From nobody Fri Dec 19 04:28:21 2025 Delivered-To: importer@patchew.org Authentication-Results: mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass(p=none dis=none) header.from=linaro.org ARC-Seal: i=1; a=rsa-sha256; t=1589467011; cv=none; d=zohomail.com; s=zohoarc; b=lda3HGmHREv0eB5YzXVfkpdTyx92am2ILYvvXvpRpLnmIZtfU6ciM0LaAACNTu/LrbY1QXOpPzQ92lOokElbrmW+TZWm2Bzp7fXBAt/pWijp7jMOhmUApJjBq9mdU5TrBeegPmyaBTKjNvoAbpqUrx/scVge15fpKGx86VjPc98= ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=zohomail.com; s=zohoarc; t=1589467011; h=Content-Transfer-Encoding:Date:From:In-Reply-To:List-Subscribe:List-Post:List-Id:List-Archive:List-Help:List-Unsubscribe:MIME-Version:Message-ID:References:Sender:Subject:To; bh=MdcTI/rl5C0T34P66Bf2339mKZikYaOI25Vw6yFmsUU=; b=JrcZ6iRO76B4OjUQEXHt3hHeoHCvsWhkquqC8wht/LakQYUMfSKTK576jq/9zno09lBlsSL55t2FcwyR1E7kbtKYTcHq/HMT7lxwKtN5wNwI9wy7QRbYkK7zTOpKzvzq/y/GMJndu7ebkjmPfILGYwL+hCz4QAUKTHvvsE6OBtQ= ARC-Authentication-Results: i=1; mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass header.from= (p=none dis=none) header.from= Return-Path: Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) by mx.zohomail.com with SMTPS id 1589467011219194.39663761156908; Thu, 14 May 2020 07:36:51 -0700 (PDT) Received: from localhost ([::1]:44426 helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1jZEyg-0006Za-87 for importer@patchew.org; Thu, 14 May 2020 10:36:50 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]:39582) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1jZEkU-0003UX-3S for qemu-devel@nongnu.org; Thu, 14 May 2020 10:22:11 -0400 Received: from mail-wm1-x32c.google.com ([2a00:1450:4864:20::32c]:55895) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1jZEkP-0005j0-87 for qemu-devel@nongnu.org; Thu, 14 May 2020 10:22:09 -0400 Received: by mail-wm1-x32c.google.com with SMTP id f13so2231197wmc.5 for ; Thu, 14 May 2020 07:22:04 -0700 (PDT) Received: from orth.archaic.org.uk (orth.archaic.org.uk. [81.2.115.148]) by smtp.gmail.com with ESMTPSA id 88sm4077443wrq.77.2020.05.14.07.22.01 for (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Thu, 14 May 2020 07:22:02 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=from:to:subject:date:message-id:in-reply-to:references:mime-version :content-transfer-encoding; bh=MdcTI/rl5C0T34P66Bf2339mKZikYaOI25Vw6yFmsUU=; b=shm73zUrM5UwewE1y40RNlF0PYJVF0CCDx52kg4+TIeTC0XVrVcV3gnigKB0p0bdu/ 1ODB4nIxHPfXIaIgwiZzZSMk7h0ZZ+D9emSZGDJp5rZbQhIHwvSE7VANTFCvNZEIn8hF hWop6q1AFt32pcVnvtWD1oWwHtfN49GW7HkcmzE+k84b+Q7pu/rCQSPdXnk3IlBe7ugb i1vHYUIjEUKQzfHbqhv2tp/lAINrZ9LmohZqYbkulzS1fjgYu9niVMgas14bZvP8HOwV ks8pHOjbQAAyl0O8EvUz6i9+szLycptNH9g7zOZi+pVPsDVRWWaUA0xHKnmVTzNQemOp 3VDA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=MdcTI/rl5C0T34P66Bf2339mKZikYaOI25Vw6yFmsUU=; b=uYWkjzKGFJig+URYTuQjSHV3hECYa1NGEHsQ+mSJPALPzFXDl0sdPSwWD17ijmZN+V tcikuaeYaiN0uUuf7Ags3n4M7gkuyOZjELBJWwIDnvlN8hbbdV3a09wO9Ya3i0FXQvqS HusX9Daa2g60W87LC803aF/r9oV9hubIHKM5x6UVgyUOIccASmdL0oBxk4t8/nDdRHxk +og/S9wOF7XMv/1UVjboUjFyG3DPV/UNC41pElvdr/NqXa559FhQ0SQ8eHpysTUc1NJD BOY5TTniZkpDuXkK+dtdP+OkrxbOE39CJ48EVytEA5KlqC5UW1Npi043rlizcGjLOGnP 6zug== X-Gm-Message-State: AGi0PuYZ09H3gHNQxWW9SIACf6YUn8OMEpZ7PGH9LYrXp8t6CONZ++JE HQ22AwPF5oqijaM/UxPw60+sjm6Cu+LTdA== X-Google-Smtp-Source: APiQypI2DlPuv/zjsyYmSweEXvf+fDBroFEBhdki8b9CYusM/Eylv5KA0+2OrhrxD1WaR91DTrjV/Q== X-Received: by 2002:a7b:c3d2:: with SMTP id t18mr40757362wmj.100.1589466123326; Thu, 14 May 2020 07:22:03 -0700 (PDT) From: Peter Maydell To: qemu-devel@nongnu.org Subject: [PULL 17/45] target/arm: Vectorize SABA/UABA Date: Thu, 14 May 2020 15:21:10 +0100 Message-Id: <20200514142138.20875-18-peter.maydell@linaro.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20200514142138.20875-1-peter.maydell@linaro.org> References: <20200514142138.20875-1-peter.maydell@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Received-SPF: pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) client-ip=209.51.188.17; envelope-from=qemu-devel-bounces+importer=patchew.org@nongnu.org; helo=lists.gnu.org; Received-SPF: pass client-ip=2a00:1450:4864:20::32c; envelope-from=peter.maydell@linaro.org; helo=mail-wm1-x32c.google.com X-detected-operating-system: by eggs.gnu.org: No matching host in p0f cache. That's all we know. X-Spam_score_int: -20 X-Spam_score: -2.1 X-Spam_bar: -- X-Spam_report: (-2.1 / 5.0 requ) BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_NONE=-0.0001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001 autolearn=_AUTOLEARN X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+importer=patchew.org@nongnu.org Sender: "Qemu-devel" X-ZohoMail-DKIM: pass (identity @linaro.org) Content-Type: text/plain; charset="utf-8" From: Richard Henderson Include 64-bit element size in preparation for SVE2. Reviewed-by: Peter Maydell Signed-off-by: Richard Henderson Message-id: 20200513163245.17915-17-richard.henderson@linaro.org Signed-off-by: Peter Maydell --- target/arm/helper.h | 17 +++-- target/arm/translate.h | 5 ++ target/arm/neon_helper.c | 10 --- target/arm/translate-a64.c | 17 ++--- target/arm/translate.c | 134 +++++++++++++++++++++++++++++++++++-- target/arm/vec_helper.c | 24 +++++++ 6 files changed, 174 insertions(+), 33 deletions(-) diff --git a/target/arm/helper.h b/target/arm/helper.h index 4678d3a6f4e..1857f4ee46a 100644 --- a/target/arm/helper.h +++ b/target/arm/helper.h @@ -284,13 +284,6 @@ DEF_HELPER_2(neon_pmax_s8, i32, i32, i32) DEF_HELPER_2(neon_pmax_u16, i32, i32, i32) DEF_HELPER_2(neon_pmax_s16, i32, i32, i32) =20 -DEF_HELPER_2(neon_abd_u8, i32, i32, i32) -DEF_HELPER_2(neon_abd_s8, i32, i32, i32) -DEF_HELPER_2(neon_abd_u16, i32, i32, i32) -DEF_HELPER_2(neon_abd_s16, i32, i32, i32) -DEF_HELPER_2(neon_abd_u32, i32, i32, i32) -DEF_HELPER_2(neon_abd_s32, i32, i32, i32) - DEF_HELPER_2(neon_shl_u16, i32, i32, i32) DEF_HELPER_2(neon_shl_s16, i32, i32, i32) DEF_HELPER_2(neon_rshl_u8, i32, i32, i32) @@ -741,6 +734,16 @@ DEF_HELPER_FLAGS_4(gvec_uabd_h, TCG_CALL_NO_RWG, void,= ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(gvec_uabd_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(gvec_uabd_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) =20 +DEF_HELPER_FLAGS_4(gvec_saba_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(gvec_saba_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(gvec_saba_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(gvec_saba_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) + +DEF_HELPER_FLAGS_4(gvec_uaba_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(gvec_uaba_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(gvec_uaba_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(gvec_uaba_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) + #ifdef TARGET_AARCH64 #include "helper-a64.h" #include "helper-sve.h" diff --git a/target/arm/translate.h b/target/arm/translate.h index bbfe3d13930..c937dfe9bf0 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -342,6 +342,11 @@ void gen_gvec_sabd(unsigned vece, uint32_t rd_ofs, uin= t32_t rn_ofs, void gen_gvec_uabd(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz); =20 +void gen_gvec_saba(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, + uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz); +void gen_gvec_uaba(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, + uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz); + /* * Forward to the isar_feature_* tests given a DisasContext pointer. */ diff --git a/target/arm/neon_helper.c b/target/arm/neon_helper.c index 448be93fa12..2ef75e04c83 100644 --- a/target/arm/neon_helper.c +++ b/target/arm/neon_helper.c @@ -576,16 +576,6 @@ NEON_POP(pmax_s16, neon_s16, 2) NEON_POP(pmax_u16, neon_u16, 2) #undef NEON_FN =20 -#define NEON_FN(dest, src1, src2) \ - dest =3D (src1 > src2) ? (src1 - src2) : (src2 - src1) -NEON_VOP(abd_s8, neon_s8, 4) -NEON_VOP(abd_u8, neon_u8, 4) -NEON_VOP(abd_s16, neon_s16, 2) -NEON_VOP(abd_u16, neon_u16, 2) -NEON_VOP(abd_s32, neon_s32, 1) -NEON_VOP(abd_u32, neon_u32, 1) -#undef NEON_FN - #define NEON_FN(dest, src1, src2) do { \ int8_t tmp; \ tmp =3D (int8_t)src2; \ diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index 54b06553a65..991e451644c 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -11197,6 +11197,13 @@ static void disas_simd_3same_int(DisasContext *s, = uint32_t insn) gen_gvec_fn3(s, is_q, rd, rn, rm, gen_gvec_sabd, size); } return; + case 0xf: /* SABA, UABA */ + if (u) { + gen_gvec_fn3(s, is_q, rd, rn, rm, gen_gvec_uaba, size); + } else { + gen_gvec_fn3(s, is_q, rd, rn, rm, gen_gvec_saba, size); + } + return; case 0x10: /* ADD, SUB */ if (u) { gen_gvec_fn3(s, is_q, rd, rn, rm, tcg_gen_gvec_sub, size); @@ -11329,16 +11336,6 @@ static void disas_simd_3same_int(DisasContext *s, = uint32_t insn) genenvfn =3D fns[size][u]; break; } - case 0xf: /* SABA, UABA */ - { - static NeonGenTwoOpFn * const fns[3][2] =3D { - { gen_helper_neon_abd_s8, gen_helper_neon_abd_u8 }, - { gen_helper_neon_abd_s16, gen_helper_neon_abd_u16 }, - { gen_helper_neon_abd_s32, gen_helper_neon_abd_u32 }, - }; - genfn =3D fns[size][u]; - break; - } case 0x16: /* SQDMULH, SQRDMULH */ { static NeonGenTwoOpEnvFn * const fns[2][2] =3D { diff --git a/target/arm/translate.c b/target/arm/translate.c index d288721c230..e3d37ef2e99 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -5222,6 +5222,124 @@ void gen_gvec_uabd(unsigned vece, uint32_t rd_ofs, = uint32_t rn_ofs, tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, opr_sz, max_sz, &ops[vece]); } =20 +static void gen_saba_i32(TCGv_i32 d, TCGv_i32 a, TCGv_i32 b) +{ + TCGv_i32 t =3D tcg_temp_new_i32(); + gen_sabd_i32(t, a, b); + tcg_gen_add_i32(d, d, t); + tcg_temp_free_i32(t); +} + +static void gen_saba_i64(TCGv_i64 d, TCGv_i64 a, TCGv_i64 b) +{ + TCGv_i64 t =3D tcg_temp_new_i64(); + gen_sabd_i64(t, a, b); + tcg_gen_add_i64(d, d, t); + tcg_temp_free_i64(t); +} + +static void gen_saba_vec(unsigned vece, TCGv_vec d, TCGv_vec a, TCGv_vec b) +{ + TCGv_vec t =3D tcg_temp_new_vec_matching(d); + gen_sabd_vec(vece, t, a, b); + tcg_gen_add_vec(vece, d, d, t); + tcg_temp_free_vec(t); +} + +void gen_gvec_saba(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, + uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz) +{ + static const TCGOpcode vecop_list[] =3D { + INDEX_op_sub_vec, INDEX_op_add_vec, + INDEX_op_smin_vec, INDEX_op_smax_vec, 0 + }; + static const GVecGen3 ops[4] =3D { + { .fniv =3D gen_saba_vec, + .fno =3D gen_helper_gvec_saba_b, + .opt_opc =3D vecop_list, + .load_dest =3D true, + .vece =3D MO_8 }, + { .fniv =3D gen_saba_vec, + .fno =3D gen_helper_gvec_saba_h, + .opt_opc =3D vecop_list, + .load_dest =3D true, + .vece =3D MO_16 }, + { .fni4 =3D gen_saba_i32, + .fniv =3D gen_saba_vec, + .fno =3D gen_helper_gvec_saba_s, + .opt_opc =3D vecop_list, + .load_dest =3D true, + .vece =3D MO_32 }, + { .fni8 =3D gen_saba_i64, + .fniv =3D gen_saba_vec, + .fno =3D gen_helper_gvec_saba_d, + .prefer_i64 =3D TCG_TARGET_REG_BITS =3D=3D 64, + .opt_opc =3D vecop_list, + .load_dest =3D true, + .vece =3D MO_64 }, + }; + tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, opr_sz, max_sz, &ops[vece]); +} + +static void gen_uaba_i32(TCGv_i32 d, TCGv_i32 a, TCGv_i32 b) +{ + TCGv_i32 t =3D tcg_temp_new_i32(); + gen_uabd_i32(t, a, b); + tcg_gen_add_i32(d, d, t); + tcg_temp_free_i32(t); +} + +static void gen_uaba_i64(TCGv_i64 d, TCGv_i64 a, TCGv_i64 b) +{ + TCGv_i64 t =3D tcg_temp_new_i64(); + gen_uabd_i64(t, a, b); + tcg_gen_add_i64(d, d, t); + tcg_temp_free_i64(t); +} + +static void gen_uaba_vec(unsigned vece, TCGv_vec d, TCGv_vec a, TCGv_vec b) +{ + TCGv_vec t =3D tcg_temp_new_vec_matching(d); + gen_uabd_vec(vece, t, a, b); + tcg_gen_add_vec(vece, d, d, t); + tcg_temp_free_vec(t); +} + +void gen_gvec_uaba(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, + uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz) +{ + static const TCGOpcode vecop_list[] =3D { + INDEX_op_sub_vec, INDEX_op_add_vec, + INDEX_op_umin_vec, INDEX_op_umax_vec, 0 + }; + static const GVecGen3 ops[4] =3D { + { .fniv =3D gen_uaba_vec, + .fno =3D gen_helper_gvec_uaba_b, + .opt_opc =3D vecop_list, + .load_dest =3D true, + .vece =3D MO_8 }, + { .fniv =3D gen_uaba_vec, + .fno =3D gen_helper_gvec_uaba_h, + .opt_opc =3D vecop_list, + .load_dest =3D true, + .vece =3D MO_16 }, + { .fni4 =3D gen_uaba_i32, + .fniv =3D gen_uaba_vec, + .fno =3D gen_helper_gvec_uaba_s, + .opt_opc =3D vecop_list, + .load_dest =3D true, + .vece =3D MO_32 }, + { .fni8 =3D gen_uaba_i64, + .fniv =3D gen_uaba_vec, + .fno =3D gen_helper_gvec_uaba_d, + .prefer_i64 =3D TCG_TARGET_REG_BITS =3D=3D 64, + .opt_opc =3D vecop_list, + .load_dest =3D true, + .vece =3D MO_64 }, + }; + tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, opr_sz, max_sz, &ops[vece]); +} + /* Translate a NEON data processing instruction. Return nonzero if the instruction is invalid. We process data in a mixture of 32-bit and 64-bit chunks. @@ -5366,6 +5484,16 @@ static int disas_neon_data_insn(DisasContext *s, uin= t32_t insn) } return 0; =20 + case NEON_3R_VABA: + if (u) { + gen_gvec_uaba(size, rd_ofs, rn_ofs, rm_ofs, + vec_size, vec_size); + } else { + gen_gvec_saba(size, rd_ofs, rn_ofs, rm_ofs, + vec_size, vec_size); + } + return 0; + case NEON_3R_VADD_VSUB: case NEON_3R_LOGIC: case NEON_3R_VMAX: @@ -5510,12 +5638,6 @@ static int disas_neon_data_insn(DisasContext *s, uin= t32_t insn) case NEON_3R_VQRSHL: GEN_NEON_INTEGER_OP_ENV(qrshl); break; - case NEON_3R_VABA: - GEN_NEON_INTEGER_OP(abd); - tcg_temp_free_i32(tmp2); - tmp2 =3D neon_load_reg(rd, pass); - gen_neon_add(size, tmp, tmp2); - break; case NEON_3R_VPMAX: GEN_NEON_INTEGER_OP(pmax); break; diff --git a/target/arm/vec_helper.c b/target/arm/vec_helper.c index a4972d02fc9..fa33df859e5 100644 --- a/target/arm/vec_helper.c +++ b/target/arm/vec_helper.c @@ -1431,3 +1431,27 @@ DO_ABD(gvec_uabd_s, uint32_t) DO_ABD(gvec_uabd_d, uint64_t) =20 #undef DO_ABD + +#define DO_ABA(NAME, TYPE) \ +void HELPER(NAME)(void *vd, void *vn, void *vm, uint32_t desc) \ +{ \ + intptr_t i, opr_sz =3D simd_oprsz(desc); \ + TYPE *d =3D vd, *n =3D vn, *m =3D vm; \ + \ + for (i =3D 0; i < opr_sz / sizeof(TYPE); ++i) { \ + d[i] +=3D n[i] < m[i] ? m[i] - n[i] : n[i] - m[i]; \ + } \ + clear_tail(d, opr_sz, simd_maxsz(desc)); \ +} + +DO_ABA(gvec_saba_b, int8_t) +DO_ABA(gvec_saba_h, int16_t) +DO_ABA(gvec_saba_s, int32_t) +DO_ABA(gvec_saba_d, int64_t) + +DO_ABA(gvec_uaba_b, uint8_t) +DO_ABA(gvec_uaba_h, uint16_t) +DO_ABA(gvec_uaba_s, uint32_t) +DO_ABA(gvec_uaba_d, uint64_t) + +#undef DO_ABA --=20 2.20.1