From nobody Fri Dec 19 18:52:47 2025 Delivered-To: importer@patchew.org Received-SPF: pass (zoho.com: domain of gnu.org designates 208.118.235.17 as permitted sender) client-ip=208.118.235.17; envelope-from=qemu-devel-bounces+importer=patchew.org@nongnu.org; helo=lists.gnu.org; Authentication-Results: mx.zohomail.com; dkim=fail; spf=pass (zoho.com: domain of gnu.org designates 208.118.235.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=fail(p=none dis=none) header.from=linaro.org Return-Path: Received: from lists.gnu.org (lists.gnu.org [208.118.235.17]) by mx.zohomail.com with SMTPS id 1518893181733135.5760606031438; Sat, 17 Feb 2018 10:46:21 -0800 (PST) Received: from localhost ([::1]:48236 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1en7V6-00022D-OZ for importer@patchew.org; Sat, 17 Feb 2018 13:46:20 -0500 Received: from eggs.gnu.org ([2001:4830:134:3::10]:39935) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1en79U-0000cC-KU for qemu-devel@nongnu.org; Sat, 17 Feb 2018 13:24:04 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1en79T-0001gs-3T for qemu-devel@nongnu.org; Sat, 17 Feb 2018 13:24:00 -0500 Received: from mail-pg0-x243.google.com ([2607:f8b0:400e:c05::243]:44352) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1en79S-0001gM-Rz for qemu-devel@nongnu.org; Sat, 17 Feb 2018 13:23:59 -0500 Received: by mail-pg0-x243.google.com with SMTP id l4so2191244pgp.11 for ; Sat, 17 Feb 2018 10:23:58 -0800 (PST) Received: from cloudburst.twiddle.net ([50.0.192.64]) by smtp.gmail.com with ESMTPSA id h15sm13466712pfi.56.2018.02.17.10.23.56 (version=TLS1_2 cipher=ECDHE-RSA-CHACHA20-POLY1305 bits=256/256); Sat, 17 Feb 2018 10:23:56 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=flMxh8Nvg2soBYq5gB6IVe/v9x76mKBYlcXjLqsvSu8=; b=SmON7BiB1Oa4kH7rcwQy2TmkZau4ONwPUQS/n8dEP1UstUk+wmYz9R70otlU1K6/xo tfnTG63eQREzoSNSGuffcw2zweJENbZbXzw5j0k4XrEb26pZS+OVOCUGfVMMgOv1WMci 1WI+w7gSQIROucxLE2y77dhKez/w1ecelyiZM= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=flMxh8Nvg2soBYq5gB6IVe/v9x76mKBYlcXjLqsvSu8=; b=HSBdK0On+ahWE+BfOewzRRvP0fV36BL8qKhlDmiIN+JdUNlsBWLKRBchxqxnZRa53v bPL2VXY1D6ibBdZFwsWWpi4tWWwqRF8gyIjWveyu8Veo78qoGjHz2Api1r5BoSjGHxSx cSLaiWkn6Dkfsdi1pBLVFSOl0KkpWn4phTJDllC7+9+fpbbZkSFSiv99XRmW97vqoBOL o/S5IfkMmnM9s+/+5sFN4oj9oj/mz69bgEoBRmPtoYDG/Y3dizGLr3UL42Xu39U7jMz4 tYUdJ10fUODIwnEeQGW6Y+rA3z1qaad5TI2tpL68m68NEImQXB5pe+rt4I4GfxLUmmpd E/9A== X-Gm-Message-State: APf1xPAo1H5FBvkGuf/AcRMX4vG+US7I3SFFKu3HwSMkzrvmjoFIpzC0 lt9hiuI7goNWz1bu+9JYB+LJSi/4H3Q= X-Google-Smtp-Source: AH8x2274+Azxc84h6CqkBvw5gwwCIMz+vfQXtIvmz4g4WXh1TPnF/N2Wb2Ii/6Tt9AcVOSeKxZOMWg== X-Received: by 10.98.38.134 with SMTP id m128mr9653690pfm.154.1518891837624; Sat, 17 Feb 2018 10:23:57 -0800 (PST) From: Richard Henderson To: qemu-devel@nongnu.org Date: Sat, 17 Feb 2018 10:22:35 -0800 Message-Id: <20180217182323.25885-20-richard.henderson@linaro.org> X-Mailer: git-send-email 2.14.3 In-Reply-To: <20180217182323.25885-1-richard.henderson@linaro.org> References: <20180217182323.25885-1-richard.henderson@linaro.org> X-detected-operating-system: by eggs.gnu.org: Genre and OS details not recognized. X-Received-From: 2607:f8b0:400e:c05::243 Subject: [Qemu-devel] [PATCH v2 19/67] target/arm: Implement SVE Bitwise Shift - Unpredicated Group X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.21 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: qemu-arm@nongnu.org Errors-To: qemu-devel-bounces+importer=patchew.org@nongnu.org Sender: "Qemu-devel" X-ZohoMail-DKIM: fail (Header signature does not verify) X-ZohoMail: RDKM_2 RSF_0 Z_629925259 SPT_0 Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell --- target/arm/helper-sve.h | 12 +++++++ target/arm/sve_helper.c | 30 +++++++++++++++++ target/arm/translate-sve.c | 81 ++++++++++++++++++++++++++++++++++++++++++= ++++ target/arm/sve.decode | 26 +++++++++++++++ 4 files changed, 149 insertions(+) diff --git a/target/arm/helper-sve.h b/target/arm/helper-sve.h index 2a2dbe98dd..00e3cd48bb 100644 --- a/target/arm/helper-sve.h +++ b/target/arm/helper-sve.h @@ -368,6 +368,18 @@ DEF_HELPER_FLAGS_4(sve_index_h, TCG_CALL_NO_RWG, void,= ptr, i32, i32, i32) DEF_HELPER_FLAGS_4(sve_index_s, TCG_CALL_NO_RWG, void, ptr, i32, i32, i32) DEF_HELPER_FLAGS_4(sve_index_d, TCG_CALL_NO_RWG, void, ptr, i64, i64, i32) =20 +DEF_HELPER_FLAGS_4(sve_asr_zzw_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i3= 2) +DEF_HELPER_FLAGS_4(sve_asr_zzw_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i3= 2) +DEF_HELPER_FLAGS_4(sve_asr_zzw_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i3= 2) + +DEF_HELPER_FLAGS_4(sve_lsr_zzw_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i3= 2) +DEF_HELPER_FLAGS_4(sve_lsr_zzw_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i3= 2) +DEF_HELPER_FLAGS_4(sve_lsr_zzw_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i3= 2) + +DEF_HELPER_FLAGS_4(sve_lsl_zzw_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i3= 2) +DEF_HELPER_FLAGS_4(sve_lsl_zzw_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i3= 2) +DEF_HELPER_FLAGS_4(sve_lsl_zzw_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i3= 2) + DEF_HELPER_FLAGS_5(sve_and_pppp, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr= , i32) DEF_HELPER_FLAGS_5(sve_bic_pppp, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr= , i32) DEF_HELPER_FLAGS_5(sve_eor_pppp, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr= , i32) diff --git a/target/arm/sve_helper.c b/target/arm/sve_helper.c index 950012e70a..4c6e2713fa 100644 --- a/target/arm/sve_helper.c +++ b/target/arm/sve_helper.c @@ -614,6 +614,36 @@ DO_ZPZ(sve_neg_h, uint16_t, H1_2, DO_NEG) DO_ZPZ(sve_neg_s, uint32_t, H1_4, DO_NEG) DO_ZPZ_D(sve_neg_d, uint64_t, DO_NEG) =20 +/* Three-operand expander, unpredicated, in which the third operand is "wi= de". + */ +#define DO_ZZW(NAME, TYPE, TYPEW, H, OP) \ +void HELPER(NAME)(void *vd, void *vn, void *vm, uint32_t desc) \ +{ \ + intptr_t i, opr_sz =3D simd_oprsz(desc); \ + for (i =3D 0; i < opr_sz; ) { \ + TYPEW mm =3D *(TYPEW *)(vm + i); \ + do { \ + TYPE nn =3D *(TYPE *)(vn + H(i)); \ + *(TYPE *)(vd + H(i)) =3D OP(nn, mm); \ + i +=3D sizeof(TYPE); \ + } while (i & 7); \ + } \ +} + +DO_ZZW(sve_asr_zzw_b, int8_t, uint64_t, H1, DO_ASR) +DO_ZZW(sve_lsr_zzw_b, uint8_t, uint64_t, H1, DO_LSR) +DO_ZZW(sve_lsl_zzw_b, uint8_t, uint64_t, H1, DO_LSL) + +DO_ZZW(sve_asr_zzw_h, int16_t, uint64_t, H1_2, DO_ASR) +DO_ZZW(sve_lsr_zzw_h, uint16_t, uint64_t, H1_2, DO_LSR) +DO_ZZW(sve_lsl_zzw_h, uint16_t, uint64_t, H1_2, DO_LSL) + +DO_ZZW(sve_asr_zzw_s, int32_t, uint64_t, H1_4, DO_ASR) +DO_ZZW(sve_lsr_zzw_s, uint32_t, uint64_t, H1_4, DO_LSR) +DO_ZZW(sve_lsl_zzw_s, uint32_t, uint64_t, H1_4, DO_LSL) + +#undef DO_ZZW + #undef DO_CLS_B #undef DO_CLS_H #undef DO_CLZ_B diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index 4a38020c8a..43e9f1ad08 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -130,6 +130,13 @@ static void do_mov_z(DisasContext *s, int rd, int rn) do_vector2_z(s, tcg_gen_gvec_mov, 0, rd, rn); } =20 +/* Initialize a Zreg with replications of a 64-bit immediate. */ +static void do_dupi_z(DisasContext *s, int rd, uint64_t word) +{ + unsigned vsz =3D vec_full_reg_size(s); + tcg_gen_gvec_dup64i(vec_full_reg_offset(s, rd), vsz, vsz, word); +} + /* Invoke a vector expander on two Pregs. */ static void do_vector2_p(DisasContext *s, GVecGen2Fn *gvec_fn, int esz, int rd, int rn) @@ -644,6 +651,80 @@ DO_ZPZW(LSL, lsl) =20 #undef DO_ZPZW =20 +/* + *** SVE Bitwise Shift - Unpredicated Group + */ + +static void do_shift_imm(DisasContext *s, arg_rri_esz *a, bool asr, + void (*gvec_fn)(unsigned, uint32_t, uint32_t, + int64_t, uint32_t, uint32_t)) +{ + unsigned vsz =3D vec_full_reg_size(s); + if (a->esz < 0) { + /* Invalid tsz encoding -- see tszimm_esz. */ + unallocated_encoding(s); + return; + } + /* Shift by element size is architecturally valid. For + arithmetic right-shift, it's the same as by one less. + Otherwise it is a zeroing operation. */ + if (a->imm >=3D 8 << a->esz) { + if (asr) { + a->imm =3D (8 << a->esz) - 1; + } else { + do_dupi_z(s, a->rd, 0); + return; + } + } + gvec_fn(a->esz, vec_full_reg_offset(s, a->rd), + vec_full_reg_offset(s, a->rn), a->imm, vsz, vsz); +} + +static void trans_ASR_zzi(DisasContext *s, arg_rri_esz *a, uint32_t insn) +{ + do_shift_imm(s, a, true, tcg_gen_gvec_sari); +} + +static void trans_LSR_zzi(DisasContext *s, arg_rri_esz *a, uint32_t insn) +{ + do_shift_imm(s, a, false, tcg_gen_gvec_shri); +} + +static void trans_LSL_zzi(DisasContext *s, arg_rri_esz *a, uint32_t insn) +{ + do_shift_imm(s, a, false, tcg_gen_gvec_shli); +} + +static void do_zzw_ool(DisasContext *s, arg_rrr_esz *a, gen_helper_gvec_3 = *fn) +{ + unsigned vsz =3D vec_full_reg_size(s); + if (fn =3D=3D NULL) { + unallocated_encoding(s); + return; + } + tcg_gen_gvec_3_ool(vec_full_reg_offset(s, a->rd), + vec_full_reg_offset(s, a->rn), + vec_full_reg_offset(s, a->rm), + vsz, vsz, 0, fn); +} + +#define DO_ZZW(NAME, name) \ +static void trans_##NAME##_zzw(DisasContext *s, arg_rrr_esz *a, \ + uint32_t insn) \ +{ \ + static gen_helper_gvec_3 * const fns[4] =3D { = \ + gen_helper_sve_##name##_zzw_b, gen_helper_sve_##name##_zzw_h, \ + gen_helper_sve_##name##_zzw_s, NULL \ + }; \ + do_zzw_ool(s, a, fns[a->esz]); \ +} + +DO_ZZW(ASR, asr) +DO_ZZW(LSR, lsr) +DO_ZZW(LSL, lsl) + +#undef DO_ZZW + /* *** SVE Integer Multiply-Add Group */ diff --git a/target/arm/sve.decode b/target/arm/sve.decode index 0b47869dcd..f71ea1b60d 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -33,6 +33,11 @@ # A combination of tsz:imm3 -- extract (tsz:imm3) - esize %tszimm_shl 22:2 5:5 !function=3Dtszimm_shl =20 +# Similarly for the tszh/tszl pair at 22/16 for zzi +%tszimm16_esz 22:2 16:5 !function=3Dtszimm_esz +%tszimm16_shr 22:2 16:5 !function=3Dtszimm_shr +%tszimm16_shl 22:2 16:5 !function=3Dtszimm_shl + # Either a copy of rd (at bit 0), or a different source # as propagated via the MOVPRFX instruction. %reg_movprfx 0:5 @@ -44,6 +49,7 @@ =20 &rr_esz rd rn esz &rri rd rn imm +&rri_esz rd rn imm esz &rrr_esz rd rn rm esz &rpr_esz rd pg rn esz &rprr_s rd pg rn rm s @@ -94,6 +100,10 @@ @rdn_pg_tszimm ........ .. ... ... ... pg:3 ..... rd:5 \ &rpri_esz rn=3D%reg_movprfx esz=3D%tszimm_esz =20 +# Similarly without predicate. +@rd_rn_tszimm ........ .. ... ... ...... rn:5 rd:5 \ + &rri_esz esz=3D%tszimm16_esz + # Basic Load/Store with 9-bit immediate offset @pd_rn_i9 ........ ........ ...... rn:5 . rd:4 \ &rri imm=3D%imm9_16_10 @@ -252,6 +262,22 @@ ADDPL 00000100 011 ..... 01010 ...... ..... @rd_rn_i6 # SVE stack frame size RDVL 00000100 101 11111 01010 imm:s6 rd:5 =20 +### SVE Bitwise Shift - Unpredicated Group + +# SVE bitwise shift by immediate (unpredicated) +ASR_zzi 00000100 .. 1 ..... 1001 00 ..... ..... \ + @rd_rn_tszimm imm=3D%tszimm16_shr +LSR_zzi 00000100 .. 1 ..... 1001 01 ..... ..... \ + @rd_rn_tszimm imm=3D%tszimm16_shr +LSL_zzi 00000100 .. 1 ..... 1001 11 ..... ..... \ + @rd_rn_tszimm imm=3D%tszimm16_shl + +# SVE bitwise shift by wide elements (unpredicated) +# Note esz !=3D 3 +ASR_zzw 00000100 .. 1 ..... 1000 00 ..... ..... @rd_rn_rm +LSR_zzw 00000100 .. 1 ..... 1000 01 ..... ..... @rd_rn_rm +LSL_zzw 00000100 .. 1 ..... 1000 11 ..... ..... @rd_rn_rm + ### SVE Predicate Logical Operations Group =20 # SVE predicate logical operations --=20 2.14.3