@@ -2235,6 +2235,11 @@ INSN_LASX(xvbitrevi_h, vv_i)
INSN_LASX(xvbitrevi_w, vv_i)
INSN_LASX(xvbitrevi_d, vv_i)
+INSN_LASX(xvfrstp_b, vvv)
+INSN_LASX(xvfrstp_h, vvv)
+INSN_LASX(xvfrstpi_b, vv_i)
+INSN_LASX(xvfrstpi_h, vv_i)
+
INSN_LASX(xvreplgr2vr_b, vr)
INSN_LASX(xvreplgr2vr_h, vr)
INSN_LASX(xvreplgr2vr_w, vr)
@@ -526,10 +526,10 @@ DEF_HELPER_FLAGS_4(vbitrevi_h, TCG_CALL_NO_RWG, void, ptr, ptr, i64, i32)
DEF_HELPER_FLAGS_4(vbitrevi_w, TCG_CALL_NO_RWG, void, ptr, ptr, i64, i32)
DEF_HELPER_FLAGS_4(vbitrevi_d, TCG_CALL_NO_RWG, void, ptr, ptr, i64, i32)
-DEF_HELPER_4(vfrstp_b, void, env, i32, i32, i32)
-DEF_HELPER_4(vfrstp_h, void, env, i32, i32, i32)
-DEF_HELPER_4(vfrstpi_b, void, env, i32, i32, i32)
-DEF_HELPER_4(vfrstpi_h, void, env, i32, i32, i32)
+DEF_HELPER_5(vfrstp_b, void, env, i32, i32, i32, i32)
+DEF_HELPER_5(vfrstp_h, void, env, i32, i32, i32, i32)
+DEF_HELPER_5(vfrstpi_b, void, env, i32, i32, i32, i32)
+DEF_HELPER_5(vfrstpi_h, void, env, i32, i32, i32, i32)
DEF_HELPER_4(vfadd_s, void, env, i32, i32, i32)
DEF_HELPER_4(vfadd_d, void, env, i32, i32, i32)
@@ -613,6 +613,11 @@ TRANS(xvbitrevi_h, gvec_vv_i, 32, MO_16, do_vbitrevi)
TRANS(xvbitrevi_w, gvec_vv_i, 32, MO_32, do_vbitrevi)
TRANS(xvbitrevi_d, gvec_vv_i, 32, MO_64, do_vbitrevi)
+TRANS(xvfrstp_b, gen_vvv, 32, gen_helper_vfrstp_b)
+TRANS(xvfrstp_h, gen_vvv, 32, gen_helper_vfrstp_h)
+TRANS(xvfrstpi_b, gen_vv_i, 32, gen_helper_vfrstpi_b)
+TRANS(xvfrstpi_h, gen_vv_i, 32, gen_helper_vfrstpi_h)
+
TRANS(xvreplgr2vr_b, gvec_dup, 32, MO_8)
TRANS(xvreplgr2vr_h, gvec_dup, 32, MO_16)
TRANS(xvreplgr2vr_w, gvec_dup, 32, MO_32)
@@ -1811,6 +1811,11 @@ xvbitrevi_h 0111 01110001 10000 1 .... ..... ..... @vv_ui4
xvbitrevi_w 0111 01110001 10001 ..... ..... ..... @vv_ui5
xvbitrevi_d 0111 01110001 1001 ...... ..... ..... @vv_ui6
+xvfrstp_b 0111 01010010 10110 ..... ..... ..... @vvv
+xvfrstp_h 0111 01010010 10111 ..... ..... ..... @vvv
+xvfrstpi_b 0111 01101001 10100 ..... ..... ..... @vv_ui5
+xvfrstpi_h 0111 01101001 10101 ..... ..... ..... @vv_ui5
+
xvreplgr2vr_b 0111 01101001 11110 00000 ..... ..... @vr
xvreplgr2vr_h 0111 01101001 11110 00001 ..... ..... @vr
xvreplgr2vr_w 0111 01101001 11110 00010 ..... ..... @vr
@@ -2387,42 +2387,61 @@ DO_BITI(vbitrevi_h, 16, UH, DO_BITREV)
DO_BITI(vbitrevi_w, 32, UW, DO_BITREV)
DO_BITI(vbitrevi_d, 64, UD, DO_BITREV)
-#define VFRSTP(NAME, BIT, MASK, E) \
-void HELPER(NAME)(CPULoongArchState *env, \
- uint32_t vd, uint32_t vj, uint32_t vk) \
-{ \
- int i, m; \
- VReg *Vd = &(env->fpr[vd].vreg); \
- VReg *Vj = &(env->fpr[vj].vreg); \
- VReg *Vk = &(env->fpr[vk].vreg); \
- \
- for (i = 0; i < LSX_LEN/BIT; i++) { \
- if (Vj->E(i) < 0) { \
- break; \
- } \
- } \
- m = Vk->E(0) & MASK; \
- Vd->E(m) = i; \
-}
-
-VFRSTP(vfrstp_b, 8, 0xf, B)
-VFRSTP(vfrstp_h, 16, 0x7, H)
-
-#define VFRSTPI(NAME, BIT, E) \
-void HELPER(NAME)(CPULoongArchState *env, \
- uint32_t vd, uint32_t vj, uint32_t imm) \
+#define VFRSTP(NAME, BIT, MASK, E) \
+void HELPER(NAME)(CPULoongArchState *env, uint32_t oprsz, \
+ uint32_t vd, uint32_t vj, uint32_t vk) \
{ \
- int i, m; \
+ int i, j, m, max; \
VReg *Vd = &(env->fpr[vd].vreg); \
VReg *Vj = &(env->fpr[vj].vreg); \
+ VReg *Vk = &(env->fpr[vk].vreg); \
\
- for (i = 0; i < LSX_LEN/BIT; i++) { \
+ max = LSX_LEN / BIT; \
+ m = Vk->E(0) & MASK; \
+ for (i = 0; i < max; i++) { \
if (Vj->E(i) < 0) { \
break; \
} \
} \
- m = imm % (LSX_LEN/BIT); \
Vd->E(m) = i; \
+ if (oprsz == 32) { \
+ for (j = 0; j < max; j++) { \
+ if (Vj->E(j + max) < 0) { \
+ break; \
+ } \
+ } \
+ m = Vk->E(max) & MASK; \
+ Vd->E(m + max) = j; \
+ } \
+}
+
+VFRSTP(vfrstp_b, 8, 0xf, B)
+VFRSTP(vfrstp_h, 16, 0x7, H)
+
+#define VFRSTPI(NAME, BIT, E) \
+void HELPER(NAME)(CPULoongArchState *env, uint32_t oprsz, \
+ uint32_t vd, uint32_t vj, uint32_t imm) \
+{ \
+ int i, j, m, max; \
+ VReg *Vd = &(env->fpr[vd].vreg); \
+ VReg *Vj = &(env->fpr[vj].vreg); \
+ \
+ max = LSX_LEN / BIT; \
+ m = imm % max; \
+ for (i = 0; i < max; i++) { \
+ if (Vj->E(i) < 0) { \
+ break; \
+ } \
+ } \
+ Vd->E(m) = i; \
+ if (oprsz == 32) { \
+ for(j = 0; j < max; j++) { \
+ if(Vj->E(j + max) < 0) { \
+ break; \
+ } \
+ } \
+ Vd->E(m + max) = j; \
+ } \
}
VFRSTPI(vfrstpi_b, 8, B)
This patch includes: - XVFRSTP[I].{B/H}. Signed-off-by: Song Gao <gaosong@loongson.cn> --- target/loongarch/disas.c | 5 ++ target/loongarch/helper.h | 8 +-- target/loongarch/insn_trans/trans_lasx.c.inc | 5 ++ target/loongarch/insns.decode | 5 ++ target/loongarch/vec_helper.c | 73 ++++++++++++-------- 5 files changed, 65 insertions(+), 31 deletions(-)