From patchwork Fri Apr 12 07:37:07 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Huang Tao X-Patchwork-Id: 13627312 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 3BE77C4345F for ; Fri, 12 Apr 2024 08:54:01 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1rvCfJ-00053Y-Qd; Fri, 12 Apr 2024 04:53:45 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1rvCfH-00052k-IU; Fri, 12 Apr 2024 04:53:43 -0400 Received: from out30-133.freemail.mail.aliyun.com ([115.124.30.133]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1rvCf5-0005zM-Vh; Fri, 12 Apr 2024 04:53:43 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.alibaba.com; s=default; t=1712912007; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=RIg6PWSzgeHhBKgFWo9JZiHi0QPSvyckGd64FC9/oh8=; b=g+YUsxQ7ROetcFgdA2nNZcpi06h1dT3rWsjz3Cjyr5Nhp/+SnewLaDrkKP2yy24eNQOb0kvSd3FvJepuakWFMUF4UcKVus/TItTTbtA6jVCyLOVcG+ovukeqGxUIY/SdLK1GDQvc/eM21bNjO1BI675xMApEDmE0hfbkAKM4yv0= X-Alimail-AntiSpam: AC=PASS; BC=-1|-1; BR=01201311R141e4; CH=green; DM=||false|; DS=||; FP=0|-1|-1|-1|0|-1|-1|-1; HT=ay29a033018045192; MF=eric.huang@linux.alibaba.com; NM=1; PH=DS; RN=9; SR=0; TI=SMTPD_---0W4Nc.B4_1712912005; Received: from localhost.localdomain(mailfrom:eric.huang@linux.alibaba.com fp:SMTPD_---0W4Nc.B4_1712912005) by smtp.aliyun-inc.com; Fri, 12 Apr 2024 16:53:26 +0800 From: Huang Tao To: qemu-devel@nongnu.org Cc: qemu-riscv@nongnu.org, zhiwei_liu@linux.alibaba.com, dbarboza@ventanamicro.com, liwei1518@gmail.com, bin.meng@windriver.com, alistair.francis@wdc.com, palmer@dabbelt.com, Huang Tao Subject: [PATCH 37/65] target/riscv: Add widening floating-point add/sub instructions for XTheadVector Date: Fri, 12 Apr 2024 15:37:07 +0800 Message-ID: <20240412073735.76413-38-eric.huang@linux.alibaba.com> X-Mailer: git-send-email 2.44.0 In-Reply-To: <20240412073735.76413-1-eric.huang@linux.alibaba.com> References: <20240412073735.76413-1-eric.huang@linux.alibaba.com> MIME-Version: 1.0 Received-SPF: pass client-ip=115.124.30.133; envelope-from=eric.huang@linux.alibaba.com; helo=out30-133.freemail.mail.aliyun.com X-Spam_score_int: -174 X-Spam_score: -17.5 X-Spam_bar: ----------------- X-Spam_report: (-17.5 / 5.0 requ) BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, ENV_AND_HDR_SPF_MATCH=-0.5, RCVD_IN_DNSWL_NONE=-0.0001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, UNPARSEABLE_RELAY=0.001, USER_IN_DEF_DKIM_WL=-7.5, USER_IN_DEF_SPF_WL=-7.5 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org The instructions have the same function as RVV1.0. Overall there are only general differences between XTheadVector and RVV1.0. Signed-off-by: Huang Tao --- target/riscv/helper.h | 17 ++ .../riscv/insn_trans/trans_xtheadvector.c.inc | 162 +++++++++++++++++- target/riscv/vector_helper.c | 16 +- target/riscv/vector_internals.h | 9 + target/riscv/xtheadvector_helper.c | 38 ++++ 5 files changed, 226 insertions(+), 16 deletions(-) diff --git a/target/riscv/helper.h b/target/riscv/helper.h index 04bd363ac0..21916e9e3c 100644 --- a/target/riscv/helper.h +++ b/target/riscv/helper.h @@ -2012,3 +2012,20 @@ DEF_HELPER_6(th_vfsub_vf_d, void, ptr, ptr, i64, ptr, env, i32) DEF_HELPER_6(th_vfrsub_vf_h, void, ptr, ptr, i64, ptr, env, i32) DEF_HELPER_6(th_vfrsub_vf_w, void, ptr, ptr, i64, ptr, env, i32) DEF_HELPER_6(th_vfrsub_vf_d, void, ptr, ptr, i64, ptr, env, i32) + +DEF_HELPER_6(th_vfwadd_vv_h, void, ptr, ptr, ptr, ptr, env, i32) +DEF_HELPER_6(th_vfwadd_vv_w, void, ptr, ptr, ptr, ptr, env, i32) +DEF_HELPER_6(th_vfwsub_vv_h, void, ptr, ptr, ptr, ptr, env, i32) +DEF_HELPER_6(th_vfwsub_vv_w, void, ptr, ptr, ptr, ptr, env, i32) +DEF_HELPER_6(th_vfwadd_wv_h, void, ptr, ptr, ptr, ptr, env, i32) +DEF_HELPER_6(th_vfwadd_wv_w, void, ptr, ptr, ptr, ptr, env, i32) +DEF_HELPER_6(th_vfwsub_wv_h, void, ptr, ptr, ptr, ptr, env, i32) +DEF_HELPER_6(th_vfwsub_wv_w, void, ptr, ptr, ptr, ptr, env, i32) +DEF_HELPER_6(th_vfwadd_vf_h, void, ptr, ptr, i64, ptr, env, i32) +DEF_HELPER_6(th_vfwadd_vf_w, void, ptr, ptr, i64, ptr, env, i32) +DEF_HELPER_6(th_vfwsub_vf_h, void, ptr, ptr, i64, ptr, env, i32) +DEF_HELPER_6(th_vfwsub_vf_w, void, ptr, ptr, i64, ptr, env, i32) +DEF_HELPER_6(th_vfwadd_wf_h, void, ptr, ptr, i64, ptr, env, i32) +DEF_HELPER_6(th_vfwadd_wf_w, void, ptr, ptr, i64, ptr, env, i32) +DEF_HELPER_6(th_vfwsub_wf_h, void, ptr, ptr, i64, ptr, env, i32) +DEF_HELPER_6(th_vfwsub_wf_w, void, ptr, ptr, i64, ptr, env, i32) diff --git a/target/riscv/insn_trans/trans_xtheadvector.c.inc b/target/riscv/insn_trans/trans_xtheadvector.c.inc index a18c661f24..64d7a7fb76 100644 --- a/target/riscv/insn_trans/trans_xtheadvector.c.inc +++ b/target/riscv/insn_trans/trans_xtheadvector.c.inc @@ -1854,20 +1854,166 @@ GEN_OPFVF_TRANS_TH(th_vfadd_vf, opfvf_check_th) GEN_OPFVF_TRANS_TH(th_vfsub_vf, opfvf_check_th) GEN_OPFVF_TRANS_TH(th_vfrsub_vf, opfvf_check_th) +/* Vector Widening Floating-Point Add/Subtract Instructions */ +static bool opfvv_widen_check_th(DisasContext *s, arg_rmrr *a) +{ + return (require_xtheadvector(s) && + vext_check_isa_ill(s) && + th_check_overlap_mask(s, a->rd, a->vm, true) && + th_check_reg(s, a->rd, true) && + th_check_reg(s, a->rs2, false) && + th_check_reg(s, a->rs1, false) && + th_check_overlap_group(a->rd, 2 << s->lmul, a->rs2, + 1 << s->lmul) && + th_check_overlap_group(a->rd, 2 << s->lmul, a->rs1, + 1 << s->lmul) && + (s->lmul < 0x3) && (s->sew < 0x3) && (s->sew != 0)); +} + +/* OPFVV with WIDEN */ +#define GEN_OPFVV_WIDEN_TRANS_TH(NAME, CHECK) \ +static bool trans_##NAME(DisasContext *s, arg_rmrr *a) \ +{ \ + if (CHECK(s, a)) { \ + uint32_t data = 0; \ + static gen_helper_gvec_4_ptr * const fns[2] = { \ + gen_helper_##NAME##_h, gen_helper_##NAME##_w, \ + }; \ + gen_set_rm(s, RISCV_FRM_DYN); \ + \ + data = FIELD_DP32(data, VDATA_TH, MLEN, s->mlen); \ + data = FIELD_DP32(data, VDATA_TH, VM, a->vm); \ + data = FIELD_DP32(data, VDATA_TH, LMUL, s->lmul); \ + tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0), \ + vreg_ofs(s, a->rs1), \ + vreg_ofs(s, a->rs2), tcg_env, \ + s->cfg_ptr->vlenb, \ + s->cfg_ptr->vlenb, data, \ + fns[s->sew - 1]); \ + finalize_rvv_inst(s); \ + return true; \ + } \ + return false; \ +} + +GEN_OPFVV_WIDEN_TRANS_TH(th_vfwadd_vv, opfvv_widen_check_th) +GEN_OPFVV_WIDEN_TRANS_TH(th_vfwsub_vv, opfvv_widen_check_th) + +static bool opfvf_widen_check_th(DisasContext *s, arg_rmrr *a) +{ + return (require_xtheadvector(s) && + vext_check_isa_ill(s) && + th_check_overlap_mask(s, a->rd, a->vm, true) && + th_check_reg(s, a->rd, true) && + th_check_reg(s, a->rs2, false) && + th_check_overlap_group(a->rd, 2 << s->lmul, a->rs2, + 1 << s->lmul) && + (s->lmul < 0x3) && (s->sew < 0x3) && (s->sew != 0)); +} + +/* OPFVF with WIDEN */ +#define GEN_OPFVF_WIDEN_TRANS_TH(NAME) \ +static bool trans_##NAME(DisasContext *s, arg_rmrr *a) \ +{ \ + if (opfvf_widen_check_th(s, a)) { \ + uint32_t data = 0; \ + static gen_helper_opfvf *const fns[2] = { \ + gen_helper_##NAME##_h, gen_helper_##NAME##_w, \ + }; \ + gen_set_rm(s, RISCV_FRM_DYN); \ + data = FIELD_DP32(data, VDATA_TH, MLEN, s->mlen); \ + data = FIELD_DP32(data, VDATA_TH, VM, a->vm); \ + data = FIELD_DP32(data, VDATA_TH, LMUL, s->lmul); \ + return opfvf_trans(a->rd, a->rs1, a->rs2, data, \ + fns[s->sew - 1], s); \ + } \ + return false; \ +} + +GEN_OPFVF_WIDEN_TRANS_TH(th_vfwadd_vf) +GEN_OPFVF_WIDEN_TRANS_TH(th_vfwsub_vf) + +static bool opfwv_widen_check_th(DisasContext *s, arg_rmrr *a) +{ + return (require_xtheadvector(s) && + vext_check_isa_ill(s) && + th_check_overlap_mask(s, a->rd, a->vm, true) && + th_check_reg(s, a->rd, true) && + th_check_reg(s, a->rs2, true) && + th_check_reg(s, a->rs1, false) && + th_check_overlap_group(a->rd, 2 << s->lmul, a->rs1, + 1 << s->lmul) && + (s->lmul < 0x3) && (s->sew < 0x3) && (s->sew != 0)); +} + +/* WIDEN OPFVV with WIDEN */ +#define GEN_OPFWV_WIDEN_TRANS_TH(NAME) \ +static bool trans_##NAME(DisasContext *s, arg_rmrr *a) \ +{ \ + if (opfwv_widen_check_th(s, a)) { \ + uint32_t data = 0; \ + static gen_helper_gvec_4_ptr * const fns[2] = { \ + gen_helper_##NAME##_h, gen_helper_##NAME##_w, \ + }; \ + gen_set_rm(s, RISCV_FRM_DYN); \ + \ + data = FIELD_DP32(data, VDATA_TH, MLEN, s->mlen); \ + data = FIELD_DP32(data, VDATA_TH, VM, a->vm); \ + data = FIELD_DP32(data, VDATA_TH, LMUL, s->lmul); \ + tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), \ + vreg_ofs(s, 0), \ + vreg_ofs(s, a->rs1), \ + vreg_ofs(s, a->rs2), tcg_env, \ + s->cfg_ptr->vlenb, \ + s->cfg_ptr->vlenb, data, \ + fns[s->sew - 1]); \ + finalize_rvv_inst(s); \ + return true; \ + } \ + return false; \ +} + +GEN_OPFWV_WIDEN_TRANS_TH(th_vfwadd_wv) +GEN_OPFWV_WIDEN_TRANS_TH(th_vfwsub_wv) + +static bool opfwf_widen_check_th(DisasContext *s, arg_rmrr *a) +{ + return (require_xtheadvector(s) && + vext_check_isa_ill(s) && + th_check_overlap_mask(s, a->rd, a->vm, true) && + th_check_reg(s, a->rd, true) && + th_check_reg(s, a->rs2, true) && + (s->lmul < 0x3) && (s->sew < 0x3) && (s->sew != 0)); +} + +/* WIDEN OPFVF with WIDEN */ +#define GEN_OPFWF_WIDEN_TRANS_TH(NAME) \ +static bool trans_##NAME(DisasContext *s, arg_rmrr *a) \ +{ \ + if (opfwf_widen_check_th(s, a)) { \ + uint32_t data = 0; \ + static gen_helper_opfvf *const fns[2] = { \ + gen_helper_##NAME##_h, gen_helper_##NAME##_w, \ + }; \ + gen_set_rm(s, RISCV_FRM_DYN); \ + data = FIELD_DP32(data, VDATA_TH, MLEN, s->mlen); \ + data = FIELD_DP32(data, VDATA_TH, VM, a->vm); \ + data = FIELD_DP32(data, VDATA_TH, LMUL, s->lmul); \ + return opfvf_trans(a->rd, a->rs1, a->rs2, data, \ + fns[s->sew - 1], s); \ + } \ + return false; \ +} + +GEN_OPFWF_WIDEN_TRANS_TH(th_vfwadd_wf) +GEN_OPFWF_WIDEN_TRANS_TH(th_vfwsub_wf) + #define TH_TRANS_STUB(NAME) \ static bool trans_##NAME(DisasContext *s, arg_##NAME *a) \ { \ return require_xtheadvector(s); \ } -TH_TRANS_STUB(th_vfwadd_vv) -TH_TRANS_STUB(th_vfwadd_vf) -TH_TRANS_STUB(th_vfwadd_wv) -TH_TRANS_STUB(th_vfwadd_wf) -TH_TRANS_STUB(th_vfwsub_vv) -TH_TRANS_STUB(th_vfwsub_vf) -TH_TRANS_STUB(th_vfwsub_wv) -TH_TRANS_STUB(th_vfwsub_wf) TH_TRANS_STUB(th_vfmul_vv) TH_TRANS_STUB(th_vfmul_vf) TH_TRANS_STUB(th_vfdiv_vv) diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c index 3784096da2..6d0358876a 100644 --- a/target/riscv/vector_helper.c +++ b/target/riscv/vector_helper.c @@ -2927,13 +2927,13 @@ GEN_VEXT_VF(vfrsub_vf_w, 4) GEN_VEXT_VF(vfrsub_vf_d, 8) /* Vector Widening Floating-Point Add/Subtract Instructions */ -static uint32_t vfwadd16(uint16_t a, uint16_t b, float_status *s) +uint32_t vfwadd16(uint16_t a, uint16_t b, float_status *s) { return float32_add(float16_to_float32(a, true, s), float16_to_float32(b, true, s), s); } -static uint64_t vfwadd32(uint32_t a, uint32_t b, float_status *s) +uint64_t vfwadd32(uint32_t a, uint32_t b, float_status *s) { return float64_add(float32_to_float64(a, s), float32_to_float64(b, s), s); @@ -2949,13 +2949,13 @@ RVVCALL(OPFVF2, vfwadd_vf_w, WOP_UUU_W, H8, H4, vfwadd32) GEN_VEXT_VF(vfwadd_vf_h, 4) GEN_VEXT_VF(vfwadd_vf_w, 8) -static uint32_t vfwsub16(uint16_t a, uint16_t b, float_status *s) +uint32_t vfwsub16(uint16_t a, uint16_t b, float_status *s) { return float32_sub(float16_to_float32(a, true, s), float16_to_float32(b, true, s), s); } -static uint64_t vfwsub32(uint32_t a, uint32_t b, float_status *s) +uint64_t vfwsub32(uint32_t a, uint32_t b, float_status *s) { return float64_sub(float32_to_float64(a, s), float32_to_float64(b, s), s); @@ -2971,12 +2971,12 @@ RVVCALL(OPFVF2, vfwsub_vf_w, WOP_UUU_W, H8, H4, vfwsub32) GEN_VEXT_VF(vfwsub_vf_h, 4) GEN_VEXT_VF(vfwsub_vf_w, 8) -static uint32_t vfwaddw16(uint32_t a, uint16_t b, float_status *s) +uint32_t vfwaddw16(uint32_t a, uint16_t b, float_status *s) { return float32_add(a, float16_to_float32(b, true, s), s); } -static uint64_t vfwaddw32(uint64_t a, uint32_t b, float_status *s) +uint64_t vfwaddw32(uint64_t a, uint32_t b, float_status *s) { return float64_add(a, float32_to_float64(b, s), s); } @@ -2990,12 +2990,12 @@ RVVCALL(OPFVF2, vfwadd_wf_w, WOP_WUUU_W, H8, H4, vfwaddw32) GEN_VEXT_VF(vfwadd_wf_h, 4) GEN_VEXT_VF(vfwadd_wf_w, 8) -static uint32_t vfwsubw16(uint32_t a, uint16_t b, float_status *s) +uint32_t vfwsubw16(uint32_t a, uint16_t b, float_status *s) { return float32_sub(a, float16_to_float32(b, true, s), s); } -static uint64_t vfwsubw32(uint64_t a, uint32_t b, float_status *s) +uint64_t vfwsubw32(uint64_t a, uint32_t b, float_status *s) { return float64_sub(a, float32_to_float64(b, s), s); } diff --git a/target/riscv/vector_internals.h b/target/riscv/vector_internals.h index 5f250ab7ba..0786f5a4e1 100644 --- a/target/riscv/vector_internals.h +++ b/target/riscv/vector_internals.h @@ -344,4 +344,13 @@ uint16_t float16_rsub(uint16_t a, uint16_t b, float_status *s); uint32_t float32_rsub(uint32_t a, uint32_t b, float_status *s); uint64_t float64_rsub(uint64_t a, uint64_t b, float_status *s); +uint32_t vfwadd16(uint16_t a, uint16_t b, float_status *s); +uint64_t vfwadd32(uint32_t a, uint32_t b, float_status *s); +uint32_t vfwsub16(uint16_t a, uint16_t b, float_status *s); +uint64_t vfwsub32(uint32_t a, uint32_t b, float_status *s); +uint32_t vfwaddw16(uint32_t a, uint16_t b, float_status *s); +uint64_t vfwaddw32(uint64_t a, uint32_t b, float_status *s); +uint32_t vfwsubw16(uint32_t a, uint16_t b, float_status *s); +uint64_t vfwsubw32(uint64_t a, uint32_t b, float_status *s); + #endif /* TARGET_RISCV_VECTOR_INTERNALS_H */ diff --git a/target/riscv/xtheadvector_helper.c b/target/riscv/xtheadvector_helper.c index 60811ca813..cab489a4ae 100644 --- a/target/riscv/xtheadvector_helper.c +++ b/target/riscv/xtheadvector_helper.c @@ -2696,3 +2696,41 @@ THCALL(TH_OPFVF2, th_vfrsub_vf_d, OP_UUU_D, H8, H8, float64_rsub) GEN_TH_VF(th_vfrsub_vf_h, 2, 2, clearh_th) GEN_TH_VF(th_vfrsub_vf_w, 4, 4, clearl_th) GEN_TH_VF(th_vfrsub_vf_d, 8, 8, clearq_th) + +/* Vector Widening Floating-Point Add/Subtract Instructions */ + +THCALL(TH_OPFVV2, th_vfwadd_vv_h, WOP_UUU_H, H4, H2, H2, vfwadd16) +THCALL(TH_OPFVV2, th_vfwadd_vv_w, WOP_UUU_W, H8, H4, H4, vfwadd32) +GEN_TH_VV_ENV(th_vfwadd_vv_h, 2, 4, clearl_th) +GEN_TH_VV_ENV(th_vfwadd_vv_w, 4, 8, clearq_th) +THCALL(TH_OPFVF2, th_vfwadd_vf_h, WOP_UUU_H, H4, H2, vfwadd16) +THCALL(TH_OPFVF2, th_vfwadd_vf_w, WOP_UUU_W, H8, H4, vfwadd32) +GEN_TH_VF(th_vfwadd_vf_h, 2, 4, clearl_th) +GEN_TH_VF(th_vfwadd_vf_w, 4, 8, clearq_th) + +THCALL(TH_OPFVV2, th_vfwsub_vv_h, WOP_UUU_H, H4, H2, H2, vfwsub16) +THCALL(TH_OPFVV2, th_vfwsub_vv_w, WOP_UUU_W, H8, H4, H4, vfwsub32) +GEN_TH_VV_ENV(th_vfwsub_vv_h, 2, 4, clearl_th) +GEN_TH_VV_ENV(th_vfwsub_vv_w, 4, 8, clearq_th) +THCALL(TH_OPFVF2, th_vfwsub_vf_h, WOP_UUU_H, H4, H2, vfwsub16) +THCALL(TH_OPFVF2, th_vfwsub_vf_w, WOP_UUU_W, H8, H4, vfwsub32) +GEN_TH_VF(th_vfwsub_vf_h, 2, 4, clearl_th) +GEN_TH_VF(th_vfwsub_vf_w, 4, 8, clearq_th) + +THCALL(TH_OPFVV2, th_vfwadd_wv_h, WOP_WUUU_H, H4, H2, H2, vfwaddw16) +THCALL(TH_OPFVV2, th_vfwadd_wv_w, WOP_WUUU_W, H8, H4, H4, vfwaddw32) +GEN_TH_VV_ENV(th_vfwadd_wv_h, 2, 4, clearl_th) +GEN_TH_VV_ENV(th_vfwadd_wv_w, 4, 8, clearq_th) +THCALL(TH_OPFVF2, th_vfwadd_wf_h, WOP_WUUU_H, H4, H2, vfwaddw16) +THCALL(TH_OPFVF2, th_vfwadd_wf_w, WOP_WUUU_W, H8, H4, vfwaddw32) +GEN_TH_VF(th_vfwadd_wf_h, 2, 4, clearl_th) +GEN_TH_VF(th_vfwadd_wf_w, 4, 8, clearq_th) + +THCALL(TH_OPFVV2, th_vfwsub_wv_h, WOP_WUUU_H, H4, H2, H2, vfwsubw16) +THCALL(TH_OPFVV2, th_vfwsub_wv_w, WOP_WUUU_W, H8, H4, H4, vfwsubw32) +GEN_TH_VV_ENV(th_vfwsub_wv_h, 2, 4, clearl_th) +GEN_TH_VV_ENV(th_vfwsub_wv_w, 4, 8, clearq_th) +THCALL(TH_OPFVF2, th_vfwsub_wf_h, WOP_WUUU_H, H4, H2, vfwsubw16) +THCALL(TH_OPFVF2, th_vfwsub_wf_w, WOP_WUUU_W, H8, H4, vfwsubw32) +GEN_TH_VF(th_vfwsub_wf_h, 2, 4, clearl_th) +GEN_TH_VF(th_vfwsub_wf_w, 4, 8, clearq_th)