From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from zen.linaroharston ([51.148.130.216]) by smtp.gmail.com with ESMTPSA id n3sm6289464wrr.50.2019.10.18.06.40.16 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 18 Oct 2019 06:40:16 -0700 (PDT) Received: from zen (localhost [127.0.0.1]) by zen.linaroharston (Postfix) with ESMTP id AB12A1FF87; Fri, 18 Oct 2019 14:40:15 +0100 (BST) References: <20191017044232.27601-1-richard.henderson@linaro.org> <20191017044232.27601-4-richard.henderson@linaro.org> <871rvap7gf.fsf@linaro.org> User-agent: mu4e 1.3.5; emacs 27.0.50 From: Alex =?utf-8?Q?Benn=C3=A9e?= To: qemu-devel@nongnu.org Cc: peter.maydell@linaro.org, qemu-arm@nongnu.org Subject: Re: [PATCH 3/4] target/arm: Convert PMULL.64 to gvec In-reply-to: <871rvap7gf.fsf@linaro.org> Date: Fri, 18 Oct 2019 14:40:15 +0100 Message-ID: <87zhhynpds.fsf@linaro.org> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable X-TUID: 43CnjtchtsJl Alex Benn=C3=A9e writes: > Richard Henderson writes: > >> The gvec form will be needed for implementing SVE2. > > Hmm I'm seeing a failure against: > > aarch32-all-v80/insn_VMULL__INC.risu.bin I take it back, after monkey patching cortex-a53 into qemu-arm it passes. Reviewed-by: Alex Benn=C3=A9e Tested-by: Alex Benn=C3=A9e > > From: > > https://fileserver.linaro.org/owncloud/index.php/s/hvEXM2eJ3uZVhlH > https://fileserver.linaro.org/owncloud/index.php/s/hvEXM2eJ3uZVhlH/down= load?path=3D%2F&files=3Daarch32-all-v80.tar.xz > > And some others. But this seems to be broken in master as well so I > don't know if this is a regression or because I have my -cpu wrong for > qemu-arm for something recorded on a cortex-a53 in aarch32. > >> >> Signed-off-by: Richard Henderson >> --- >> target/arm/helper.h | 4 +--- >> target/arm/neon_helper.c | 30 ------------------------------ >> target/arm/translate-a64.c | 28 +++------------------------- >> target/arm/translate.c | 16 ++-------------- >> target/arm/vec_helper.c | 33 +++++++++++++++++++++++++++++++++ >> 5 files changed, 39 insertions(+), 72 deletions(-) >> >> diff --git a/target/arm/helper.h b/target/arm/helper.h >> index 800446e537..d954399b7e 100644 >> --- a/target/arm/helper.h >> +++ b/target/arm/helper.h >> @@ -555,9 +555,6 @@ DEF_HELPER_FLAGS_3(crc32, TCG_CALL_NO_RWG_SE, i32, i= 32, i32, i32) >> DEF_HELPER_FLAGS_3(crc32c, TCG_CALL_NO_RWG_SE, i32, i32, i32, i32) >> DEF_HELPER_2(dc_zva, void, env, i64) >> >> -DEF_HELPER_FLAGS_2(neon_pmull_64_lo, TCG_CALL_NO_RWG_SE, i64, i64, i64) >> -DEF_HELPER_FLAGS_2(neon_pmull_64_hi, TCG_CALL_NO_RWG_SE, i64, i64, i64) >> - >> DEF_HELPER_FLAGS_5(gvec_qrdmlah_s16, TCG_CALL_NO_RWG, >> void, ptr, ptr, ptr, ptr, i32) >> DEF_HELPER_FLAGS_5(gvec_qrdmlsh_s16, TCG_CALL_NO_RWG, >> @@ -689,6 +686,7 @@ DEF_HELPER_FLAGS_4(gvec_ushl_b, TCG_CALL_NO_RWG, voi= d, ptr, ptr, ptr, i32) >> DEF_HELPER_FLAGS_4(gvec_ushl_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i= 32) >> >> DEF_HELPER_FLAGS_4(gvec_pmul_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i= 32) >> +DEF_HELPER_FLAGS_4(gvec_pmull_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, = i32) >> >> #ifdef TARGET_AARCH64 >> #include "helper-a64.h" >> diff --git a/target/arm/neon_helper.c b/target/arm/neon_helper.c >> index 9e7a9a1ac5..6a107da0e1 100644 >> --- a/target/arm/neon_helper.c >> +++ b/target/arm/neon_helper.c >> @@ -2152,33 +2152,3 @@ void HELPER(neon_zip16)(void *vd, void *vm) >> rm[0] =3D m0; >> rd[0] =3D d0; >> } >> - >> -/* Helper function for 64 bit polynomial multiply case: >> - * perform PolynomialMult(op1, op2) and return either the top or >> - * bottom half of the 128 bit result. >> - */ >> -uint64_t HELPER(neon_pmull_64_lo)(uint64_t op1, uint64_t op2) >> -{ >> - int bitnum; >> - uint64_t res =3D 0; >> - >> - for (bitnum =3D 0; bitnum < 64; bitnum++) { >> - if (op1 & (1ULL << bitnum)) { >> - res ^=3D op2 << bitnum; >> - } >> - } >> - return res; >> -} >> -uint64_t HELPER(neon_pmull_64_hi)(uint64_t op1, uint64_t op2) >> -{ >> - int bitnum; >> - uint64_t res =3D 0; >> - >> - /* bit 0 of op1 can't influence the high 64 bits at all */ >> - for (bitnum =3D 1; bitnum < 64; bitnum++) { >> - if (op1 & (1ULL << bitnum)) { >> - res ^=3D op2 >> (64 - bitnum); >> - } >> - } >> - return res; >> -} >> diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c >> index 04e25cfe06..12588d18df 100644 >> --- a/target/arm/translate-a64.c >> +++ b/target/arm/translate-a64.c >> @@ -10598,30 +10598,6 @@ static void handle_3rd_narrowing(DisasContext *= s, int is_q, int is_u, int size, >> clear_vec_high(s, is_q, rd); >> } >> >> -static void handle_pmull_64(DisasContext *s, int is_q, int rd, int rn, = int rm) >> -{ >> - /* PMULL of 64 x 64 -> 128 is an odd special case because it >> - * is the only three-reg-diff instruction which produces a >> - * 128-bit wide result from a single operation. However since >> - * it's possible to calculate the two halves more or less >> - * separately we just use two helper calls. >> - */ >> - TCGv_i64 tcg_op1 =3D tcg_temp_new_i64(); >> - TCGv_i64 tcg_op2 =3D tcg_temp_new_i64(); >> - TCGv_i64 tcg_res =3D tcg_temp_new_i64(); >> - >> - read_vec_element(s, tcg_op1, rn, is_q, MO_64); >> - read_vec_element(s, tcg_op2, rm, is_q, MO_64); >> - gen_helper_neon_pmull_64_lo(tcg_res, tcg_op1, tcg_op2); >> - write_vec_element(s, tcg_res, rd, 0, MO_64); >> - gen_helper_neon_pmull_64_hi(tcg_res, tcg_op1, tcg_op2); >> - write_vec_element(s, tcg_res, rd, 1, MO_64); >> - >> - tcg_temp_free_i64(tcg_op1); >> - tcg_temp_free_i64(tcg_op2); >> - tcg_temp_free_i64(tcg_res); >> -} >> - >> /* AdvSIMD three different >> * 31 30 29 28 24 23 22 21 20 16 15 12 11 10 9 5 4 = 0 >> * +---+---+---+-----------+------+---+------+--------+-----+------+---= ---+ >> @@ -10686,7 +10662,9 @@ static void disas_simd_three_reg_diff(DisasConte= xt *s, uint32_t insn) >> if (!fp_access_check(s)) { >> return; >> } >> - handle_pmull_64(s, is_q, rd, rn, rm); >> + /* The Q field specifies lo/hi half input for this insn. */ >> + gen_gvec_op3_ool(s, true, rd, rn, rm, is_q, >> + gen_helper_gvec_pmull_q); >> return; >> } >> goto is_widening; >> diff --git a/target/arm/translate.c b/target/arm/translate.c >> index b66a2f6b71..4e34249672 100644 >> --- a/target/arm/translate.c >> +++ b/target/arm/translate.c >> @@ -5877,23 +5877,11 @@ static int disas_neon_data_insn(DisasContext *s,= uint32_t insn) >> * outside the loop below as it only performs a single = pass. >> */ >> if (op =3D=3D 14 && size =3D=3D 2) { >> - TCGv_i64 tcg_rn, tcg_rm, tcg_rd; >> - >> if (!dc_isar_feature(aa32_pmull, s)) { >> return 1; >> } >> - tcg_rn =3D tcg_temp_new_i64(); >> - tcg_rm =3D tcg_temp_new_i64(); >> - tcg_rd =3D tcg_temp_new_i64(); >> - neon_load_reg64(tcg_rn, rn); >> - neon_load_reg64(tcg_rm, rm); >> - gen_helper_neon_pmull_64_lo(tcg_rd, tcg_rn, tcg_rm); >> - neon_store_reg64(tcg_rd, rd); >> - gen_helper_neon_pmull_64_hi(tcg_rd, tcg_rn, tcg_rm); >> - neon_store_reg64(tcg_rd, rd + 1); >> - tcg_temp_free_i64(tcg_rn); >> - tcg_temp_free_i64(tcg_rm); >> - tcg_temp_free_i64(tcg_rd); >> + tcg_gen_gvec_3_ool(rd_ofs, rn_ofs, rm_ofs, 16, 16, >> + 0, gen_helper_gvec_pmull_q); >> return 0; >> } >> >> diff --git a/target/arm/vec_helper.c b/target/arm/vec_helper.c >> index d401282c6f..5c1074374e 100644 >> --- a/target/arm/vec_helper.c >> +++ b/target/arm/vec_helper.c >> @@ -1164,3 +1164,36 @@ void HELPER(gvec_pmul_b)(void *vd, void *vn, void= *vm, uint32_t desc) >> } >> clear_tail(d, opr_sz, simd_maxsz(desc)); >> } >> + >> +/* >> + * 64x64->128 polynomial multiply. >> + * Because of the lanes are not accessed in strict columns, >> + * this probably cannot be turned into a generic helper. >> + */ >> +void HELPER(gvec_pmull_q)(void *vd, void *vn, void *vm, uint32_t desc) >> +{ >> + intptr_t i, j, opr_sz =3D simd_oprsz(desc); >> + intptr_t hi =3D simd_data(desc); >> + uint64_t *d =3D vd, *n =3D vn, *m =3D vm; >> + >> + for (i =3D 0; i < opr_sz / 8; i +=3D 2) { >> + uint64_t nn =3D n[i + hi]; >> + uint64_t mm =3D m[i + hi]; >> + uint64_t rhi =3D 0; >> + uint64_t rlo =3D 0; >> + >> + /* Bit 0 can only influence the low 64-bit result. */ >> + if (nn & 1) { >> + rlo =3D mm; >> + } >> + >> + for (j =3D 1; j < 64; ++j) { >> + uint64_t mask =3D -((nn >> j) & 1); >> + rlo ^=3D (mm << j) & mask; >> + rhi ^=3D (mm >> (64 - j)) & mask; >> + } >> + d[i] =3D rlo; >> + d[i + 1] =3D rhi; >> + } >> + clear_tail(d, opr_sz, simd_maxsz(desc)); >> +} -- Alex Benn=C3=A9e