From 3d50b7287e3c11b769945fd7352788d69b1a5a5e Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Sat, 4 Nov 2023 14:59:24 -0700 Subject: [PATCH] target/sparc: Implement FHADD, FHSUB, FNHADD, FNADD, FNMUL Signed-off-by: Richard Henderson --- target/sparc/fop_helper.c | 68 +++++++++++++++++++++++++++++++++++ target/sparc/helper.h | 5 +++ target/sparc/insns.decode | 11 ++++++ target/sparc/translate.c | 76 +++++++++++++++++++++++++++++++++++++++ 4 files changed, 160 insertions(+) diff --git a/target/sparc/fop_helper.c b/target/sparc/fop_helper.c index 1de44d79c1..08b5f96f95 100644 --- a/target/sparc/fop_helper.c +++ b/target/sparc/fop_helper.c @@ -359,6 +359,74 @@ float64 helper_fmaddd(CPUSPARCState *env, float64 s1, return ret; } +float32 helper_fnadds(CPUSPARCState *env, float32 src1, float32 src2) +{ + float32 ret = float32_add(src1, src2, &env->fp_status); + + /* + * NaN inputs or result do not get a sign change. + * Nor, apparently, does zero: on hardware, -(x + -x) yields +0. + */ + if (!float32_is_any_nan(ret) && !float32_is_zero(ret)) { + ret = float32_chs(ret); + } + check_ieee_exceptions(env, GETPC()); + return ret; +} + +float32 helper_fnmuls(CPUSPARCState *env, float32 src1, float32 src2) +{ + float32 ret = float32_mul(src1, src2, &env->fp_status); + + /* NaN inputs or result do not get a sign change. */ + if (!float32_is_any_nan(ret)) { + ret = float32_chs(ret); + } + check_ieee_exceptions(env, GETPC()); + return ret; +} + +float64 helper_fnaddd(CPUSPARCState *env, float64 src1, float64 src2) +{ + float64 ret = float64_add(src1, src2, &env->fp_status); + + /* + * NaN inputs or result do not get a sign change. + * Nor, apparently, does zero: on hardware, -(x + -x) yields +0. + */ + if (!float64_is_any_nan(ret) && !float64_is_zero(ret)) { + ret = float64_chs(ret); + } + check_ieee_exceptions(env, GETPC()); + return ret; +} + +float64 helper_fnmuld(CPUSPARCState *env, float64 src1, float64 src2) +{ + float64 ret = float64_mul(src1, src2, &env->fp_status); + + /* NaN inputs or result do not get a sign change. */ + if (!float64_is_any_nan(ret)) { + ret = float64_chs(ret); + } + check_ieee_exceptions(env, GETPC()); + return ret; +} + +float64 helper_fnsmuld(CPUSPARCState *env, float32 src1, float32 src2) +{ + float64 ret = float64_mul(float32_to_float64(src1, &env->fp_status), + float32_to_float64(src2, &env->fp_status), + &env->fp_status); + + /* NaN inputs or result do not get a sign change. */ + if (!float64_is_any_nan(ret)) { + ret = float64_chs(ret); + } + check_ieee_exceptions(env, GETPC()); + return ret; +} + static uint32_t finish_fcmp(CPUSPARCState *env, FloatRelation r, uintptr_t ra) { check_ieee_exceptions(env, ra); diff --git a/target/sparc/helper.h b/target/sparc/helper.h index e59307efc2..15f0907a1b 100644 --- a/target/sparc/helper.h +++ b/target/sparc/helper.h @@ -57,6 +57,8 @@ DEF_HELPER_FLAGS_3(fsubd, TCG_CALL_NO_WG, f64, env, f64, f64) DEF_HELPER_FLAGS_3(fmuld, TCG_CALL_NO_WG, f64, env, f64, f64) DEF_HELPER_FLAGS_3(fdivd, TCG_CALL_NO_WG, f64, env, f64, f64) DEF_HELPER_FLAGS_5(fmaddd, TCG_CALL_NO_WG, f64, env, f64, f64, f64, i32) +DEF_HELPER_FLAGS_3(fnaddd, TCG_CALL_NO_WG, f64, env, f64, f64) +DEF_HELPER_FLAGS_3(fnmuld, TCG_CALL_NO_WG, f64, env, f64, f64) DEF_HELPER_FLAGS_3(faddq, TCG_CALL_NO_WG, i128, env, i128, i128) DEF_HELPER_FLAGS_3(fsubq, TCG_CALL_NO_WG, i128, env, i128, i128) @@ -68,8 +70,11 @@ DEF_HELPER_FLAGS_3(fsubs, TCG_CALL_NO_WG, f32, env, f32, f32) DEF_HELPER_FLAGS_3(fmuls, TCG_CALL_NO_WG, f32, env, f32, f32) DEF_HELPER_FLAGS_3(fdivs, TCG_CALL_NO_WG, f32, env, f32, f32) DEF_HELPER_FLAGS_5(fmadds, TCG_CALL_NO_WG, f32, env, f32, f32, f32, i32) +DEF_HELPER_FLAGS_3(fnadds, TCG_CALL_NO_WG, f32, env, f32, f32) +DEF_HELPER_FLAGS_3(fnmuls, TCG_CALL_NO_WG, f32, env, f32, f32) DEF_HELPER_FLAGS_3(fsmuld, TCG_CALL_NO_WG, f64, env, f32, f32) +DEF_HELPER_FLAGS_3(fnsmuld, TCG_CALL_NO_WG, f64, env, f32, f32) DEF_HELPER_FLAGS_3(fdmulq, TCG_CALL_NO_WG, i128, env, f64, f64) DEF_HELPER_FLAGS_2(fitod, TCG_CALL_NO_WG, f64, env, s32) diff --git a/target/sparc/insns.decode b/target/sparc/insns.decode index 18d068d2af..6ec3838865 100644 --- a/target/sparc/insns.decode +++ b/target/sparc/insns.decode @@ -307,8 +307,19 @@ FMULq 10 ..... 110100 ..... 0 0100 1011 ..... @q_q_q FDIVs 10 ..... 110100 ..... 0 0100 1101 ..... @r_r_r FDIVd 10 ..... 110100 ..... 0 0100 1110 ..... @d_d_d FDIVq 10 ..... 110100 ..... 0 0100 1111 ..... @q_q_q +FNADDs 10 ..... 110100 ..... 0 0101 0001 ..... @r_r_r +FNADDd 10 ..... 110100 ..... 0 0101 0010 ..... @d_d_d +FNMULs 10 ..... 110100 ..... 0 0101 1001 ..... @r_r_r +FNMULd 10 ..... 110100 ..... 0 0101 1010 ..... @d_d_d +FHADDs 10 ..... 110100 ..... 0 0110 0001 ..... @r_r_r +FHADDd 10 ..... 110100 ..... 0 0110 0010 ..... @d_d_d +FHSUBs 10 ..... 110100 ..... 0 0110 0101 ..... @r_r_r +FHSUBd 10 ..... 110100 ..... 0 0110 0110 ..... @d_d_d FsMULd 10 ..... 110100 ..... 0 0110 1001 ..... @d_r_r FdMULq 10 ..... 110100 ..... 0 0110 1110 ..... @q_d_d +FNHADDs 10 ..... 110100 ..... 0 0111 0001 ..... @r_r_r +FNHADDd 10 ..... 110100 ..... 0 0111 0010 ..... @d_d_d +FNsMULd 10 ..... 110100 ..... 0 0111 1001 ..... @d_r_r FsTOx 10 ..... 110100 00000 0 1000 0001 ..... @r_r2 FdTOx 10 ..... 110100 00000 0 1000 0010 ..... @r_d2 FqTOx 10 ..... 110100 00000 0 1000 0011 ..... @r_q2 diff --git a/target/sparc/translate.c b/target/sparc/translate.c index b8ba8eea74..d81b9ce5a8 100644 --- a/target/sparc/translate.c +++ b/target/sparc/translate.c @@ -1234,6 +1234,51 @@ static void gen_op_fnmaddd(TCGv_i64 d, TCGv_i64 s1, TCGv_i64 s2, TCGv_i64 s3) gen_helper_fmaddd(d, tcg_env, s1, s2, s3, tcg_constant_i32(op)); } +/* Use muladd to compute (1 * src1) + src2 / 2 with one rounding. */ +static void gen_op_fhadds(TCGv_i32 d, TCGv_i32 s1, TCGv_i32 s2) +{ + TCGv_i32 one = tcg_constant_i32(float32_one); + int op = float_muladd_halve_result; + gen_helper_fmadds(d, tcg_env, one, s1, s2, tcg_constant_i32(op)); +} + +static void gen_op_fhaddd(TCGv_i64 d, TCGv_i64 s1, TCGv_i64 s2) +{ + TCGv_i64 one = tcg_constant_i64(float64_one); + int op = float_muladd_halve_result; + gen_helper_fmaddd(d, tcg_env, one, s1, s2, tcg_constant_i32(op)); +} + +/* Use muladd to compute (1 * src1) - src2 / 2 with one rounding. */ +static void gen_op_fhsubs(TCGv_i32 d, TCGv_i32 s1, TCGv_i32 s2) +{ + TCGv_i32 one = tcg_constant_i32(float32_one); + int op = float_muladd_negate_c | float_muladd_halve_result; + gen_helper_fmadds(d, tcg_env, one, s1, s2, tcg_constant_i32(op)); +} + +static void gen_op_fhsubd(TCGv_i64 d, TCGv_i64 s1, TCGv_i64 s2) +{ + TCGv_i64 one = tcg_constant_i64(float64_one); + int op = float_muladd_negate_c | float_muladd_halve_result; + gen_helper_fmaddd(d, tcg_env, one, s1, s2, tcg_constant_i32(op)); +} + +/* Use muladd to compute -((1 * src1) + src2 / 2) with one rounding. */ +static void gen_op_fnhadds(TCGv_i32 d, TCGv_i32 s1, TCGv_i32 s2) +{ + TCGv_i32 one = tcg_constant_i32(float32_one); + int op = float_muladd_negate_result | float_muladd_halve_result; + gen_helper_fmadds(d, tcg_env, one, s1, s2, tcg_constant_i32(op)); +} + +static void gen_op_fnhaddd(TCGv_i64 d, TCGv_i64 s1, TCGv_i64 s2) +{ + TCGv_i64 one = tcg_constant_i64(float64_one); + int op = float_muladd_negate_result | float_muladd_halve_result; + gen_helper_fmaddd(d, tcg_env, one, s1, s2, tcg_constant_i32(op)); +} + static void gen_op_fpexception_im(DisasContext *dc, int ftt) { /* @@ -4710,6 +4755,10 @@ TRANS(FXNORs, VIS1, do_fff, a, tcg_gen_eqv_i32) TRANS(FORNOTs, VIS1, do_fff, a, tcg_gen_orc_i32) TRANS(FORs, VIS1, do_fff, a, tcg_gen_or_i32) +TRANS(FHADDs, VIS3, do_fff, a, gen_op_fhadds) +TRANS(FHSUBs, VIS3, do_fff, a, gen_op_fhsubs) +TRANS(FNHADDs, VIS3, do_fff, a, gen_op_fnhadds) + static bool do_env_fff(DisasContext *dc, arg_r_r_r *a, void (*func)(TCGv_i32, TCGv_env, TCGv_i32, TCGv_i32)) { @@ -4730,6 +4779,8 @@ TRANS(FADDs, ALL, do_env_fff, a, gen_helper_fadds) TRANS(FSUBs, ALL, do_env_fff, a, gen_helper_fsubs) TRANS(FMULs, ALL, do_env_fff, a, gen_helper_fmuls) TRANS(FDIVs, ALL, do_env_fff, a, gen_helper_fdivs) +TRANS(FNADDs, VIS3, do_env_fff, a, gen_helper_fnadds) +TRANS(FNMULs, VIS3, do_env_fff, a, gen_helper_fnmuls) static bool do_dff(DisasContext *dc, arg_r_r_r *a, void (*func)(TCGv_i64, TCGv_i32, TCGv_i32)) @@ -4827,6 +4878,10 @@ TRANS(FPACK32, VIS1, do_ddd, a, gen_op_fpack32) TRANS(FALIGNDATAg, VIS1, do_ddd, a, gen_op_faligndata) TRANS(BSHUFFLE, VIS2, do_ddd, a, gen_op_bshuffle) +TRANS(FHADDd, VIS3, do_ddd, a, gen_op_fhaddd) +TRANS(FHSUBd, VIS3, do_ddd, a, gen_op_fhsubd) +TRANS(FNHADDd, VIS3, do_ddd, a, gen_op_fnhaddd) + static bool do_rdd(DisasContext *dc, arg_r_r_r *a, void (*func)(TCGv, TCGv_i64, TCGv_i64)) { @@ -4876,6 +4931,8 @@ TRANS(FADDd, ALL, do_env_ddd, a, gen_helper_faddd) TRANS(FSUBd, ALL, do_env_ddd, a, gen_helper_fsubd) TRANS(FMULd, ALL, do_env_ddd, a, gen_helper_fmuld) TRANS(FDIVd, ALL, do_env_ddd, a, gen_helper_fdivd) +TRANS(FNADDd, VIS3, do_env_ddd, a, gen_helper_fnaddd) +TRANS(FNMULd, VIS3, do_env_ddd, a, gen_helper_fnmuld) static bool trans_FsMULd(DisasContext *dc, arg_r_r_r *a) { @@ -4897,6 +4954,25 @@ static bool trans_FsMULd(DisasContext *dc, arg_r_r_r *a) return advance_pc(dc); } +static bool trans_FNsMULd(DisasContext *dc, arg_r_r_r *a) +{ + TCGv_i64 dst; + TCGv_i32 src1, src2; + + if (!avail_VIS3(dc)) { + return false; + } + if (gen_trap_ifnofpu(dc)) { + return true; + } + dst = tcg_temp_new_i64(); + src1 = gen_load_fpr_F(dc, a->rs1); + src2 = gen_load_fpr_F(dc, a->rs2); + gen_helper_fnsmuld(dst, tcg_env, src1, src2); + gen_store_fpr_D(dc, a->rd, dst); + return advance_pc(dc); +} + static bool do_ffff(DisasContext *dc, arg_r_r_r_r *a, void (*func)(TCGv_i32, TCGv_i32, TCGv_i32, TCGv_i32)) {