vitalif
/
mirror_qemu
mirror of https://github.com/proxmox/mirror_qemu
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

target/mips: msa: Split helpers for PCNT.<B|H|W|D> 

Achieves clearer code and slightly better performance.

Signed-off-by: Aleksandar Markovic <amarkovic@wavecomp.com>
Reviewed-by: Aleksandar Rikalo <arikalo@wavecomp.com>
Message-Id: <1569415572-19635-9-git-send-email-aleksandar.markovic@rt-rk.com>
master
Aleksandar Markovic 2019-09-25 14:46:00 +02:00
parent 81c4b05995
commit 4c5daf386f
3 changed files with 95 additions and 73 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
target/mips/helper.h
View File

@ -788,6 +788,11 @@ DEF_HELPER_3(msa_nlzc_h, void, env, i32, i32)
DEF_HELPER_3(msa_nlzc_w, void, env, i32, i32)
DEF_HELPER_3(msa_nlzc_d, void, env, i32, i32)
DEF_HELPER_3(msa_pcnt_b, void, env, i32, i32)
DEF_HELPER_3(msa_pcnt_h, void, env, i32, i32)
DEF_HELPER_3(msa_pcnt_w, void, env, i32, i32)
DEF_HELPER_3(msa_pcnt_d, void, env, i32, i32)
DEF_HELPER_4(msa_andi_b, void, env, i32, i32, i32)
DEF_HELPER_4(msa_ori_b, void, env, i32, i32, i32)
@ -946,7 +951,6 @@ DEF_HELPER_4(msa_bmnz_v, void, env, i32, i32, i32)
DEF_HELPER_4(msa_bmz_v, void, env, i32, i32, i32)
DEF_HELPER_4(msa_bsel_v, void, env, i32, i32, i32)
DEF_HELPER_4(msa_fill_df, void, env, i32, i32, i32)
DEF_HELPER_4(msa_pcnt_df, void, env, i32, i32, i32)
DEF_HELPER_4(msa_copy_s_b, void, env, i32, i32, i32)
DEF_HELPER_4(msa_copy_s_h, void, env, i32, i32, i32)

143
target/mips/msa_helper.c
View File

@ -207,6 +207,80 @@ void helper_msa_nlzc_d(CPUMIPSState *env, uint32_t wd, uint32_t ws)
pwd->d[1] = msa_nlzc_df(DF_DOUBLE, pws->d[1]);
}
static inline int64_t msa_pcnt_df(uint32_t df, int64_t arg)
{
uint64_t x;
x = UNSIGNED(arg, df);
x = (x & 0x5555555555555555ULL) + ((x >> 1) & 0x5555555555555555ULL);
x = (x & 0x3333333333333333ULL) + ((x >> 2) & 0x3333333333333333ULL);
x = (x & 0x0F0F0F0F0F0F0F0FULL) + ((x >> 4) & 0x0F0F0F0F0F0F0F0FULL);
x = (x & 0x00FF00FF00FF00FFULL) + ((x >> 8) & 0x00FF00FF00FF00FFULL);
x = (x & 0x0000FFFF0000FFFFULL) + ((x >> 16) & 0x0000FFFF0000FFFFULL);
x = (x & 0x00000000FFFFFFFFULL) + ((x >> 32));
return x;
}
void helper_msa_pcnt_b(CPUMIPSState *env, uint32_t wd, uint32_t ws)
{
wr_t *pwd = &(env->active_fpu.fpr[wd].wr);
wr_t *pws = &(env->active_fpu.fpr[ws].wr);
pwd->b[0] = msa_pcnt_df(DF_BYTE, pws->b[0]);
pwd->b[1] = msa_pcnt_df(DF_BYTE, pws->b[1]);
pwd->b[2] = msa_pcnt_df(DF_BYTE, pws->b[2]);
pwd->b[3] = msa_pcnt_df(DF_BYTE, pws->b[3]);
pwd->b[4] = msa_pcnt_df(DF_BYTE, pws->b[4]);
pwd->b[5] = msa_pcnt_df(DF_BYTE, pws->b[5]);
pwd->b[6] = msa_pcnt_df(DF_BYTE, pws->b[6]);
pwd->b[7] = msa_pcnt_df(DF_BYTE, pws->b[7]);
pwd->b[8] = msa_pcnt_df(DF_BYTE, pws->b[8]);
pwd->b[9] = msa_pcnt_df(DF_BYTE, pws->b[9]);
pwd->b[10] = msa_pcnt_df(DF_BYTE, pws->b[10]);
pwd->b[11] = msa_pcnt_df(DF_BYTE, pws->b[11]);
pwd->b[12] = msa_pcnt_df(DF_BYTE, pws->b[12]);
pwd->b[13] = msa_pcnt_df(DF_BYTE, pws->b[13]);
pwd->b[14] = msa_pcnt_df(DF_BYTE, pws->b[14]);
pwd->b[15] = msa_pcnt_df(DF_BYTE, pws->b[15]);
}
void helper_msa_pcnt_h(CPUMIPSState *env, uint32_t wd, uint32_t ws)
{
wr_t *pwd = &(env->active_fpu.fpr[wd].wr);
wr_t *pws = &(env->active_fpu.fpr[ws].wr);
pwd->h[0] = msa_pcnt_df(DF_HALF, pws->h[0]);
pwd->h[1] = msa_pcnt_df(DF_HALF, pws->h[1]);
pwd->h[2] = msa_pcnt_df(DF_HALF, pws->h[2]);
pwd->h[3] = msa_pcnt_df(DF_HALF, pws->h[3]);
pwd->h[4] = msa_pcnt_df(DF_HALF, pws->h[4]);
pwd->h[5] = msa_pcnt_df(DF_HALF, pws->h[5]);
pwd->h[6] = msa_pcnt_df(DF_HALF, pws->h[6]);
pwd->h[7] = msa_pcnt_df(DF_HALF, pws->h[7]);
}
void helper_msa_pcnt_w(CPUMIPSState *env, uint32_t wd, uint32_t ws)
{
wr_t *pwd = &(env->active_fpu.fpr[wd].wr);
wr_t *pws = &(env->active_fpu.fpr[ws].wr);
pwd->w[0] = msa_pcnt_df(DF_WORD, pws->w[0]);
pwd->w[1] = msa_pcnt_df(DF_WORD, pws->w[1]);
pwd->w[2] = msa_pcnt_df(DF_WORD, pws->w[2]);
pwd->w[3] = msa_pcnt_df(DF_WORD, pws->w[3]);
}
void helper_msa_pcnt_d(CPUMIPSState *env, uint32_t wd, uint32_t ws)
{
wr_t *pwd = &(env->active_fpu.fpr[wd].wr);
wr_t *pws = &(env->active_fpu.fpr[ws].wr);
pwd->d[0] = msa_pcnt_df(DF_DOUBLE, pws->d[0]);
pwd->d[1] = msa_pcnt_df(DF_DOUBLE, pws->d[1]);
}
/*
* Bit Move
@ -2648,22 +2722,6 @@ void helper_msa_move_v(CPUMIPSState *env, uint32_t wd, uint32_t ws)
msa_move_v(pwd, pws);
}
static inline int64_t msa_pcnt_df(uint32_t df, int64_t arg)
{
uint64_t x;
x = UNSIGNED(arg, df);
x = (x & 0x5555555555555555ULL) + ((x >> 1) & 0x5555555555555555ULL);
x = (x & 0x3333333333333333ULL) + ((x >> 2) & 0x3333333333333333ULL);
x = (x & 0x0F0F0F0F0F0F0F0FULL) + ((x >> 4) & 0x0F0F0F0F0F0F0F0FULL);
x = (x & 0x00FF00FF00FF00FFULL) + ((x >> 8) & 0x00FF00FF00FF00FFULL);
x = (x & 0x0000FFFF0000FFFFULL) + ((x >> 16) & 0x0000FFFF0000FFFFULL);
x = (x & 0x00000000FFFFFFFFULL) + ((x >> 32));
return x;
}
void helper_msa_fill_df(CPUMIPSState *env, uint32_t df, uint32_t wd,
uint32_t rs)
{
@ -2696,59 +2754,6 @@ void helper_msa_fill_df(CPUMIPSState *env, uint32_t df, uint32_t wd,
}
}
#define MSA_UNOP_DF(func) \
void helper_msa_ ## func ## _df(CPUMIPSState *env, uint32_t df, \
uint32_t wd, uint32_t ws) \
{ \
wr_t *pwd = &(env->active_fpu.fpr[wd].wr); \
wr_t *pws = &(env->active_fpu.fpr[ws].wr); \
\
switch (df) { \
case DF_BYTE: \
pwd->b[0] = msa_ ## func ## _df(df, pws->b[0]); \
pwd->b[1] = msa_ ## func ## _df(df, pws->b[1]); \
pwd->b[2] = msa_ ## func ## _df(df, pws->b[2]); \
pwd->b[3] = msa_ ## func ## _df(df, pws->b[3]); \
pwd->b[4] = msa_ ## func ## _df(df, pws->b[4]); \
pwd->b[5] = msa_ ## func ## _df(df, pws->b[5]); \
pwd->b[6] = msa_ ## func ## _df(df, pws->b[6]); \
pwd->b[7] = msa_ ## func ## _df(df, pws->b[7]); \
pwd->b[8] = msa_ ## func ## _df(df, pws->b[8]); \
pwd->b[9] = msa_ ## func ## _df(df, pws->b[9]); \
pwd->b[10] = msa_ ## func ## _df(df, pws->b[10]); \
pwd->b[11] = msa_ ## func ## _df(df, pws->b[11]); \
pwd->b[12] = msa_ ## func ## _df(df, pws->b[12]); \
pwd->b[13] = msa_ ## func ## _df(df, pws->b[13]); \
pwd->b[14] = msa_ ## func ## _df(df, pws->b[14]); \
pwd->b[15] = msa_ ## func ## _df(df, pws->b[15]); \
break; \
case DF_HALF: \
pwd->h[0] = msa_ ## func ## _df(df, pws->h[0]); \
pwd->h[1] = msa_ ## func ## _df(df, pws->h[1]); \
pwd->h[2] = msa_ ## func ## _df(df, pws->h[2]); \
pwd->h[3] = msa_ ## func ## _df(df, pws->h[3]); \
pwd->h[4] = msa_ ## func ## _df(df, pws->h[4]); \
pwd->h[5] = msa_ ## func ## _df(df, pws->h[5]); \
pwd->h[6] = msa_ ## func ## _df(df, pws->h[6]); \
pwd->h[7] = msa_ ## func ## _df(df, pws->h[7]); \
break; \
case DF_WORD: \
pwd->w[0] = msa_ ## func ## _df(df, pws->w[0]); \
pwd->w[1] = msa_ ## func ## _df(df, pws->w[1]); \
pwd->w[2] = msa_ ## func ## _df(df, pws->w[2]); \
pwd->w[3] = msa_ ## func ## _df(df, pws->w[3]); \
break; \
case DF_DOUBLE: \
pwd->d[0] = msa_ ## func ## _df(df, pws->d[0]); \
pwd->d[1] = msa_ ## func ## _df(df, pws->d[1]); \
break; \
default: \
assert(0); \
} \
}
MSA_UNOP_DF(pcnt)
#undef MSA_UNOP_DF
#define FLOAT_ONE32 make_float32(0x3f8 << 20)
#define FLOAT_ONE64 make_float64(0x3ffULL << 52)

19
target/mips/translate.c
View File

@ -28958,9 +28958,6 @@ static void gen_msa_2r(CPUMIPSState *env, DisasContext *ctx)
#endif
gen_helper_msa_fill_df(cpu_env, tdf, twd, tws); /* trs */
break;
case OPC_PCNT_df:
gen_helper_msa_pcnt_df(cpu_env, tdf, twd, tws);
break;
case OPC_NLOC_df:
switch (df) {
case DF_BYTE:
@ -28993,6 +28990,22 @@ static void gen_msa_2r(CPUMIPSState *env, DisasContext *ctx)
break;
}
break;
case OPC_PCNT_df:
switch (df) {
case DF_BYTE:
gen_helper_msa_pcnt_b(cpu_env, twd, tws);
break;
case DF_HALF:
gen_helper_msa_pcnt_h(cpu_env, twd, tws);
break;
case DF_WORD:
gen_helper_msa_pcnt_w(cpu_env, twd, tws);
break;
case DF_DOUBLE:
gen_helper_msa_pcnt_d(cpu_env, twd, tws);
break;
}
break;
default:
MIPS_INVAL("MSA instruction");
generate_exception_end(ctx, EXCP_RI);