yuzu/unicorn
1
0
Fork 0
mirror of https://github.com/yuzu-emu/unicorn.git synced 2025-11-04 13:44:49 +00:00
Code Issues Packages Projects Releases Wiki Activity

target/arm: Convert Neon 3-same-fp size field to MO_* in decode

Browse source 
In the Neon instructions, some instruction formats have a 2-bit size
field which corresponds exactly to QEMU's MO_8/16/32/64. However the
floating-point insns in the 3-same group have a 1-bit size field
which is "0 for 32-bit float and 1 for 16-bit float". Currently we
pass these values directly through to trans_ functions, which means
that when reading a particular trans_ function you need to know if
that insn uses a 2-bit size or a 1-bit size.

Move the handling of the 1-bit size to the decodetree file, so that
all these insns consistently pass a size to the trans_ function which
is an MO_8/16/32/64 value.

In this commit we switch over the insns using the 3same_fp and
3same_fp_q0 formats.

Backports commit 6cf0f240e0b980a877abed12d2995f740eae6515
This commit is contained in:
Peter Maydell 2021-03-01 18:15:11 -05:00 committed by Lioncash
parent cd79d2a915
commit 524b54bc7b
2 changed files with 21 additions and 10 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

15
qemu/target/arm/neon-dp.decode
View file

@ -45,11 +45,16 @@
@3same_q0 .... ... . . . size:2 .... .... .... . 0 . . .... \ @3same_q0 .... ... . . . size:2 .... .... .... . 0 . . .... \
&3same vm=%vm_dp vn=%vn_dp vd=%vd_dp q=0 &3same vm=%vm_dp vn=%vn_dp vd=%vd_dp q=0
# For FP insns the high bit of 'size' is used as part of opcode decode # For FP insns the high bit of 'size' is used as part of opcode decode,
@3same_fp .... ... . . . . size:1 .... .... .... . q:1 . . .... \ # and the 'size' bit is 0 for 32-bit float and 1 for 16-bit float.
&3same vm=%vm_dp vn=%vn_dp vd=%vd_dp # This converts this encoding to the same MO_8/16/32/64 values that the
@3same_fp_q0 .... ... . . . . size:1 .... .... .... . 0 . . .... \ # integer neon insns use.
&3same vm=%vm_dp vn=%vn_dp vd=%vd_dp q=0 %3same_fp_size 20:1 !function=neon_3same_fp_size
@3same_fp .... ... . . . . . .... .... .... . q:1 . . .... \
&3same vm=%vm_dp vn=%vn_dp vd=%vd_dp size=%3same_fp_size
@3same_fp_q0 .... ... . . . . . .... .... .... . 0 . . .... \
&3same vm=%vm_dp vn=%vn_dp vd=%vd_dp q=0 size=%3same_fp_size
VHADD_S_3s 1111 001 0 0 . .. .... .... 0000 . . . 0 .... @3same VHADD_S_3s 1111 001 0 0 . .. .... .... 0000 . . . 0 .... @3same
VHADD_U_3s 1111 001 1 0 . .. .... .... 0000 . . . 0 .... @3same VHADD_U_3s 1111 001 1 0 . .. .... .... 0000 . . . 0 .... @3same

16
qemu/target/arm/translate-neon.inc.c
View file

@ -49,6 +49,12 @@ static inline int rsub_8(DisasContext *s, int x)
return 8 - x; return 8 - x;
} }
static inline int neon_3same_fp_size(DisasContext *s, int x)
{
/* Convert 0==fp32, 1==fp16 into a MO_* value */
return MO_32 - x;
}
/* Include the generated Neon decoder */ /* Include the generated Neon decoder */
#include "decode-neon-dp.inc.c" #include "decode-neon-dp.inc.c"
#include "decode-neon-ls.inc.c" #include "decode-neon-ls.inc.c"
@ -1067,7 +1073,7 @@ DO_3SAME_VQDMULH(VQRDMULH, qrdmulh)
WRAP_FP_GVEC(gen_##INSN##_fp16_3s, FPST_STD_F16, HFUNC) \ WRAP_FP_GVEC(gen_##INSN##_fp16_3s, FPST_STD_F16, HFUNC) \
static bool trans_##INSN##_fp_3s(DisasContext *s, arg_3same *a) \ static bool trans_##INSN##_fp_3s(DisasContext *s, arg_3same *a) \
{ \ { \
if (a->size != 0) { \ if (a->size == MO_16) { \
if (!dc_isar_feature(aa32_fp16_arith, s)) { \ if (!dc_isar_feature(aa32_fp16_arith, s)) { \
return false; \ return false; \
} \ } \
@ -1106,7 +1112,7 @@ static bool trans_VMAXNM_fp_3s(DisasContext *s, arg_3same *a)
return false; return false;
} }
if (a->size != 0) { if (a->size == MO_16) {
if (!dc_isar_feature(aa32_fp16_arith, s)) { if (!dc_isar_feature(aa32_fp16_arith, s)) {
return false; return false;
} }
@ -1122,7 +1128,7 @@ static bool trans_VMINNM_fp_3s(DisasContext *s, arg_3same *a)
return false; return false;
} }
if (a->size != 0) { if (a->size == MO_16) {
if (!dc_isar_feature(aa32_fp16_arith, s)) { if (!dc_isar_feature(aa32_fp16_arith, s)) {
return false; return false;
} }
@ -1155,7 +1161,7 @@ static bool do_3same_fp_pair(DisasContext *s, arg_3same *a,
assert(a->q == 0); /* enforced by decode patterns */ assert(a->q == 0); /* enforced by decode patterns */
fpstatus = fpstatus_ptr(tcg_ctx, a->size != 0 ? FPST_STD_F16 : FPST_STD); fpstatus = fpstatus_ptr(tcg_ctx, a->size == MO_16 ? FPST_STD_F16 : FPST_STD);
tcg_gen_gvec_3_ptr(tcg_ctx, vfp_reg_offset(1, a->vd), tcg_gen_gvec_3_ptr(tcg_ctx, vfp_reg_offset(1, a->vd),
vfp_reg_offset(1, a->vn), vfp_reg_offset(1, a->vn),
vfp_reg_offset(1, a->vm), vfp_reg_offset(1, a->vm),
@ -1172,7 +1178,7 @@ static bool do_3same_fp_pair(DisasContext *s, arg_3same *a,
#define DO_3S_FP_PAIR(INSN,FUNC) \ #define DO_3S_FP_PAIR(INSN,FUNC) \
static bool trans_##INSN##_fp_3s(DisasContext *s, arg_3same *a) \ static bool trans_##INSN##_fp_3s(DisasContext *s, arg_3same *a) \
{ \ { \
if (a->size != 0) { \ if (a->size == MO_16) { \
if (!dc_isar_feature(aa32_fp16_arith, s)) { \ if (!dc_isar_feature(aa32_fp16_arith, s)) { \
return false; \ return false; \
} \ } \