = { 0xf0, 0xf1, 0xf0, 0xf1, 0xf0, 0xf1, 0xf0, 0xf1 }; VECT_VAR_DECL(expected_vld2_0,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff0, 0xfff1 }; +VECT_VAR_DECL(expected_vld2_0,hfloat,16,4) [] = {0xcc00, 0xcb80, 0xcc00, 0xcb80 }; VECT_VAR_DECL(expected_vld2_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 }; /* vld2_dup/chunk 1. */ @@ -35,6 +36,7 @@ VECT_VAR_DECL(expected_vld2_1,poly,8,8) [] = { 0xf0, 0xf1, 0xf0, 0xf1, 0xf0, 0xf1, 0xf0, 0xf1 }; VECT_VAR_DECL(expected_vld2_1,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff0, 0xfff1 }; +VECT_VAR_DECL(expected_vld2_1,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcc00, 0xcb80 }; VECT_VAR_DECL(expected_vld2_1,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 }; /* vld3_dup/chunk 0. */ @@ -54,6 +56,7 @@ VECT_VAR_DECL(expected_vld3_0,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf0, 0xf1, 0xf2, 0xf0, 0xf1 }; VECT_VAR_DECL(expected_vld3_0,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff0 }; +VECT_VAR_DECL(expected_vld3_0,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xcc00 }; VECT_VAR_DECL(expected_vld3_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 }; /* vld3_dup/chunk 1. */ @@ -73,6 +76,7 @@ VECT_VAR_DECL(expected_vld3_1,poly,8,8) [] = { 0xf2, 0xf0, 0xf1, 0xf2, 0xf0, 0xf1, 0xf2, 0xf0 }; VECT_VAR_DECL(expected_vld3_1,poly,16,4) [] = { 0xfff1, 0xfff2, 0xfff0, 0xfff1 }; +VECT_VAR_DECL(expected_vld3_1,hfloat,16,4) [] = { 0xcb80, 0xcb00, 0xcc00, 0xcb80 }; VECT_VAR_DECL(expected_vld3_1,hfloat,32,2) [] = { 0xc1600000, 0xc1800000 }; /* vld3_dup/chunk 2. */ @@ -92,6 +96,7 @@ VECT_VAR_DECL(expected_vld3_2,poly,8,8) [] = { 0xf1, 0xf2, 0xf0, 0xf1, 0xf2, 0xf0, 0xf1, 0xf2 }; VECT_VAR_DECL(expected_vld3_2,poly,16,4) [] = { 0xfff2, 0xfff0, 0xfff1, 0xfff2 }; +VECT_VAR_DECL(expected_vld3_2,hfloat,16,4) [] = { 0xcb00, 0xcc00, 0xcb80, 0xcb00 }; VECT_VAR_DECL(expected_vld3_2,hfloat,32,2) [] = { 0xc1700000, 0xc1600000 }; /* vld4_dup/chunk 0. */ @@ -109,6 +114,7 @@ VECT_VAR_DECL(expected_vld4_0,uint,64,1) [] = { 0xfffffffffffffff0 }; VECT_VAR_DECL(expected_vld4_0,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf0, 0xf1, 0xf2, 0xf3 }; VECT_VAR_DECL(expected_vld4_0,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 }; +VECT_VAR_DECL(expected_vld4_0,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 }; VECT_VAR_DECL(expected_vld4_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 }; /* vld4_dup/chunk 1. */ @@ -125,6 +131,7 @@ VECT_VAR_DECL(expected_vld4_1,uint,64,1) [] = { 0xfffffffffffffff1 }; VECT_VAR_DECL(expected_vld4_1,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf0, 0xf1, 0xf2, 0xf3 }; VECT_VAR_DECL(expected_vld4_1,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 }; +VECT_VAR_DECL(expected_vld4_1,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 }; VECT_VAR_DECL(expected_vld4_1,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 }; /* vld4_dup/chunk 2. */ @@ -141,6 +148,7 @@ VECT_VAR_DECL(expected_vld4_2,uint,64,1) [] = { 0xfffffffffffffff2 }; VECT_VAR_DECL(expected_vld4_2,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf0, 0xf1, 0xf2, 0xf3 }; VECT_VAR_DECL(expected_vld4_2,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 }; +VECT_VAR_DECL(expected_vld4_2,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 }; VECT_VAR_DECL(expected_vld4_2,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 }; /* vld4_dup/chunk3. */ @@ -157,6 +165,7 @@ VECT_VAR_DECL(expected_vld4_3,uint,64,1) [] = { 0xfffffffffffffff3 }; VECT_VAR_DECL(expected_vld4_3,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf0, 0xf1, 0xf2, 0xf3 }; VECT_VAR_DECL(expected_vld4_3,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 }; +VECT_VAR_DECL(expected_vld4_3,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 }; VECT_VAR_DECL(expected_vld4_3,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 }; void exec_vldX_dup (void) @@ -188,7 +197,7 @@ void exec_vldX_dup (void) &(VECT_VAR(result_bis_##X, T1, W, N)[Y*N]), \ sizeof(VECT_VAR(result, T1, W, N))); -#define DECL_ALL_VLDX_DUP(X) \ +#define DECL_ALL_VLDX_DUP_NO_FP16(X) \ DECL_VLDX_DUP(int, 8, 8, X); \ DECL_VLDX_DUP(int, 16, 4, X); \ DECL_VLDX_DUP(int, 32, 2, X); \ @@ -201,7 +210,15 @@ void exec_vldX_dup (void) DECL_VLDX_DUP(poly, 16, 4, X); \ DECL_VLDX_DUP(float, 32, 2, X) -#define TEST_ALL_VLDX_DUP(X) \ +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define DECL_ALL_VLDX_DUP(X) \ + DECL_ALL_VLDX_DUP_NO_FP16(X); \ + DECL_VLDX_DUP(float, 16, 4, X) +#else +#define DECL_ALL_VLDX_DUP(X) DECL_ALL_VLDX_DUP_NO_FP16(X) +#endif + +#define TEST_ALL_VLDX_DUP_NO_FP16(X) \ TEST_VLDX_DUP(, int, s, 8, 8, X); \ TEST_VLDX_DUP(, int, s, 16, 4, X); \ TEST_VLDX_DUP(, int, s, 32, 2, X); \ @@ -214,7 +231,15 @@ void exec_vldX_dup (void) TEST_VLDX_DUP(, poly, p, 16, 4, X); \ TEST_VLDX_DUP(, float, f, 32, 2, X) -#define TEST_ALL_EXTRA_CHUNKS(X, Y) \ +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define TEST_ALL_VLDX_DUP(X) \ + TEST_ALL_VLDX_DUP_NO_FP16(X); \ + TEST_VLDX_DUP(, float, f, 16, 4, X) +#else +#define TEST_ALL_VLDX_DUP(X) TEST_ALL_VLDX_DUP_NO_FP16(X) +#endif + +#define TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y) \ TEST_EXTRA_CHUNK(int, 8, 8, X, Y); \ TEST_EXTRA_CHUNK(int, 16, 4, X, Y); \ TEST_EXTRA_CHUNK(int, 32, 2, X, Y); \ @@ -227,9 +252,16 @@ void exec_vldX_dup (void) TEST_EXTRA_CHUNK(poly, 16, 4, X, Y); \ TEST_EXTRA_CHUNK(float, 32, 2, X, Y) +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define TEST_ALL_EXTRA_CHUNKS(X, Y) \ + TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y); \ + TEST_EXTRA_CHUNK(float, 16, 4, X, Y) +#else +#define TEST_ALL_EXTRA_CHUNKS(X, Y) TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y) +#endif + /* vldX_dup supports only 64-bit inputs. */ -#define CHECK_RESULTS_VLDX_DUP(test_name,EXPECTED,comment) \ - { \ +#define CHECK_RESULTS_VLDX_DUP_NO_FP16(test_name,EXPECTED,comment) \ CHECK(test_name, int, 8, 8, PRIx8, EXPECTED, comment); \ CHECK(test_name, int, 16, 4, PRIx16, EXPECTED, comment); \ CHECK(test_name, int, 32, 2, PRIx32, EXPECTED, comment); \ @@ -240,8 +272,20 @@ void exec_vldX_dup (void) CHECK(test_name, uint, 64, 1, PRIx64, EXPECTED, comment); \ CHECK(test_name, poly, 8, 8, PRIx8, EXPECTED, comment); \ CHECK(test_name, poly, 16, 4, PRIx16, EXPECTED, comment); \ - CHECK_FP(test_name, float, 32, 2, PRIx32, EXPECTED, comment); \ - } \ + CHECK_FP(test_name, float, 32, 2, PRIx32, EXPECTED, comment) + +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define CHECK_RESULTS_VLDX_DUP(test_name,EXPECTED,comment) \ + { \ + CHECK_RESULTS_VLDX_DUP_NO_FP16(test_name,EXPECTED,comment); \ + CHECK_FP(test_name, float, 16, 4, PRIx16, EXPECTED, comment); \ + } +#else +#define CHECK_RESULTS_VLDX_DUP(test_name,EXPECTED,comment) \ + { \ + CHECK_RESULTS_VLDX_DUP_NO_FP16(test_name,EXPECTED,comment); \ + } +#endif DECL_ALL_VLDX_DUP(2); DECL_ALL_VLDX_DUP(3); @@ -269,6 +313,10 @@ void exec_vldX_dup (void) PAD(buffer_vld2_pad, poly, 8, 8); VECT_ARRAY_INIT2(buffer_vld2, poly, 16, 4); PAD(buffer_vld2_pad, poly, 16, 4); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VECT_ARRAY_INIT2(buffer_vld2, float, 16, 4); + PAD(buffer_vld2_pad, float, 16, 4); +#endif VECT_ARRAY_INIT2(buffer_vld2, float, 32, 2); PAD(buffer_vld2_pad, float, 32, 2); @@ -292,6 +340,10 @@ void exec_vldX_dup (void) PAD(buffer_vld2_pad, poly, 8, 16); VECT_ARRAY_INIT2(buffer_vld2, poly, 16, 8); PAD(buffer_vld2_pad, poly, 16, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VECT_ARRAY_INIT2(buffer_vld2, float, 16, 8); + PAD(buffer_vld2_pad, float, 16, 8); +#endif VECT_ARRAY_INIT2(buffer_vld2, float, 32, 4); PAD(buffer_vld2_pad, float, 32, 4); @@ -316,6 +368,10 @@ void exec_vldX_dup (void) PAD(buffer_vld3_pad, poly, 8, 8); VECT_ARRAY_INIT3(buffer_vld3, poly, 16, 4); PAD(buffer_vld3_pad, poly, 16, 4); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VECT_ARRAY_INIT3(buffer_vld3, float, 16, 4); + PAD(buffer_vld3_pad, float, 16, 4); +#endif VECT_ARRAY_INIT3(buffer_vld3, float, 32, 2); PAD(buffer_vld3_pad, float, 32, 2); @@ -339,6 +395,10 @@ void exec_vldX_dup (void) PAD(buffer_vld3_pad, poly, 8, 16); VECT_ARRAY_INIT3(buffer_vld3, poly, 16, 8); PAD(buffer_vld3_pad, poly, 16, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VECT_ARRAY_INIT3(buffer_vld3, float, 16, 8); + PAD(buffer_vld3_pad, float, 16, 8); +#endif VECT_ARRAY_INIT3(buffer_vld3, float, 32, 4); PAD(buffer_vld3_pad, float, 32, 4); @@ -363,6 +423,10 @@ void exec_vldX_dup (void) PAD(buffer_vld4_pad, poly, 8, 8); VECT_ARRAY_INIT4(buffer_vld4, poly, 16, 4); PAD(buffer_vld4_pad, poly, 16, 4); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VECT_ARRAY_INIT4(buffer_vld4, float, 16, 4); + PAD(buffer_vld4_pad, float, 16, 4); +#endif VECT_ARRAY_INIT4(buffer_vld4, float, 32, 2); PAD(buffer_vld4_pad, float, 32, 2); @@ -386,6 +450,10 @@ void exec_vldX_dup (void) PAD(buffer_vld4_pad, poly, 8, 16); VECT_ARRAY_INIT4(buffer_vld4, poly, 16, 8); PAD(buffer_vld4_pad, poly, 16, 8); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VECT_ARRAY_INIT4(buffer_vld4, float, 16, 8); + PAD(buffer_vld4_pad, float, 16, 8); +#endif VECT_ARRAY_INIT4(buffer_vld4, float, 32, 4); PAD(buffer_vld4_pad, float, 32, 4); diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vldX_lane.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vldX_lane.c index 2f2e62f0e3e0f43770d66e2255907a0abc4df919..33b0eafbadb475c6cfe7702205bcd88e0e344b0c 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vldX_lane.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vldX_lane.c @@ -18,6 +18,7 @@ VECT_VAR_DECL(expected_vld2_0,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa }; VECT_VAR_DECL(expected_vld2_0,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; +VECT_VAR_DECL(expected_vld2_0,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld2_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 }; VECT_VAR_DECL(expected_vld2_0,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; @@ -29,6 +30,8 @@ VECT_VAR_DECL(expected_vld2_0,uint,32,4) [] = { 0xfffffff0, 0xfffffff1, 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld2_0,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; +VECT_VAR_DECL(expected_vld2_0,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, + 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa } ; VECT_VAR_DECL(expected_vld2_0,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa }; @@ -44,6 +47,7 @@ VECT_VAR_DECL(expected_vld2_1,uint,32,2) [] = { 0xfffffff0, 0xfffffff1 }; VECT_VAR_DECL(expected_vld2_1,poly,8,8) [] = { 0xf0, 0xf1, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa }; VECT_VAR_DECL(expected_vld2_1,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xfff0, 0xfff1 }; +VECT_VAR_DECL(expected_vld2_1,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld2_1,hfloat,32,2) [] = { 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld2_1,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xfff0, 0xfff1, 0xaaaa, 0xaaaa }; @@ -55,6 +59,8 @@ VECT_VAR_DECL(expected_vld2_1,uint,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld2_1,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xfff0, 0xfff1, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; +VECT_VAR_DECL(expected_vld2_1,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, + 0xcc00, 0xcb80, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld2_1,hfloat,32,4) [] = { 0xc1800000, 0xc1700000, 0xaaaaaaaa, 0xaaaaaaaa }; @@ -70,6 +76,7 @@ VECT_VAR_DECL(expected_vld3_0,uint,32,2) [] = { 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld3_0,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa }; VECT_VAR_DECL(expected_vld3_0,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; +VECT_VAR_DECL(expected_vld3_0,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld3_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 }; VECT_VAR_DECL(expected_vld3_0,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; @@ -81,6 +88,8 @@ VECT_VAR_DECL(expected_vld3_0,uint,32,4) [] = { 0xfffffff0, 0xfffffff1, 0xfffffff2, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld3_0,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; +VECT_VAR_DECL(expected_vld3_0,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, + 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld3_0,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa }; @@ -96,6 +105,7 @@ VECT_VAR_DECL(expected_vld3_1,uint,32,2) [] = { 0xaaaaaaaa, 0xfffffff0 }; VECT_VAR_DECL(expected_vld3_1,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa, 0xf0, 0xf1, 0xf2, 0xaa }; VECT_VAR_DECL(expected_vld3_1,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; +VECT_VAR_DECL(expected_vld3_1,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xcc00, 0xcb80 }; VECT_VAR_DECL(expected_vld3_1,hfloat,32,2) [] = { 0xc1600000, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld3_1,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; @@ -107,6 +117,8 @@ VECT_VAR_DECL(expected_vld3_1,uint,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld3_1,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xfff0 }; +VECT_VAR_DECL(expected_vld3_1,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, + 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld3_1,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa, 0xc1800000, 0xc1700000 }; @@ -122,6 +134,7 @@ VECT_VAR_DECL(expected_vld3_2,uint,32,2) [] = { 0xfffffff1, 0xfffffff2 }; VECT_VAR_DECL(expected_vld3_2,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa }; VECT_VAR_DECL(expected_vld3_2,poly,16,4) [] = { 0xaaaa, 0xfff0, 0xfff1, 0xfff2 }; +VECT_VAR_DECL(expected_vld3_2,hfloat,16,4) [] = { 0xcb00, 0xaaaa, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld3_2,hfloat,32,2) [] = { 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld3_2,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xfff0, 0xfff1, 0xfff2, 0xaaaa, 0xaaaa, 0xaaaa }; @@ -133,6 +146,8 @@ VECT_VAR_DECL(expected_vld3_2,uint,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld3_2,poly,16,8) [] = { 0xfff1, 0xfff2, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; +VECT_VAR_DECL(expected_vld3_2,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xcc00, 0xcb80, + 0xcb00, 0xaaaa, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld3_2,hfloat,32,4) [] = { 0xc1600000, 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa }; @@ -148,6 +163,7 @@ VECT_VAR_DECL(expected_vld4_0,uint,32,2) [] = { 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld4_0,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa }; VECT_VAR_DECL(expected_vld4_0,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; +VECT_VAR_DECL(expected_vld4_0,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld4_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 }; VECT_VAR_DECL(expected_vld4_0,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; @@ -159,6 +175,8 @@ VECT_VAR_DECL(expected_vld4_0,uint,32,4) [] = { 0xfffffff0, 0xfffffff1, 0xfffffff2, 0xfffffff3 }; VECT_VAR_DECL(expected_vld4_0,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; +VECT_VAR_DECL(expected_vld4_0,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, + 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld4_0,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa }; @@ -174,6 +192,7 @@ VECT_VAR_DECL(expected_vld4_1,uint,32,2) [] = { 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld4_1,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa }; VECT_VAR_DECL(expected_vld4_1,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; +VECT_VAR_DECL(expected_vld4_1,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld4_1,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 }; VECT_VAR_DECL(expected_vld4_1,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; @@ -185,6 +204,8 @@ VECT_VAR_DECL(expected_vld4_1,uint,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld4_1,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; +VECT_VAR_DECL(expected_vld4_1,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, + 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld4_1,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa }; @@ -200,6 +221,7 @@ VECT_VAR_DECL(expected_vld4_2,uint,32,2) [] = { 0xfffffff0, 0xfffffff1 }; VECT_VAR_DECL(expected_vld4_2,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xaa, 0xaa, 0xaa, 0xaa }; VECT_VAR_DECL(expected_vld4_2,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; +VECT_VAR_DECL(expected_vld4_2,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 }; VECT_VAR_DECL(expected_vld4_2,hfloat,32,2) [] = { 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld4_2,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; @@ -211,6 +233,8 @@ VECT_VAR_DECL(expected_vld4_2,uint,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld4_2,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xfff0, 0xfff1, 0xfff2, 0xfff3 }; +VECT_VAR_DECL(expected_vld4_2,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, + 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld4_2,hfloat,32,4) [] = { 0xc1800000, 0xc1700000, 0xc1600000, 0xc1500000 }; @@ -226,6 +250,7 @@ VECT_VAR_DECL(expected_vld4_3,uint,32,2) [] = { 0xfffffff2, 0xfffffff3 }; VECT_VAR_DECL(expected_vld4_3,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa, 0xaa }; VECT_VAR_DECL(expected_vld4_3,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 }; +VECT_VAR_DECL(expected_vld4_3,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld4_3,hfloat,32,2) [] = { 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld4_3,int,16,8) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; @@ -237,6 +262,8 @@ VECT_VAR_DECL(expected_vld4_3,uint,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa }; VECT_VAR_DECL(expected_vld4_3,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; +VECT_VAR_DECL(expected_vld4_3,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80, + 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa }; VECT_VAR_DECL(expected_vld4_3,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa }; @@ -252,6 +279,9 @@ VECT_VAR_DECL_INIT(buffer_vld2_lane, uint, 32, 2); VECT_VAR_DECL_INIT(buffer_vld2_lane, uint, 64, 2); VECT_VAR_DECL_INIT(buffer_vld2_lane, poly, 8, 2); VECT_VAR_DECL_INIT(buffer_vld2_lane, poly, 16, 2); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +VECT_VAR_DECL_INIT(buffer_vld2_lane, float, 16, 2); +#endif VECT_VAR_DECL_INIT(buffer_vld2_lane, float, 32, 2); /* Input buffers for vld3_lane */ @@ -265,6 +295,9 @@ VECT_VAR_DECL_INIT(buffer_vld3_lane, uint, 32, 3); VECT_VAR_DECL_INIT(buffer_vld3_lane, uint, 64, 3); VECT_VAR_DECL_INIT(buffer_vld3_lane, poly, 8, 3); VECT_VAR_DECL_INIT(buffer_vld3_lane, poly, 16, 3); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +VECT_VAR_DECL_INIT(buffer_vld3_lane, float, 16, 3); +#endif VECT_VAR_DECL_INIT(buffer_vld3_lane, float, 32, 3); /* Input buffers for vld4_lane */ @@ -278,6 +311,9 @@ VECT_VAR_DECL_INIT(buffer_vld4_lane, uint, 32, 4); VECT_VAR_DECL_INIT(buffer_vld4_lane, uint, 64, 4); VECT_VAR_DECL_INIT(buffer_vld4_lane, poly, 8, 4); VECT_VAR_DECL_INIT(buffer_vld4_lane, poly, 16, 4); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +VECT_VAR_DECL_INIT(buffer_vld4_lane, float, 16, 4); +#endif VECT_VAR_DECL_INIT(buffer_vld4_lane, float, 32, 4); void exec_vldX_lane (void) @@ -321,7 +357,7 @@ void exec_vldX_lane (void) sizeof(VECT_VAR(result, T1, W, N))); /* We need all variants in 64 bits, but there is no 64x2 variant. */ -#define DECL_ALL_VLDX_LANE(X) \ +#define DECL_ALL_VLDX_LANE_NO_FP16(X) \ DECL_VLDX_LANE(int, 8, 8, X); \ DECL_VLDX_LANE(int, 16, 4, X); \ DECL_VLDX_LANE(int, 32, 2, X); \ @@ -338,6 +374,15 @@ void exec_vldX_lane (void) DECL_VLDX_LANE(float, 32, 2, X); \ DECL_VLDX_LANE(float, 32, 4, X) +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define DECL_ALL_VLDX_LANE(X) \ + DECL_ALL_VLDX_LANE_NO_FP16(X); \ + DECL_VLDX_LANE(float, 16, 4, X); \ + DECL_VLDX_LANE(float, 16, 8, X) +#else +#define DECL_ALL_VLDX_LANE(X) DECL_ALL_VLDX_LANE_NO_FP16(X) +#endif + /* Add some padding to try to catch out of bound accesses. */ #define ARRAY1(V, T, W, N) VECT_VAR_DECL(V,T,W,N)[1]={42} #define DUMMY_ARRAY(V, T, W, N, L) \ @@ -346,7 +391,7 @@ void exec_vldX_lane (void) /* Use the same lanes regardless of the size of the array (X), for simplicity. */ -#define TEST_ALL_VLDX_LANE(X) \ +#define TEST_ALL_VLDX_LANE_NO_FP16(X) \ TEST_VLDX_LANE(, int, s, 8, 8, X, 7); \ TEST_VLDX_LANE(, int, s, 16, 4, X, 2); \ TEST_VLDX_LANE(, int, s, 32, 2, X, 0); \ @@ -363,7 +408,16 @@ void exec_vldX_lane (void) TEST_VLDX_LANE(, float, f, 32, 2, X, 0); \ TEST_VLDX_LANE(q, float, f, 32, 4, X, 2) -#define TEST_ALL_EXTRA_CHUNKS(X, Y) \ +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define TEST_ALL_VLDX_LANE(X) \ + TEST_ALL_VLDX_LANE_NO_FP16(X); \ + TEST_VLDX_LANE(, float, f, 16, 4, X, 2); \ + TEST_VLDX_LANE(q, float, f, 16, 8, X, 6) +#else +#define TEST_ALL_VLDX_LANE(X) TEST_ALL_VLDX_LANE_NO_FP16(X) +#endif + +#define TEST_ALL_EXTRA_CHUNKS_NO_FP16(X,Y) \ TEST_EXTRA_CHUNK(int, 8, 8, X, Y); \ TEST_EXTRA_CHUNK(int, 16, 4, X, Y); \ TEST_EXTRA_CHUNK(int, 32, 2, X, Y); \ @@ -380,9 +434,17 @@ void exec_vldX_lane (void) TEST_EXTRA_CHUNK(float, 32, 2, X, Y); \ TEST_EXTRA_CHUNK(float, 32, 4, X, Y) +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define TEST_ALL_EXTRA_CHUNKS(X,Y) \ + TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y); \ + TEST_EXTRA_CHUNK(float, 16, 4, X, Y); \ + TEST_EXTRA_CHUNK(float, 16, 8, X, Y) +#else +#define TEST_ALL_EXTRA_CHUNKS(X,Y) TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y) +#endif + /* vldX_lane supports only a subset of all variants. */ -#define CHECK_RESULTS_VLDX_LANE(test_name,EXPECTED,comment) \ - { \ +#define CHECK_RESULTS_VLDX_LANE_NO_FP16(test_name,EXPECTED,comment) \ CHECK(test_name, int, 8, 8, PRIx8, EXPECTED, comment); \ CHECK(test_name, int, 16, 4, PRIx16, EXPECTED, comment); \ CHECK(test_name, int, 32, 2, PRIx32, EXPECTED, comment); \ @@ -397,8 +459,21 @@ void exec_vldX_lane (void) CHECK(test_name, uint, 16, 8, PRIx16, EXPECTED, comment); \ CHECK(test_name, uint, 32, 4, PRIx32, EXPECTED, comment); \ CHECK(test_name, poly, 16, 8, PRIx16, EXPECTED, comment); \ - CHECK_FP(test_name, float, 32, 4, PRIx32, EXPECTED, comment); \ - } \ + CHECK_FP(test_name, float, 32, 4, PRIx32, EXPECTED, comment) + +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) +#define CHECK_RESULTS_VLDX_LANE(test_name,EXPECTED,comment) \ + { \ + CHECK_RESULTS_VLDX_LANE_NO_FP16(test_name,EXPECTED,comment); \ + CHECK_FP(test_name, float, 16, 4, PRIx16, EXPECTED, comment); \ + CHECK_FP(test_name, float, 16, 8, PRIx16, EXPECTED, comment); \ + } +#else +#define CHECK_RESULTS_VLDX_LANE(test_name,EXPECTED,comment) \ + { \ + CHECK_RESULTS_VLDX_LANE_NO_FP16(test_name,EXPECTED,comment); \ + } +#endif /* Declare the temporary buffers / variables. */ DECL_ALL_VLDX_LANE(2); @@ -419,6 +494,10 @@ void exec_vldX_lane (void) DUMMY_ARRAY(buffer_src, uint, 16, 8, 4); DUMMY_ARRAY(buffer_src, uint, 32, 4, 4); DUMMY_ARRAY(buffer_src, poly, 16, 8, 4); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + DUMMY_ARRAY(buffer_src, float, 16, 4, 4); + DUMMY_ARRAY(buffer_src, float, 16, 8, 4); +#endif DUMMY_ARRAY(buffer_src, float, 32, 2, 4); DUMMY_ARRAY(buffer_src, float, 32, 4, 4); diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vset_lane.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vset_lane.c index 51594068364676f04661f4ed442934bb0c0c9652..e0499df5170594b0c97afc55b2dad3829aca57ad 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vset_lane.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vset_lane.c @@ -16,6 +16,7 @@ VECT_VAR_DECL(expected,uint,64,1) [] = { 0x88 }; VECT_VAR_DECL(expected,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0x55, 0xf7 }; VECT_VAR_DECL(expected,poly,16,4) [] = { 0xfff0, 0xfff1, 0x66, 0xfff3 }; +VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0x4840, 0xca80 }; VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0xc1800000, 0x4204cccd }; VECT_VAR_DECL(expected,int,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, @@ -41,6 +42,8 @@ VECT_VAR_DECL(expected,poly,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3, 0xfc, 0xfd, 0xdd, 0xff }; VECT_VAR_DECL(expected,poly,16,8) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3, 0xfff4, 0xfff5, 0xee, 0xfff7 }; +VECT_VAR_DECL(expected,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80, + 0xca00, 0x4480, 0xc900, 0xc880 }; VECT_VAR_DECL(expected,hfloat,32,4) [] = { 0xc1800000, 0xc1700000, 0xc1600000, 0x41333333 }; @@ -61,6 +64,10 @@ void exec_vset_lane (void) /* Initialize input "vector" from "buffer". */ TEST_MACRO_ALL_VARIANTS_2_5(VLOAD, vector, buffer); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + VLOAD(vector, buffer, , float, f, 16, 4); + VLOAD(vector, buffer, q, float, f, 16, 8); +#endif VLOAD(vector, buffer, , float, f, 32, 2); VLOAD(vector, buffer, q, float, f, 32, 4); @@ -75,6 +82,9 @@ void exec_vset_lane (void) TEST_VSET_LANE(, uint, u, 64, 1, 0x88, 0); TEST_VSET_LANE(, poly, p, 8, 8, 0x55, 6); TEST_VSET_LANE(, poly, p, 16, 4, 0x66, 2); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + TEST_VSET_LANE(, float, f, 16, 4, 8.5f, 2); +#endif TEST_VSET_LANE(, float, f, 32, 2, 33.2f, 1); TEST_VSET_LANE(q, int, s, 8, 16, 0x99, 15); @@ -87,6 +97,9 @@ void exec_vset_lane (void) TEST_VSET_LANE(q, uint, u, 64, 2, 0x11, 1); TEST_VSET_LANE(q, poly, p, 8, 16, 0xDD, 14); TEST_VSET_LANE(q, poly, p, 16, 8, 0xEE, 6); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + TEST_VSET_LANE(q, float, f, 16, 8, 4.5f, 5); +#endif TEST_VSET_LANE(q, float, f, 32, 4, 11.2f, 3); CHECK_RESULTS(TEST_MSG, ""); diff --git a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vst1_lane.c b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vst1_lane.c index 08583b88cf3705d059e0824b5199ec77739b1b97..825d07dbf77fc54b5ef796b57a42c81d6dd6d611 100644 --- a/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vst1_lane.c +++ b/gcc/testsuite/gcc.target/aarch64/advsimd-intrinsics/vst1_lane.c @@ -16,6 +16,7 @@ VECT_VAR_DECL(expected,uint,64,1) [] = { 0xfffffffffffffff0 }; VECT_VAR_DECL(expected,poly,8,8) [] = { 0xf6, 0x33, 0x33, 0x33, 0x33, 0x33, 0x33, 0x33 }; VECT_VAR_DECL(expected,poly,16,4) [] = { 0xfff2, 0x3333, 0x3333, 0x3333 }; +VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xcb80, 0x3333, 0x3333, 0x3333 }; VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0xc1700000, 0x33333333 }; VECT_VAR_DECL(expected,int,8,16) [] = { 0xff, 0x33, 0x33, 0x33, 0x33, 0x33, 0x33, 0x33, @@ -42,6 +43,8 @@ VECT_VAR_DECL(expected,poly,8,16) [] = { 0xfa, 0x33, 0x33, 0x33, 0x33, 0x33, 0x33, 0x33 }; VECT_VAR_DECL(expected,poly,16,8) [] = { 0xfff4, 0x3333, 0x3333, 0x3333, 0x3333, 0x3333, 0x3333, 0x3333 }; +VECT_VAR_DECL(expected,hfloat,16,8) [] = { 0xc900, 0x3333, 0x3333, 0x3333, + 0x3333, 0x3333, 0x3333, 0x3333 }; VECT_VAR_DECL(expected,hfloat,32,4) [] = { 0xc1700000, 0x33333333, 0x33333333, 0x33333333 }; @@ -69,6 +72,9 @@ void exec_vst1_lane (void) TEST_VST1_LANE(, uint, u, 64, 1, 0); TEST_VST1_LANE(, poly, p, 8, 8, 6); TEST_VST1_LANE(, poly, p, 16, 4, 2); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + TEST_VST1_LANE(, float, f, 16, 4, 1); +#endif TEST_VST1_LANE(, float, f, 32, 2, 1); TEST_VST1_LANE(q, int, s, 8, 16, 15); @@ -81,6 +87,9 @@ void exec_vst1_lane (void) TEST_VST1_LANE(q, uint, u, 64, 2, 0); TEST_VST1_LANE(q, poly, p, 8, 16, 10); TEST_VST1_LANE(q, poly, p, 16, 8, 4); +#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE) + TEST_VST1_LANE(q, float, f, 16, 8, 6); +#endif TEST_VST1_LANE(q, float, f, 32, 4, 1); CHECK_RESULTS(TEST_MSG, "");