[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[PATCH] target/arm: Implement SVE2 fp convert precision
|
From: |
Stephen Long |
|
Subject: |
[PATCH] target/arm: Implement SVE2 fp convert precision |
|
Date: |
Thu, 30 Apr 2020 06:02:08 -0700 |
Implements FCVTLT, FCVTX, FCVTNT, FCVTXNT
Signed-off-by: Stephen Long <address@hidden>
---
target/arm/helper-sve.h | 15 ++++++++++
target/arm/sve.decode | 8 ++++++
target/arm/sve_helper.c | 59 ++++++++++++++++++++++++++++++++++++++
target/arm/translate-sve.c | 48 +++++++++++++++++++++++++++++++
4 files changed, 130 insertions(+)
Ignore my previous patchset. I missed the insn FCVTX and I ended up
squashing those commits into this one. At the moment, I'm working on a patch
for the SVE2 bitwise shift by imm insns.
diff --git a/target/arm/helper-sve.h b/target/arm/helper-sve.h
index 0a62eef94e..22c91913ba 100644
--- a/target/arm/helper-sve.h
+++ b/target/arm/helper-sve.h
@@ -2730,4 +2730,19 @@ DEF_HELPER_FLAGS_5(sve2_cmla_idx_s, TCG_CALL_NO_RWG,
DEF_HELPER_FLAGS_5(sve2_sqrdcmlah_idx_h, TCG_CALL_NO_RWG,
void, ptr, ptr, ptr, ptr, i32)
DEF_HELPER_FLAGS_5(sve2_sqrdcmlah_idx_s, TCG_CALL_NO_RWG,
+
+DEF_HELPER_FLAGS_5(sve2_fcvtxnt, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr,
i32)
+
+DEF_HELPER_FLAGS_5(sve2_fcvtnt_sh, TCG_CALL_NO_RWG,
+ void, ptr, ptr, ptr, ptr, i32)
+DEF_HELPER_FLAGS_5(sve2_fcvtnt_ds, TCG_CALL_NO_RWG,
+ void, ptr, ptr, ptr, ptr, i32)
+
+DEF_HELPER_FLAGS_5(sve2_fcvtlt_hs, TCG_CALL_NO_RWG,
+ void, ptr, ptr, ptr, ptr, i32)
+DEF_HELPER_FLAGS_5(sve2_fcvtlt_sd, TCG_CALL_NO_RWG,
+ void, ptr, ptr, ptr, ptr, i32)
+
+DEF_HELPER_FLAGS_5(do_fcvtx, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32)
+DEF_HELPER_FLAGS_5(sve2_fcvtx_ds, TCG_CALL_NO_RWG,
void, ptr, ptr, ptr, ptr, i32)
diff --git a/target/arm/sve.decode b/target/arm/sve.decode
index 3cf824bac5..bc48094ed8 100644
--- a/target/arm/sve.decode
+++ b/target/arm/sve.decode
@@ -1568,3 +1568,11 @@ SM4E 01000101 00 10001 1 11100 0 ..... .....
@rdn_rm_e0
# SVE2 crypto constructive binary operations
SM4EKEY 01000101 00 1 ..... 11110 0 ..... ..... @rd_rn_rm_e0
RAX1 01000101 00 1 ..... 11110 1 ..... ..... @rd_rn_rm_e0
+
+### SVE2 floating-point convert precision
+FCVTXNT 01100100 00 0010 10 101 ... ..... ..... @rd_pg_rn_e0
+FCVTNT_sh 01100100 10 0010 00 101 ... ..... ..... @rd_pg_rn_e0
+FCVTLT_hs 01100100 10 0010 01 101 ... ..... ..... @rd_pg_rn_e0
+FCVTNT_ds 01100100 11 0010 10 101 ... ..... ..... @rd_pg_rn_e0
+FCVTLT_sd 01100100 11 0010 11 101 ... ..... ..... @rd_pg_rn_e0
+FCVTX_ds 01100101 00 0010 10 101 ... ..... ..... @rd_pg_rn_e0
diff --git a/target/arm/sve_helper.c b/target/arm/sve_helper.c
index aa94df302a..cb56148cdc 100644
--- a/target/arm/sve_helper.c
+++ b/target/arm/sve_helper.c
@@ -4624,6 +4624,8 @@ DO_ZPZ_FP(sve_ucvt_dh, uint64_t, , uint64_to_float16)
DO_ZPZ_FP(sve_ucvt_ds, uint64_t, , uint64_to_float32)
DO_ZPZ_FP(sve_ucvt_dd, uint64_t, , uint64_to_float64)
+DO_ZPZ_FP(do_fcvtx, uint64_t, , float64_to_float32)
+
#undef DO_ZPZ_FP
static void do_fmla_zpzzz_h(void *vd, void *vn, void *vm, void *va, void *vg,
@@ -7622,3 +7624,60 @@ void HELPER(fmmla_d)(void *vd, void *va, void *vn, void
*vm,
d[3] = float64_add(a[3], float64_add(p0, p1, status), status);
}
}
+
+#define DO_FCVTNT(NAME, TYPEW, TYPEN, HW, HN, OP) \
+void HELPER(NAME)(void *vd, void *vn, void *vg, void *status, uint32_t desc) \
+{ \
+ intptr_t i = simd_oprsz(desc); \
+ uint64_t *g = vg; \
+ do { \
+ uint64_t pg = g[(i - 1) >> 6]; \
+ do { \
+ i -= sizeof(TYPEW); \
+ if (likely((pg >> (i & 63)) & 1)) { \
+ TYPEW nn = *(TYPEW *)(vn + HW(i)); \
+ *(TYPEN *)(vd + HN(i + sizeof(TYPEN))) = OP(nn, status); \
+ } \
+ } while (i & 63); \
+ } while (i != 0); \
+}
+
+DO_FCVTNT(sve2_fcvtnt_sh, uint32_t, uint16_t, H1_4, H1_2, sve_f32_to_f16)
+DO_FCVTNT(sve2_fcvtnt_ds, uint64_t, uint32_t, H1_4, H1_2, float64_to_float32)
+
+void HELPER(sve2_fcvtxnt)(void *vd, void *vn, void *vg,
+ void *status, uint32_t desc)
+{
+ set_float_rounding_mode(float_round_to_odd, status);
+ HELPER(sve2_fcvtnt_ds)(vd, vn, vg, status, desc);
+}
+
+#define DO_FCVTLT(NAME, TYPEW, TYPEN, HW, HN, OP) \
+void HELPER(NAME)(void *vd, void *vn, void *vg, void *status, uint32_t desc) \
+{ \
+ intptr_t i = simd_oprsz(desc); \
+ uint64_t *g = vg; \
+ do { \
+ uint64_t pg = g[(i - 1) >> 6]; \
+ do { \
+ i -= sizeof(TYPEW); \
+ if (likely((pg >> (i & 63)) & 1)) { \
+ TYPEN nn = *(TYPEN *)(vn + HN(i + sizeof(TYPEN))); \
+ *(TYPEW *)(vd + HW(i)) = OP(nn, status); \
+ } \
+ } while (i & 63); \
+ } while (i != 0); \
+}
+
+DO_FCVTLT(sve2_fcvtlt_hs, uint32_t, uint16_t, H1_4, H1_2, sve_f16_to_f32)
+DO_FCVTLT(sve2_fcvtlt_sd, uint64_t, uint32_t, H1_4, H1_2, float32_to_float64)
+
+void HELPER(sve2_fcvtx_ds)(void *vd, void *vn, void *vg, void *status,
+ uint32_t desc)
+{
+ set_float_rounding_mode(float_round_to_odd, status);
+ HELPER(do_fcvtx)(vd, vn, vg, status, desc);
+}
+
+#undef DO_FCVTLT
+#undef DO_FCVTNT
diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c
index a8e57ea5f4..7e31468dc0 100644
--- a/target/arm/translate-sve.c
+++ b/target/arm/translate-sve.c
@@ -8253,3 +8253,51 @@ static bool trans_RAX1(DisasContext *s, arg_rrr_esz *a)
}
return true;
}
+
+static bool trans_FCVTXNT(DisasContext *s, arg_rpr_esz *a)
+{
+ if (!dc_isar_feature(aa64_sve2, s)) {
+ return false;
+ }
+ return do_zpz_ptr(s, a->rd, a->rn, a->pg, false, gen_helper_sve2_fcvtxnt);
+}
+
+static bool trans_FCVTNT_sh(DisasContext *s, arg_rpr_esz *a)
+{
+ if (!dc_isar_feature(aa64_sve2, s)) {
+ return false;
+ }
+ return do_zpz_ptr(s, a->rd, a->rn, a->pg, false,
gen_helper_sve2_fcvtnt_sh);
+}
+
+static bool trans_FCVTNT_ds(DisasContext *s, arg_rpr_esz *a)
+{
+ if (!dc_isar_feature(aa64_sve2, s)) {
+ return false;
+ }
+ return do_zpz_ptr(s, a->rd, a->rn, a->pg, false,
gen_helper_sve2_fcvtnt_ds);
+}
+
+static bool trans_FCVTLT_hs(DisasContext *s, arg_rpr_esz *a)
+{
+ if (!dc_isar_feature(aa64_sve2, s)) {
+ return false;
+ }
+ return do_zpz_ptr(s, a->rd, a->rn, a->pg, false,
gen_helper_sve2_fcvtlt_hs);
+}
+
+static bool trans_FCVTLT_sd(DisasContext *s, arg_rpr_esz *a)
+{
+ if (!dc_isar_feature(aa64_sve2, s)) {
+ return false;
+ }
+ return do_zpz_ptr(s, a->rd, a->rn, a->pg, false,
gen_helper_sve2_fcvtlt_sd);
+}
+
+static bool trans_FCVTX_ds(DisasContext *s, arg_rpr_esz *a)
+{
+ if (!dc_isar_feature(aa64_sve2, s)) {
+ return false;
+ }
+ return do_zpz_ptr(s, a->rd, a->rn, a->pg, false, gen_helper_sve2_fcvtx_ds);
+}
--
2.17.1
| [Prev in Thread] |
Current Thread |
[Next in Thread] |
- [PATCH] target/arm: Implement SVE2 fp convert precision,
Stephen Long <=