From patchwork Wed Aug 10 15:44:55 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Kito Cheng X-Patchwork-Id: 465 Return-Path: Delivered-To: ouuuleilei@gmail.com Received: by 2002:a05:6a10:20da:b0:2d3:3019:e567 with SMTP id n26csp3255437pxc; Wed, 10 Aug 2022 08:46:02 -0700 (PDT) X-Google-Smtp-Source: AA6agR7x62/tKlQNQv3kGmaAhiFvIqxe1iC9nrz+EVxZzyRUkGtis76bzng2TMsLAVC2KtO8oSJ0 X-Received: by 2002:a05:6402:27d2:b0:43e:3ff6:ad58 with SMTP id c18-20020a05640227d200b0043e3ff6ad58mr26607691ede.234.1660146362166; Wed, 10 Aug 2022 08:46:02 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1660146362; cv=none; d=google.com; s=arc-20160816; b=Z/g6jS63bSFHluVW2k9kuLe6JTVA+IS+aT+6NN8LUIetE1b8zXmrv53UeuUXVLW5Qp SLS6/KcV5m4HJIQpRoypJm0keTPrRZsNYMty+/+SQY7SW1nCndjrmjCQGCkDLgakrjqJ M4mfiK5KSAHLSr2KdlJBIUxi+8AlhLl/y2WgOwNkf+r2EkMhKpzGb3cjLW8BWZ4RVNZo rch0BZLCxdw7xwPtw00rGknKQGhuK7a4D4PCKY1nE3MDk+6UOGn8StTGVfxs3t2V1L26 l/H02SS5rgpcspq8X7cGKkxp9iRFD0yNxkxAZ3mmY++c5F4RU6Na4aTSBXS/XAcIe6LN TUEg== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=sender:errors-to:cc:list-subscribe:list-help:list-post:list-archive :list-unsubscribe:list-id:precedence:content-transfer-encoding :mime-version:message-id:date:subject:to:from:dkim-signature :dmarc-filter:delivered-to; bh=EbsHa8RK/xjynDwZUBTrzHwfN31sWmGh3S391MT7NfI=; b=XiP2PDhh0l+bBvTU8gTj2yW1teOrKWrh+srHBSl6aJ+Sw54aaQ+BZ2Vpa/RzVhWca+ NO4C1j6n6GYyHxT289gXgPWqG9ah7oRhuA09jpGUdzZxx04btrwz2euawxFgP2KRK88Y DjySYsLi4x/Vlb7pNQy8TOVQHXb8J425SZ7OTC1klVcHFR2QPcqarjEfDub9OVL9PgKk IfsNwKavCyyEOoCkWX/F9JyPeh6Pisp4aZhK6cfpnTty4I/xOJrEtEyKdPDCIVVDS0Jj cL6gPmgRgMfwP4PBGUm+YXqU0NjRt646ziLznjnOaBs73XnO6xul6PU8iCpIto/U3dxN QHfQ== ARC-Authentication-Results: i=1; mx.google.com; dkim=pass header.i=@sifive.com header.s=google header.b=X3NsPKRr; spf=pass (google.com: domain of gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org designates 8.43.85.97 as permitted sender) smtp.mailfrom="gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org" Received: from sourceware.org (server2.sourceware.org. [8.43.85.97]) by mx.google.com with ESMTPS id sa27-20020a1709076d1b00b007310a6c0836si4918695ejc.754.2022.08.10.08.46.01 for (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 10 Aug 2022 08:46:02 -0700 (PDT) Received-SPF: pass (google.com: domain of gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org designates 8.43.85.97 as permitted sender) client-ip=8.43.85.97; Authentication-Results: mx.google.com; dkim=pass header.i=@sifive.com header.s=google header.b=X3NsPKRr; spf=pass (google.com: domain of gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org designates 8.43.85.97 as permitted sender) smtp.mailfrom="gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org" Received: from server2.sourceware.org (localhost [IPv6:::1]) by sourceware.org (Postfix) with ESMTP id BA65F3858406 for ; Wed, 10 Aug 2022 15:45:30 +0000 (GMT) X-Original-To: gcc-patches@gcc.gnu.org Delivered-To: gcc-patches@gcc.gnu.org Received: from mail-pf1-x429.google.com (mail-pf1-x429.google.com [IPv6:2607:f8b0:4864:20::429]) by sourceware.org (Postfix) with ESMTPS id CB8983858401 for ; Wed, 10 Aug 2022 15:45:03 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.1 sourceware.org CB8983858401 Authentication-Results: sourceware.org; dmarc=none (p=none dis=none) header.from=sifive.com Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=sifive.com Received: by mail-pf1-x429.google.com with SMTP id q19so14054935pfg.8 for ; Wed, 10 Aug 2022 08:45:03 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=sifive.com; s=google; h=content-transfer-encoding:mime-version:message-id:date:subject:cc :to:from:from:to:cc; bh=EbsHa8RK/xjynDwZUBTrzHwfN31sWmGh3S391MT7NfI=; b=X3NsPKRrVEppeLsVgVT+nTrQkAVWGwStSBlj5moJahrjCC5d1D9JWgYJY7gUtIbd4f OHqtU11e3kYmuKFN0ZoJModIDqrtSwt76TUdTuA2dpfBvkUwQ3r1EhdilC6WVJvQ3lhC qadels0/1ZKjjrNV94OAs7mETUwtVL7lt8Pxh4CHA6cxHb0Yz3GMgh99ecdxHzctPqAZ ysHd5BHS9HYPw8y3t415Gfi42V3LQnDJC5wTBYzqzwrkPvnleh+A9iRRk30weCo4iYAi 18PiMBsuvPm75JkfP4HUXqTBr6JhoiFgn/bYJuFdv41jriPF0pgbfrG87+PAUzbI9Kke duvA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:message-id:date:subject:cc :to:from:x-gm-message-state:from:to:cc; bh=EbsHa8RK/xjynDwZUBTrzHwfN31sWmGh3S391MT7NfI=; b=4/0YP82T6mGVwPnuGcnzgOm3sMWxmcD1Vo0sDbybmhPGcy4m9sUGXzh3oRT7zvOj5S QBoaZ1fMeBHoEmlX5BXJ1D3T0PRLgp4kGE1xGG905TuQ50cohtvTvZUkpg+01wtfWNqB NVk5aVgDV4MbWla05WGUAz/Zg+Jf9WnXgw37J8LuSkNbmlaHxiOxFEowUuGr0GRKaGNr o5GB2h2t/EmQsUtN6bRt9lpabMz+cgYHzl26FRADNuG+g2gNiDnCt6CdAqR8gAh3Bdvw PGP2CbQNeTKAnNB5hBM+sm1t7Q56w1V0fwCBpgeODgp9eSV3T9cxYe2HE/eyehL5g0f8 jc2g== X-Gm-Message-State: ACgBeo3h/mVaoVwObWhdsVXzDhV6PpZ88cwU5li9tyasNjuh/csTZmAr wsB57reUR00QI1Pm0mMr0xBBN7o3SmrKBw== X-Received: by 2002:a63:f412:0:b0:41b:cbd:bcd7 with SMTP id g18-20020a63f412000000b0041b0cbdbcd7mr24096434pgi.30.1660146302132; Wed, 10 Aug 2022 08:45:02 -0700 (PDT) Received: from hsinchu02.internal.sifive.com (59-124-168-89.hinet-ip.hinet.net. [59.124.168.89]) by smtp.gmail.com with ESMTPSA id s5-20020a625e05000000b0052e6d5ee183sm2131219pfb.129.2022.08.10.08.45.00 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 10 Aug 2022 08:45:01 -0700 (PDT) From: Kito Cheng To: gcc-patches@gcc.gnu.org, kito.cheng@gmail.com, jim.wilson.gcc@gmail.com, palmer@dabbelt.com, andrew@sifive.com, juzhe.zhong@rivai.ai, joseph@codesourcery.com Subject: [PATCH v2 1/2] RISC-V: Support _Float16 type. Date: Wed, 10 Aug 2022 23:44:55 +0800 Message-Id: <20220810154456.101086-1-kito.cheng@sifive.com> X-Mailer: git-send-email 2.37.1 MIME-Version: 1.0 X-Spam-Status: No, score=-12.4 required=5.0 tests=BAYES_00, DKIM_SIGNED, DKIM_VALID, DKIM_VALID_AU, DKIM_VALID_EF, GIT_PATCH_0, KAM_SHORT, RCVD_IN_DNSWL_NONE, SPF_HELO_NONE, SPF_PASS, TXREP, T_SCC_BODY_TEXT_LINE, URIBL_BLACK 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: , Cc: Kito Cheng Errors-To: gcc-patches-bounces+ouuuleilei=gmail.com@gcc.gnu.org Sender: "Gcc-patches" X-getmail-retrieved-from-mailbox: =?utf-8?q?INBOX?= X-GMAIL-THRID: =?utf-8?q?1740789631754390263?= X-GMAIL-MSGID: =?utf-8?q?1740789631754390263?= RISC-V decide use _Float16 as primary IEEE half precision type, and this already become part of psABI, this patch has added folloing support for _Float16: - Soft-float support for _Float16. - Make sure _Float16 available on C++ mode. - Name mangling for _Float16 on C++ mode. gcc/ChangeLog * config/riscv/riscv-builtins.cc: include stringpool.h (riscv_float16_type_node): New. (riscv_init_builtin_types): Ditto. (riscv_init_builtins): Call riscv_init_builtin_types. * config/riscv/riscv-modes.def (HF): New. * gcc/config/riscv/riscv.cc (riscv_output_move): Handle HFmode. (riscv_mangle_type): New. (riscv_scalar_mode_supported_p): Ditto. (riscv_libgcc_floating_mode_supported_p): Ditto. (riscv_excess_precision): Ditto. (riscv_floatn_mode): Ditto. (riscv_init_libfuncs): Ditto. (TARGET_MANGLE_TYPE): Ditto. (TARGET_SCALAR_MODE_SUPPORTED_P): Ditto. (TARGET_LIBGCC_FLOATING_MODE_SUPPORTED_P): Ditto. (TARGET_INIT_LIBFUNCS): Ditto. (TARGET_C_EXCESS_PRECISION): Ditto. (TARGET_FLOATN_MODE): Ditto. * gcc/config/riscv/riscv.md (mode): Add HF. (softload): Add HF. (softstore): Ditto. (fmt): Ditto. (UNITMODE): Ditto. (movhf): New. (*movhf_softfloat): New. libgcc/ChangeLog: * config/riscv/sfp-machine.h (_FP_NANFRAC_H): New. (_FP_NANFRAC_H): Ditto. (_FP_NANSIGN_H): Ditto. * config/riscv/t-softfp32 (softfp_extensions): Add HF related routines. (softfp_truncations): Ditto. (softfp_extras): Ditto. * config/riscv/t-softfp64 (softfp_extras): Add HF related routines. gcc/testsuite/ChangeLog: * gcc/testsuite/g++.target/riscv/_Float16.C: New. * gcc/testsuite/gcc.target/riscv/_Float16-soft-1.c: Ditto. * gcc/testsuite/gcc.target/riscv/_Float16-soft-2.c: Ditto. * gcc/testsuite/gcc.target/riscv/_Float16-soft-3.c: Ditto. * gcc/testsuite/gcc.target/riscv/_Float16-soft-4.c: Ditto. * gcc/testsuite/gcc.target/riscv/_Float16.c: Ditto. --- gcc/config/riscv/riscv-builtins.cc | 24 +++ gcc/config/riscv/riscv-modes.def | 1 + gcc/config/riscv/riscv.cc | 171 ++++++++++++++++-- gcc/config/riscv/riscv.md | 30 ++- gcc/testsuite/g++.target/riscv/_Float16.C | 18 ++ .../gcc.target/riscv/_Float16-soft-1.c | 9 + .../gcc.target/riscv/_Float16-soft-2.c | 13 ++ .../gcc.target/riscv/_Float16-soft-3.c | 12 ++ .../gcc.target/riscv/_Float16-soft-4.c | 12 ++ gcc/testsuite/gcc.target/riscv/_Float16.c | 19 ++ libgcc/config/riscv/sfp-machine.h | 3 + libgcc/config/riscv/t-softfp32 | 5 + libgcc/config/riscv/t-softfp64 | 1 + 13 files changed, 300 insertions(+), 18 deletions(-) create mode 100644 gcc/testsuite/g++.target/riscv/_Float16.C create mode 100644 gcc/testsuite/gcc.target/riscv/_Float16-soft-1.c create mode 100644 gcc/testsuite/gcc.target/riscv/_Float16-soft-2.c create mode 100644 gcc/testsuite/gcc.target/riscv/_Float16-soft-3.c create mode 100644 gcc/testsuite/gcc.target/riscv/_Float16-soft-4.c create mode 100644 gcc/testsuite/gcc.target/riscv/_Float16.c diff --git a/gcc/config/riscv/riscv-builtins.cc b/gcc/config/riscv/riscv-builtins.cc index 1218fdfc67d..3009311604d 100644 --- a/gcc/config/riscv/riscv-builtins.cc +++ b/gcc/config/riscv/riscv-builtins.cc @@ -34,6 +34,7 @@ along with GCC; see the file COPYING3. If not see #include "recog.h" #include "diagnostic-core.h" #include "stor-layout.h" +#include "stringpool.h" #include "expr.h" #include "langhooks.h" @@ -160,6 +161,8 @@ static GTY(()) int riscv_builtin_decl_index[NUM_INSN_CODES]; #define GET_BUILTIN_DECL(CODE) \ riscv_builtin_decls[riscv_builtin_decl_index[(CODE)]] +tree riscv_float16_type_node = NULL_TREE; + /* Return the function type associated with function prototype TYPE. */ static tree @@ -185,11 +188,32 @@ riscv_build_function_type (enum riscv_function_type type) return types[(int) type]; } +static void +riscv_init_builtin_types (void) +{ + /* Provide the _Float16 type and float16_type_node if needed. */ + if (!float16_type_node) + { + riscv_float16_type_node = make_node (REAL_TYPE); + TYPE_PRECISION (riscv_float16_type_node) = 16; + SET_TYPE_MODE (riscv_float16_type_node, HFmode); + layout_type (riscv_float16_type_node); + } + else + riscv_float16_type_node = float16_type_node; + + if (!maybe_get_identifier ("_Float16")) + lang_hooks.types.register_builtin_type (riscv_float16_type_node, + "_Float16"); +} + /* Implement TARGET_INIT_BUILTINS. */ void riscv_init_builtins (void) { + riscv_init_builtin_types (); + for (size_t i = 0; i < ARRAY_SIZE (riscv_builtins); i++) { const struct riscv_builtin_description *d = &riscv_builtins[i]; diff --git a/gcc/config/riscv/riscv-modes.def b/gcc/config/riscv/riscv-modes.def index 653228409a4..5cf2fc8e9e6 100644 --- a/gcc/config/riscv/riscv-modes.def +++ b/gcc/config/riscv/riscv-modes.def @@ -19,4 +19,5 @@ You should have received a copy of the GNU General Public License along with GCC; see the file COPYING3. If not see . */ +FLOAT_MODE (HF, 2, ieee_half_format); FLOAT_MODE (TF, 16, ieee_quad_format); diff --git a/gcc/config/riscv/riscv.cc b/gcc/config/riscv/riscv.cc index 5a0adffb5ce..47e6110767c 100644 --- a/gcc/config/riscv/riscv.cc +++ b/gcc/config/riscv/riscv.cc @@ -2296,11 +2296,13 @@ riscv_output_move (rtx dest, rtx src) enum rtx_code dest_code, src_code; machine_mode mode; bool dbl_p; + unsigned width; dest_code = GET_CODE (dest); src_code = GET_CODE (src); mode = GET_MODE (dest); dbl_p = (GET_MODE_SIZE (mode) == 8); + width = GET_MODE_SIZE (mode); if (dbl_p && riscv_split_64bit_move_p (dest, src)) return "#"; @@ -2308,10 +2310,19 @@ riscv_output_move (rtx dest, rtx src) if (dest_code == REG && GP_REG_P (REGNO (dest))) { if (src_code == REG && FP_REG_P (REGNO (src))) - return dbl_p ? "fmv.x.d\t%0,%1" : "fmv.x.w\t%0,%1"; + switch (width) + { + case 2: + /* Using fmv.x.s + sign-extend to emulate fmv.x.h. */ + return "fmv.x.s\t%0,%1;slli\t%0,%0,16;srai\t%0,%0,16"; + case 4: + return "fmv.x.s\t%0,%1"; + case 8: + return "fmv.x.d\t%0,%1"; + } if (src_code == MEM) - switch (GET_MODE_SIZE (mode)) + switch (width) { case 1: return "lbu\t%0,%1"; case 2: return "lhu\t%0,%1"; @@ -2353,18 +2364,24 @@ riscv_output_move (rtx dest, rtx src) return "mv\t%0,%z1"; if (FP_REG_P (REGNO (dest))) - { - if (!dbl_p) - return "fmv.w.x\t%0,%z1"; - if (TARGET_64BIT) - return "fmv.d.x\t%0,%z1"; - /* in RV32, we can emulate fmv.d.x %0, x0 using fcvt.d.w */ - gcc_assert (src == CONST0_RTX (mode)); - return "fcvt.d.w\t%0,x0"; - } + switch (width) + { + case 2: + /* High 16 bits should be all-1, otherwise HW will treated + as a n-bit canonical NaN, but isn't matter for softfloat. */ + return "fmv.s.x\t%0,%1"; + case 4: + return "fmv.s.x\t%0,%z1"; + case 8: + if (TARGET_64BIT) + return "fmv.d.x\t%0,%z1"; + /* in RV32, we can emulate fmv.d.x %0, x0 using fcvt.d.w */ + gcc_assert (src == CONST0_RTX (mode)); + return "fcvt.d.w\t%0,x0"; + } } if (dest_code == MEM) - switch (GET_MODE_SIZE (mode)) + switch (width) { case 1: return "sb\t%z1,%0"; case 2: return "sh\t%z1,%0"; @@ -2375,7 +2392,15 @@ riscv_output_move (rtx dest, rtx src) if (src_code == REG && FP_REG_P (REGNO (src))) { if (dest_code == REG && FP_REG_P (REGNO (dest))) - return dbl_p ? "fmv.d\t%0,%1" : "fmv.s\t%0,%1"; + switch (width) + { + case 2: + return "fmv.s\t%0,%1"; + case 4: + return "fmv.s\t%0,%1"; + case 8: + return "fmv.d\t%0,%1"; + } if (dest_code == MEM) return dbl_p ? "fsd\t%1,%0" : "fsw\t%1,%0"; @@ -5638,6 +5663,107 @@ riscv_asan_shadow_offset (void) return TARGET_64BIT ? (HOST_WIDE_INT_1 << 29) : 0; } +/* Implement TARGET_MANGLE_TYPE. */ + +static const char * +riscv_mangle_type (const_tree type) +{ + /* Half-precision float, _Float16 is "DF16_". */ + if (TREE_CODE (type) == REAL_TYPE && TYPE_PRECISION (type) == 16) + return "DF16_"; + + /* Use the default mangling. */ + return NULL; +} + +/* Implement TARGET_SCALAR_MODE_SUPPORTED_P. */ + +static bool +riscv_scalar_mode_supported_p (scalar_mode mode) +{ + if (mode == HFmode) + return true; + else + return default_scalar_mode_supported_p (mode); +} + +/* Implement TARGET_LIBGCC_FLOATING_MODE_SUPPORTED_P - return TRUE + if MODE is HFmode, and punt to the generic implementation otherwise. */ + +static bool +riscv_libgcc_floating_mode_supported_p (scalar_float_mode mode) +{ + if (mode == HFmode) + return true; + else + return default_libgcc_floating_mode_supported_p (mode); +} + +/* Set the value of FLT_EVAL_METHOD. + ISO/IEC TS 18661-3 defines two values that we'd like to make use of: + + 0: evaluate all operations and constants, whose semantic type has at + most the range and precision of type float, to the range and + precision of float; evaluate all other operations and constants to + the range and precision of the semantic type; + + N, where _FloatN is a supported interchange floating type + evaluate all operations and constants, whose semantic type has at + most the range and precision of _FloatN type, to the range and + precision of the _FloatN type; evaluate all other operations and + constants to the range and precision of the semantic type; + + If we have the zfh extensions then we support _Float16 in native + precision, so we should set this to 16. */ +static enum flt_eval_method +riscv_excess_precision (enum excess_precision_type type) +{ + switch (type) + { + case EXCESS_PRECISION_TYPE_FAST: + case EXCESS_PRECISION_TYPE_STANDARD: + return FLT_EVAL_METHOD_PROMOTE_TO_FLOAT; + case EXCESS_PRECISION_TYPE_IMPLICIT: + return FLT_EVAL_METHOD_PROMOTE_TO_FLOAT16; + default: + gcc_unreachable (); + } + return FLT_EVAL_METHOD_UNPREDICTABLE; +} + +/* Implement TARGET_FLOATN_MODE. */ +static opt_scalar_float_mode +riscv_floatn_mode (int n, bool extended) +{ + if (!extended && n == 16) + return HFmode; + + return default_floatn_mode (n, extended); +} + +static void +riscv_init_libfuncs (void) +{ + /* Half-precision float operations. The compiler handles all operations + with NULL libfuncs by converting to SFmode. */ + + /* Arithmetic. */ + set_optab_libfunc (add_optab, HFmode, NULL); + set_optab_libfunc (sdiv_optab, HFmode, NULL); + set_optab_libfunc (smul_optab, HFmode, NULL); + set_optab_libfunc (neg_optab, HFmode, NULL); + set_optab_libfunc (sub_optab, HFmode, NULL); + + /* Comparisons. */ + set_optab_libfunc (eq_optab, HFmode, NULL); + set_optab_libfunc (ne_optab, HFmode, NULL); + set_optab_libfunc (lt_optab, HFmode, NULL); + set_optab_libfunc (le_optab, HFmode, NULL); + set_optab_libfunc (ge_optab, HFmode, NULL); + set_optab_libfunc (gt_optab, HFmode, NULL); + set_optab_libfunc (unord_optab, HFmode, NULL); +} + /* Initialize the GCC target structure. */ #undef TARGET_ASM_ALIGNED_HI_OP #define TARGET_ASM_ALIGNED_HI_OP "\t.half\t" @@ -5821,6 +5947,25 @@ riscv_asan_shadow_offset (void) #undef TARGET_NEW_ADDRESS_PROFITABLE_P #define TARGET_NEW_ADDRESS_PROFITABLE_P riscv_new_address_profitable_p +#undef TARGET_MANGLE_TYPE +#define TARGET_MANGLE_TYPE riscv_mangle_type + +#undef TARGET_SCALAR_MODE_SUPPORTED_P +#define TARGET_SCALAR_MODE_SUPPORTED_P riscv_scalar_mode_supported_p + +#undef TARGET_LIBGCC_FLOATING_MODE_SUPPORTED_P +#define TARGET_LIBGCC_FLOATING_MODE_SUPPORTED_P \ + riscv_libgcc_floating_mode_supported_p + +#undef TARGET_INIT_LIBFUNCS +#define TARGET_INIT_LIBFUNCS riscv_init_libfuncs + +#undef TARGET_C_EXCESS_PRECISION +#define TARGET_C_EXCESS_PRECISION riscv_excess_precision + +#undef TARGET_FLOATN_MODE +#define TARGET_FLOATN_MODE riscv_floatn_mode + #undef TARGET_ASAN_SHADOW_OFFSET #define TARGET_ASAN_SHADOW_OFFSET riscv_asan_shadow_offset diff --git a/gcc/config/riscv/riscv.md b/gcc/config/riscv/riscv.md index 0796f91dd30..7af5c90e0f3 100644 --- a/gcc/config/riscv/riscv.md +++ b/gcc/config/riscv/riscv.md @@ -134,7 +134,7 @@ (define_attr "move_type" (const_string "unknown")) ;; Main data type used by the insn -(define_attr "mode" "unknown,none,QI,HI,SI,DI,TI,SF,DF,TF" +(define_attr "mode" "unknown,none,QI,HI,SI,DI,TI,HF,SF,DF,TF" (const_string "unknown")) ;; True if the main data type is twice the size of a word. @@ -324,20 +324,20 @@ (define_mode_attr load [(QI "lb") (HI "lh") (SI "lw") (DI "ld") (SF "flw") (DF " (define_mode_attr default_load [(QI "lbu") (HI "lhu") (SI "lw") (DI "ld")]) ;; Mode attribute for FP loads into integer registers. -(define_mode_attr softload [(SF "lw") (DF "ld")]) +(define_mode_attr softload [(HF "lh") (SF "lw") (DF "ld")]) ;; Instruction names for stores. (define_mode_attr store [(QI "sb") (HI "sh") (SI "sw") (DI "sd") (SF "fsw") (DF "fsd")]) ;; Instruction names for FP stores from integer registers. -(define_mode_attr softstore [(SF "sw") (DF "sd")]) +(define_mode_attr softstore [(HF "sh") (SF "sw") (DF "sd")]) ;; This attribute gives the best constraint to use for registers of ;; a given mode. (define_mode_attr reg [(SI "d") (DI "d") (CC "d")]) ;; This attribute gives the format suffix for floating-point operations. -(define_mode_attr fmt [(SF "s") (DF "d")]) +(define_mode_attr fmt [(HF "h") (SF "s") (DF "d")]) ;; This attribute gives the integer suffix for floating-point conversions. (define_mode_attr ifmt [(SI "w") (DI "l")]) @@ -347,7 +347,7 @@ (define_mode_attr amo [(SI "w") (DI "d")]) ;; This attribute gives the upper-case mode name for one unit of a ;; floating-point mode. -(define_mode_attr UNITMODE [(SF "SF") (DF "DF")]) +(define_mode_attr UNITMODE [(HF "HF") (SF "SF") (DF "DF")]) ;; This attribute gives the integer mode that has half the size of ;; the controlling mode. @@ -1450,6 +1450,26 @@ (define_insn "extendsfdf2" [(set_attr "type" "fcvt") (set_attr "mode" "DF")]) +;; 16-bit floating point moves +(define_expand "movhf" + [(set (match_operand:HF 0 "") + (match_operand:HF 1 ""))] + "" +{ + if (riscv_legitimize_move (HFmode, operands[0], operands[1])) + DONE; +}) + + +(define_insn "*movhf_softfloat" + [(set (match_operand:HF 0 "nonimmediate_operand" "=f, r,r,m,*f,*r") + (match_operand:HF 1 "move_operand" " f,Gr,m,r,*r,*f"))] + "(register_operand (operands[0], HFmode) + || reg_or_0_operand (operands[1], HFmode))" + { return riscv_output_move (operands[0], operands[1]); } + [(set_attr "move_type" "fmove,move,load,store,mtc,mfc") + (set_attr "mode" "HF")]) + ;; ;; .................... ;; diff --git a/gcc/testsuite/g++.target/riscv/_Float16.C b/gcc/testsuite/g++.target/riscv/_Float16.C new file mode 100644 index 00000000000..b97d1fbcf1c --- /dev/null +++ b/gcc/testsuite/g++.target/riscv/_Float16.C @@ -0,0 +1,18 @@ +/* { dg-do compile } */ + +_Float16 x; + +_Float16 foo1 (_Float16 a, _Float16 b) +{ + return a + b; +} + +_Float16 foo2 (_Float16 a, _Float16 b) +{ + return a * b; +} + +int foo3 (_Float16 a, _Float16 b) +{ + return a > b; +} diff --git a/gcc/testsuite/gcc.target/riscv/_Float16-soft-1.c b/gcc/testsuite/gcc.target/riscv/_Float16-soft-1.c new file mode 100644 index 00000000000..0622588fdb9 --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/_Float16-soft-1.c @@ -0,0 +1,9 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64if -mabi=lp64f -O" } */ + +_Float16 test_soft_move (_Float16 a, _Float16 b) +{ + return b; +} + +/* { dg-final { scan-assembler-not "fmv.h" } } */ diff --git a/gcc/testsuite/gcc.target/riscv/_Float16-soft-2.c b/gcc/testsuite/gcc.target/riscv/_Float16-soft-2.c new file mode 100644 index 00000000000..3d37823fa4d --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/_Float16-soft-2.c @@ -0,0 +1,13 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64if -mabi=lp64f -O" } */ + +_Float16 test_soft_add (_Float16 a, _Float16 b) +{ + /* Make sure __addhf3 not invoked here. */ + /* { dg-final { scan-assembler-times "call\t__extendhfsf2" 2 } } */ + return a + b; + /* { dg-final { scan-assembler-not "call\t__addhf3" } } */ + /* { dg-final { scan-assembler-times "fadd.s" 1 } } */ + /* { dg-final { scan-assembler-times "call\t__truncsfhf2" 1 } } */ +} + diff --git a/gcc/testsuite/gcc.target/riscv/_Float16-soft-3.c b/gcc/testsuite/gcc.target/riscv/_Float16-soft-3.c new file mode 100644 index 00000000000..ecce364e310 --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/_Float16-soft-3.c @@ -0,0 +1,12 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64if -mabi=lp64f -O" } */ + +int test_soft_compare (_Float16 a, _Float16 b) +{ + /* Make sure __gthf2 not invoked here. */ + /* { dg-final { scan-assembler-times "call\t__extendhfsf2" 2 } } */ + return a > b; + /* { dg-final { scan-assembler-not "call\t__gthf2" } } */ + /* { dg-final { scan-assembler-times "fgt.s" 1 } } */ +} + diff --git a/gcc/testsuite/gcc.target/riscv/_Float16-soft-4.c b/gcc/testsuite/gcc.target/riscv/_Float16-soft-4.c new file mode 100644 index 00000000000..ce4bc7a6a5d --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/_Float16-soft-4.c @@ -0,0 +1,12 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64i -mabi=lp64 -O" } */ + +int test_soft_compare (_Float16 a, _Float16 b) +{ + /* Make sure __gthf2 not invoked here. */ + /* { dg-final { scan-assembler-times "call\t__extendhfsf2" 2 } } */ + return a > b; + /* { dg-final { scan-assembler-not "call\t__gthf2" } } */ + /* { dg-final { scan-assembler-times "call\t__gtsf2" 1 } } */ +} + diff --git a/gcc/testsuite/gcc.target/riscv/_Float16.c b/gcc/testsuite/gcc.target/riscv/_Float16.c new file mode 100644 index 00000000000..cc131fd3e2c --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/_Float16.c @@ -0,0 +1,19 @@ +/* { dg-do compile } */ +/* { dg-options "-O" } */ + +_Float16 x; + +_Float16 foo1 (_Float16 a, _Float16 b) +{ + return a + b; +} + +_Float16 foo2 (_Float16 a, _Float16 b) +{ + return a * b; +} + +int foo3 (_Float16 a, _Float16 b) +{ + return a > b; +} diff --git a/libgcc/config/riscv/sfp-machine.h b/libgcc/config/riscv/sfp-machine.h index 27fbe6e84d6..b2901f0b1fd 100644 --- a/libgcc/config/riscv/sfp-machine.h +++ b/libgcc/config/riscv/sfp-machine.h @@ -41,6 +41,7 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see #define _FP_DIV_MEAT_D(R,X,Y) _FP_DIV_MEAT_2_udiv(D,R,X,Y) #define _FP_DIV_MEAT_Q(R,X,Y) _FP_DIV_MEAT_4_udiv(Q,R,X,Y) +#define _FP_NANFRAC_H _FP_QNANBIT_H #define _FP_NANFRAC_S _FP_QNANBIT_S #define _FP_NANFRAC_D _FP_QNANBIT_D, 0 #define _FP_NANFRAC_Q _FP_QNANBIT_Q, 0, 0, 0 @@ -63,6 +64,7 @@ see the files COPYING3 and COPYING.RUNTIME respectively. If not, see #define _FP_DIV_MEAT_D(R,X,Y) _FP_DIV_MEAT_1_udiv_norm(D,R,X,Y) #define _FP_DIV_MEAT_Q(R,X,Y) _FP_DIV_MEAT_2_udiv(Q,R,X,Y) +#define _FP_NANFRAC_H _FP_QNANBIT_H #define _FP_NANFRAC_S _FP_QNANBIT_S #define _FP_NANFRAC_D _FP_QNANBIT_D #define _FP_NANFRAC_Q _FP_QNANBIT_Q, 0 @@ -80,6 +82,7 @@ typedef unsigned int UTItype __attribute__ ((mode (TI))); typedef int __gcc_CMPtype __attribute__ ((mode (__libgcc_cmp_return__))); #define CMPtype __gcc_CMPtype +#define _FP_NANSIGN_H 0 #define _FP_NANSIGN_S 0 #define _FP_NANSIGN_D 0 #define _FP_NANSIGN_Q 0 diff --git a/libgcc/config/riscv/t-softfp32 b/libgcc/config/riscv/t-softfp32 index 59be1df827e..55261f98383 100644 --- a/libgcc/config/riscv/t-softfp32 +++ b/libgcc/config/riscv/t-softfp32 @@ -41,3 +41,8 @@ else softfp_extras := divsf3 divdf3 divtf3 endif + +softfp_extensions += hfsf hfdf hftf +softfp_truncations += tfhf dfhf sfhf +softfp_extras += fixhfsi fixhfdi fixunshfsi fixunshfdi \ + floatsihf floatdihf floatunsihf floatundihf diff --git a/libgcc/config/riscv/t-softfp64 b/libgcc/config/riscv/t-softfp64 index 75870951202..c87d242d5c3 100644 --- a/libgcc/config/riscv/t-softfp64 +++ b/libgcc/config/riscv/t-softfp64 @@ -1,3 +1,4 @@ include $(srcdir)/config/riscv/t-softfp32 softfp_int_modes += ti +softfp_extras += fixhfti fixunshfti floattihf floatuntihf