diff options
| -rw-r--r-- | target/arm/neon-shared.decode | 18 | ||||
| -rw-r--r-- | target/arm/translate-neon.c.inc | 22 |
2 files changed, 24 insertions, 16 deletions
diff --git a/target/arm/neon-shared.decode b/target/arm/neon-shared.decode index f297ba8..a9d0108 100644 --- a/target/arm/neon-shared.decode +++ b/target/arm/neon-shared.decode @@ -34,11 +34,17 @@ %vd_dp 22:1 12:4 %vd_sp 12:4 22:1 -VCMLA 1111 110 rot:2 . 1 size:1 .... .... 1000 . q:1 . 0 .... \ - vm=%vm_dp vn=%vn_dp vd=%vd_dp +# For VCMLA/VCADD insns, convert the single-bit size field +# which is 0 for fp16 and 1 for fp32 into a MO_* constant. +# (Note that this is the reverse of the sense of the 1-bit size +# field in the 3same_fp Neon insns.) +%vcadd_size 20:1 !function=plus1 -VCADD 1111 110 rot:1 1 . 0 size:1 .... .... 1000 . q:1 . 0 .... \ - vm=%vm_dp vn=%vn_dp vd=%vd_dp +VCMLA 1111 110 rot:2 . 1 . .... .... 1000 . q:1 . 0 .... \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp size=%vcadd_size + +VCADD 1111 110 rot:1 1 . 0 . .... .... 1000 . q:1 . 0 .... \ + vm=%vm_dp vn=%vn_dp vd=%vd_dp size=%vcadd_size # VUDOT and VSDOT VDOT 1111 110 00 . 10 .... .... 1101 . q:1 . u:1 .... \ @@ -51,9 +57,9 @@ VFML 1111 110 0 s:1 . 10 .... .... 1000 . 1 . 1 .... \ vm=%vm_dp vn=%vn_dp vd=%vd_dp q=1 VCMLA_scalar 1111 1110 0 . rot:2 .... .... 1000 . q:1 index:1 0 vm:4 \ - vn=%vn_dp vd=%vd_dp size=0 + vn=%vn_dp vd=%vd_dp size=1 VCMLA_scalar 1111 1110 1 . rot:2 .... .... 1000 . q:1 . 0 .... \ - vm=%vm_dp vn=%vn_dp vd=%vd_dp size=1 index=0 + vm=%vm_dp vn=%vn_dp vd=%vd_dp size=2 index=0 VDOT_scalar 1111 1110 0 . 10 .... .... 1101 . q:1 index:1 u:1 rm:4 \ vm=%vm_dp vn=%vn_dp vd=%vd_dp diff --git a/target/arm/translate-neon.c.inc b/target/arm/translate-neon.c.inc index 213c1c2..4d1a292 100644 --- a/target/arm/translate-neon.c.inc +++ b/target/arm/translate-neon.c.inc @@ -168,7 +168,7 @@ static bool trans_VCMLA(DisasContext *s, arg_VCMLA *a) gen_helper_gvec_3_ptr *fn_gvec_ptr; if (!dc_isar_feature(aa32_vcma, s) - || (!a->size && !dc_isar_feature(aa32_fp16_arith, s))) { + || (a->size == MO_16 && !dc_isar_feature(aa32_fp16_arith, s))) { return false; } @@ -187,8 +187,9 @@ static bool trans_VCMLA(DisasContext *s, arg_VCMLA *a) } opr_sz = (1 + a->q) * 8; - fpst = fpstatus_ptr(a->size == 0 ? FPST_STD_F16 : FPST_STD); - fn_gvec_ptr = a->size ? gen_helper_gvec_fcmlas : gen_helper_gvec_fcmlah; + fpst = fpstatus_ptr(a->size == MO_16 ? FPST_STD_F16 : FPST_STD); + fn_gvec_ptr = (a->size == MO_16) ? + gen_helper_gvec_fcmlah : gen_helper_gvec_fcmlas; tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd), vfp_reg_offset(1, a->vn), vfp_reg_offset(1, a->vm), @@ -205,7 +206,7 @@ static bool trans_VCADD(DisasContext *s, arg_VCADD *a) gen_helper_gvec_3_ptr *fn_gvec_ptr; if (!dc_isar_feature(aa32_vcma, s) - || (!a->size && !dc_isar_feature(aa32_fp16_arith, s))) { + || (a->size == MO_16 && !dc_isar_feature(aa32_fp16_arith, s))) { return false; } @@ -224,8 +225,9 @@ static bool trans_VCADD(DisasContext *s, arg_VCADD *a) } opr_sz = (1 + a->q) * 8; - fpst = fpstatus_ptr(a->size == 0 ? FPST_STD_F16 : FPST_STD); - fn_gvec_ptr = a->size ? gen_helper_gvec_fcadds : gen_helper_gvec_fcaddh; + fpst = fpstatus_ptr(a->size == MO_16 ? FPST_STD_F16 : FPST_STD); + fn_gvec_ptr = (a->size == MO_16) ? + gen_helper_gvec_fcaddh : gen_helper_gvec_fcadds; tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd), vfp_reg_offset(1, a->vn), vfp_reg_offset(1, a->vm), @@ -307,7 +309,7 @@ static bool trans_VCMLA_scalar(DisasContext *s, arg_VCMLA_scalar *a) if (!dc_isar_feature(aa32_vcma, s)) { return false; } - if (a->size == 0 && !dc_isar_feature(aa32_fp16_arith, s)) { + if (a->size == MO_16 && !dc_isar_feature(aa32_fp16_arith, s)) { return false; } @@ -325,10 +327,10 @@ static bool trans_VCMLA_scalar(DisasContext *s, arg_VCMLA_scalar *a) return true; } - fn_gvec_ptr = (a->size ? gen_helper_gvec_fcmlas_idx - : gen_helper_gvec_fcmlah_idx); + fn_gvec_ptr = (a->size == MO_16) ? + gen_helper_gvec_fcmlah_idx : gen_helper_gvec_fcmlas_idx; opr_sz = (1 + a->q) * 8; - fpst = fpstatus_ptr(a->size == 0 ? FPST_STD_F16 : FPST_STD); + fpst = fpstatus_ptr(a->size == MO_16 ? FPST_STD_F16 : FPST_STD); tcg_gen_gvec_3_ptr(vfp_reg_offset(1, a->vd), vfp_reg_offset(1, a->vn), vfp_reg_offset(1, a->vm), |