From nobody Mon Apr 7 15:55:56 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; 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 1526665468092508.5869600075637; Fri, 18 May 2018 10:44:28 -0700 (PDT) Received: from localhost ([::1]:40223 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fJjQY-00051v-UO for importer@patchew.org; Fri, 18 May 2018 13:44:26 -0400 Received: from eggs.gnu.org ([2001:4830:134:3::10]:36256) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fJj3U-0000uT-F5 for qemu-devel@nongnu.org; Fri, 18 May 2018 13:20:37 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1fJj3T-0007h4-7r for qemu-devel@nongnu.org; Fri, 18 May 2018 13:20:36 -0400 Received: from orth.archaic.org.uk ([2001:8b0:1d0::2]:41802) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1fJj3S-0007fI-IY for qemu-devel@nongnu.org; Fri, 18 May 2018 13:20:35 -0400 Received: from pm215 by orth.archaic.org.uk with local (Exim 4.89) (envelope-from ) id 1fJj3R-0004xl-FK for qemu-devel@nongnu.org; Fri, 18 May 2018 18:20:33 +0100 From: Peter Maydell To: qemu-devel@nongnu.org Date: Fri, 18 May 2018 18:20:09 +0100 Message-Id: <20180518172009.14416-33-peter.maydell@linaro.org> X-Mailer: git-send-email 2.17.0 In-Reply-To: <20180518172009.14416-1-peter.maydell@linaro.org> References: <20180518172009.14416-1-peter.maydell@linaro.org> X-detected-operating-system: by eggs.gnu.org: Genre and OS details not recognized. X-Received-From: 2001:8b0:1d0::2 Subject: [Qemu-devel] [PULL 32/32] target/arm: Implement SVE Permute - Extract 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: , Errors-To: qemu-devel-bounces+importer=patchew.org@nongnu.org Sender: "Qemu-devel" X-ZohoMail: RSF_0 Z_629925259 SPT_0 Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" From: Richard Henderson Reviewed-by: Peter Maydell Signed-off-by: Richard Henderson Message-id: 20180516223007.10256-26-richard.henderson@linaro.org Signed-off-by: Peter Maydell --- target/arm/helper-sve.h | 2 + target/arm/sve_helper.c | 81 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sve.c | 34 ++++++++++++++++ target/arm/sve.decode | 7 ++++ 4 files changed, 124 insertions(+) diff --git a/target/arm/helper-sve.h b/target/arm/helper-sve.h index 79493ab647..94f4356ce9 100644 --- a/target/arm/helper-sve.h +++ b/target/arm/helper-sve.h @@ -414,6 +414,8 @@ DEF_HELPER_FLAGS_4(sve_cpy_z_h, TCG_CALL_NO_RWG, void, = ptr, ptr, i64, i32) DEF_HELPER_FLAGS_4(sve_cpy_z_s, TCG_CALL_NO_RWG, void, ptr, ptr, i64, i32) DEF_HELPER_FLAGS_4(sve_cpy_z_d, TCG_CALL_NO_RWG, void, ptr, ptr, i64, i32) =20 +DEF_HELPER_FLAGS_4(sve_ext, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) + 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 8c7ea989b1..b825e44cb5 100644 --- a/target/arm/sve_helper.c +++ b/target/arm/sve_helper.c @@ -1479,3 +1479,84 @@ void HELPER(sve_cpy_z_d)(void *vd, void *vg, uint64_= t val, uint32_t desc) d[i] =3D (pg[H1(i)] & 1 ? val : 0); } } + +/* Big-endian hosts need to frob the byte indicies. If the copy + * happens to be 8-byte aligned, then no frobbing necessary. + */ +static void swap_memmove(void *vd, void *vs, size_t n) +{ + uintptr_t d =3D (uintptr_t)vd; + uintptr_t s =3D (uintptr_t)vs; + uintptr_t o =3D (d | s | n) & 7; + size_t i; + +#ifndef HOST_WORDS_BIGENDIAN + o =3D 0; +#endif + switch (o) { + case 0: + memmove(vd, vs, n); + break; + + case 4: + if (d < s || d >=3D s + n) { + for (i =3D 0; i < n; i +=3D 4) { + *(uint32_t *)H1_4(d + i) =3D *(uint32_t *)H1_4(s + i); + } + } else { + for (i =3D n; i > 0; ) { + i -=3D 4; + *(uint32_t *)H1_4(d + i) =3D *(uint32_t *)H1_4(s + i); + } + } + break; + + case 2: + case 6: + if (d < s || d >=3D s + n) { + for (i =3D 0; i < n; i +=3D 2) { + *(uint16_t *)H1_2(d + i) =3D *(uint16_t *)H1_2(s + i); + } + } else { + for (i =3D n; i > 0; ) { + i -=3D 2; + *(uint16_t *)H1_2(d + i) =3D *(uint16_t *)H1_2(s + i); + } + } + break; + + default: + if (d < s || d >=3D s + n) { + for (i =3D 0; i < n; i++) { + *(uint8_t *)H1(d + i) =3D *(uint8_t *)H1(s + i); + } + } else { + for (i =3D n; i > 0; ) { + i -=3D 1; + *(uint8_t *)H1(d + i) =3D *(uint8_t *)H1(s + i); + } + } + break; + } +} + +void HELPER(sve_ext)(void *vd, void *vn, void *vm, uint32_t desc) +{ + intptr_t opr_sz =3D simd_oprsz(desc); + size_t n_ofs =3D simd_data(desc); + size_t n_siz =3D opr_sz - n_ofs; + + if (vd !=3D vm) { + swap_memmove(vd, vn + n_ofs, n_siz); + swap_memmove(vd + n_siz, vm, n_ofs); + } else if (vd !=3D vn) { + swap_memmove(vd + n_siz, vd, n_ofs); + swap_memmove(vd, vn + n_ofs, n_siz); + } else { + /* vd =3D=3D vn =3D=3D vm. Need temp space. */ + ARMVectorReg tmp; + swap_memmove(&tmp, vm, n_ofs); + swap_memmove(vd, vd + n_ofs, n_siz); + memcpy(vd + n_siz, &tmp, n_ofs); + } +} diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index 9bdd61ff84..c48d4b530a 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -1922,6 +1922,40 @@ static bool trans_CPY_z_i(DisasContext *s, arg_CPY_z= _i *a, uint32_t insn) return true; } =20 +/* + *** SVE Permute Extract Group + */ + +static bool trans_EXT(DisasContext *s, arg_EXT *a, uint32_t insn) +{ + if (!sve_access_check(s)) { + return true; + } + + unsigned vsz =3D vec_full_reg_size(s); + unsigned n_ofs =3D a->imm >=3D vsz ? 0 : a->imm; + unsigned n_siz =3D vsz - n_ofs; + unsigned d =3D vec_full_reg_offset(s, a->rd); + unsigned n =3D vec_full_reg_offset(s, a->rn); + unsigned m =3D vec_full_reg_offset(s, a->rm); + + /* Use host vector move insns if we have appropriate sizes + * and no unfortunate overlap. + */ + if (m !=3D d + && n_ofs =3D=3D size_for_gvec(n_ofs) + && n_siz =3D=3D size_for_gvec(n_siz) + && (d !=3D n || n_siz <=3D n_ofs)) { + tcg_gen_gvec_mov(0, d, n + n_ofs, n_siz, n_siz); + if (n_ofs !=3D 0) { + tcg_gen_gvec_mov(0, d + n_siz, m, n_ofs, n_ofs); + } + } else { + tcg_gen_gvec_3_ool(d, n, m, vsz, vsz, n_ofs, gen_helper_sve_ext); + } + return true; +} + /* *** SVE Memory - 32-bit Gather and Unsized Contiguous Group */ diff --git a/target/arm/sve.decode b/target/arm/sve.decode index 4ee7c78cda..4761d1921e 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -24,6 +24,7 @@ =20 %imm4_16_p1 16:4 !function=3Dplus1 %imm6_22_5 22:1 5:5 +%imm8_16_10 16:5 10:3 %imm9_16_10 16:s6 10:3 =20 # A combination of tsz:imm3 -- extract esize. @@ -362,6 +363,12 @@ FCPY 00000101 .. 01 .... 110 imm:8 ..... = @rdn_pg4 CPY_m_i 00000101 .. 01 .... 01 . ........ ..... @rdn_pg4 imm=3D%= sh8_i8s CPY_z_i 00000101 .. 01 .... 00 . ........ ..... @rdn_pg4 imm=3D%= sh8_i8s =20 +### SVE Permute - Extract Group + +# SVE extract vector (immediate offset) +EXT 00000101 001 ..... 000 ... rm:5 rd:5 \ + &rrri rn=3D%reg_movprfx imm=3D%imm8_16_10 + ### SVE Predicate Logical Operations Group =20 # SVE predicate logical operations --=20 2.17.0