[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[PATCH v4 13/48] target/loongarch: Implement xavg/xvagr
From: |
Song Gao |
Subject: |
[PATCH v4 13/48] target/loongarch: Implement xavg/xvagr |
Date: |
Wed, 30 Aug 2023 16:48:27 +0800 |
This patch includes:
- XVAVG.{B/H/W/D/}[U];
- XVAVGR.{B/H/W/D}[U].
Signed-off-by: Song Gao <gaosong@loongson.cn>
---
target/loongarch/vec.h | 3 +++
target/loongarch/insns.decode | 17 +++++++++++++
target/loongarch/disas.c | 17 +++++++++++++
target/loongarch/vec_helper.c | 25 ++++++++++----------
target/loongarch/insn_trans/trans_lasx.c.inc | 17 +++++++++++++
5 files changed, 66 insertions(+), 13 deletions(-)
diff --git a/target/loongarch/vec.h b/target/loongarch/vec.h
index 5332dff83c..6ac6b22f20 100644
--- a/target/loongarch/vec.h
+++ b/target/loongarch/vec.h
@@ -50,4 +50,7 @@
#define DO_ADD(a, b) (a + b)
#define DO_SUB(a, b) (a - b)
+#define DO_VAVG(a, b) ((a >> 1) + (b >> 1) + (a & b & 1))
+#define DO_VAVGR(a, b) ((a >> 1) + (b >> 1) + ((a | b) & 1))
+
#endif /* LOONGARCH_VEC_H */
diff --git a/target/loongarch/insns.decode b/target/loongarch/insns.decode
index e1d8b30179..a2cb39750d 100644
--- a/target/loongarch/insns.decode
+++ b/target/loongarch/insns.decode
@@ -1406,6 +1406,23 @@ xvaddwod_w_hu_h 0111 01000100 00001 ..... ..... .....
@vvv
xvaddwod_d_wu_w 0111 01000100 00010 ..... ..... ..... @vvv
xvaddwod_q_du_d 0111 01000100 00011 ..... ..... ..... @vvv
+xvavg_b 0111 01000110 01000 ..... ..... ..... @vvv
+xvavg_h 0111 01000110 01001 ..... ..... ..... @vvv
+xvavg_w 0111 01000110 01010 ..... ..... ..... @vvv
+xvavg_d 0111 01000110 01011 ..... ..... ..... @vvv
+xvavg_bu 0111 01000110 01100 ..... ..... ..... @vvv
+xvavg_hu 0111 01000110 01101 ..... ..... ..... @vvv
+xvavg_wu 0111 01000110 01110 ..... ..... ..... @vvv
+xvavg_du 0111 01000110 01111 ..... ..... ..... @vvv
+xvavgr_b 0111 01000110 10000 ..... ..... ..... @vvv
+xvavgr_h 0111 01000110 10001 ..... ..... ..... @vvv
+xvavgr_w 0111 01000110 10010 ..... ..... ..... @vvv
+xvavgr_d 0111 01000110 10011 ..... ..... ..... @vvv
+xvavgr_bu 0111 01000110 10100 ..... ..... ..... @vvv
+xvavgr_hu 0111 01000110 10101 ..... ..... ..... @vvv
+xvavgr_wu 0111 01000110 10110 ..... ..... ..... @vvv
+xvavgr_du 0111 01000110 10111 ..... ..... ..... @vvv
+
xvreplgr2vr_b 0111 01101001 11110 00000 ..... ..... @vr
xvreplgr2vr_h 0111 01101001 11110 00001 ..... ..... @vr
xvreplgr2vr_w 0111 01101001 11110 00010 ..... ..... @vr
diff --git a/target/loongarch/disas.c b/target/loongarch/disas.c
index 6972e33833..8296aafa98 100644
--- a/target/loongarch/disas.c
+++ b/target/loongarch/disas.c
@@ -1825,6 +1825,23 @@ INSN_LASX(xvaddwod_w_hu_h, vvv)
INSN_LASX(xvaddwod_d_wu_w, vvv)
INSN_LASX(xvaddwod_q_du_d, vvv)
+INSN_LASX(xvavg_b, vvv)
+INSN_LASX(xvavg_h, vvv)
+INSN_LASX(xvavg_w, vvv)
+INSN_LASX(xvavg_d, vvv)
+INSN_LASX(xvavg_bu, vvv)
+INSN_LASX(xvavg_hu, vvv)
+INSN_LASX(xvavg_wu, vvv)
+INSN_LASX(xvavg_du, vvv)
+INSN_LASX(xvavgr_b, vvv)
+INSN_LASX(xvavgr_h, vvv)
+INSN_LASX(xvavgr_w, vvv)
+INSN_LASX(xvavgr_d, vvv)
+INSN_LASX(xvavgr_bu, vvv)
+INSN_LASX(xvavgr_hu, vvv)
+INSN_LASX(xvavgr_wu, vvv)
+INSN_LASX(xvavgr_du, vvv)
+
INSN_LASX(xvreplgr2vr_b, vr)
INSN_LASX(xvreplgr2vr_h, vr)
INSN_LASX(xvreplgr2vr_w, vr)
diff --git a/target/loongarch/vec_helper.c b/target/loongarch/vec_helper.c
index fffc67ce93..a5d425e965 100644
--- a/target/loongarch/vec_helper.c
+++ b/target/loongarch/vec_helper.c
@@ -344,19 +344,18 @@ DO_ODD_U_S(vaddwod_h_bu_b, 16, H, UH, B, UB, DO_ADD)
DO_ODD_U_S(vaddwod_w_hu_h, 32, W, UW, H, UH, DO_ADD)
DO_ODD_U_S(vaddwod_d_wu_w, 64, D, UD, W, UW, DO_ADD)
-#define DO_VAVG(a, b) ((a >> 1) + (b >> 1) + (a & b & 1))
-#define DO_VAVGR(a, b) ((a >> 1) + (b >> 1) + ((a | b) & 1))
-
-#define DO_3OP(NAME, BIT, E, DO_OP) \
-void HELPER(NAME)(void *vd, void *vj, void *vk, uint32_t v) \
-{ \
- int i; \
- VReg *Vd = (VReg *)vd; \
- VReg *Vj = (VReg *)vj; \
- VReg *Vk = (VReg *)vk; \
- for (i = 0; i < LSX_LEN/BIT; i++) { \
- Vd->E(i) = DO_OP(Vj->E(i), Vk->E(i)); \
- } \
+#define DO_3OP(NAME, BIT, E, DO_OP) \
+void HELPER(NAME)(void *vd, void *vj, void *vk, uint32_t desc) \
+{ \
+ int i; \
+ VReg *Vd = (VReg *)vd; \
+ VReg *Vj = (VReg *)vj; \
+ VReg *Vk = (VReg *)vk; \
+ int oprsz = simd_oprsz(desc); \
+ \
+ for (i = 0; i < oprsz / (BIT / 8); i++) { \
+ Vd->E(i) = DO_OP(Vj->E(i), Vk->E(i)); \
+ } \
}
DO_3OP(vavg_b, 8, B, DO_VAVG)
diff --git a/target/loongarch/insn_trans/trans_lasx.c.inc
b/target/loongarch/insn_trans/trans_lasx.c.inc
index 922222bd78..bcd4b03afc 100644
--- a/target/loongarch/insn_trans/trans_lasx.c.inc
+++ b/target/loongarch/insn_trans/trans_lasx.c.inc
@@ -144,6 +144,23 @@ TRANS(xvaddwod_w_hu_h, LASX, gvec_vvv, 32, MO_16,
do_vaddwod_u_s)
TRANS(xvaddwod_d_wu_w, LASX, gvec_vvv, 32, MO_32, do_vaddwod_u_s)
TRANS(xvaddwod_q_du_d, LASX, gvec_vvv, 32, MO_64, do_vaddwod_u_s)
+TRANS(xvavg_b, LASX, gvec_vvv, 32, MO_8, do_vavg_s)
+TRANS(xvavg_h, LASX, gvec_vvv, 32, MO_16, do_vavg_s)
+TRANS(xvavg_w, LASX, gvec_vvv, 32, MO_32, do_vavg_s)
+TRANS(xvavg_d, LASX, gvec_vvv, 32, MO_64, do_vavg_s)
+TRANS(xvavg_bu, LASX, gvec_vvv, 32, MO_8, do_vavg_u)
+TRANS(xvavg_hu, LASX, gvec_vvv, 32, MO_16, do_vavg_u)
+TRANS(xvavg_wu, LASX, gvec_vvv, 32, MO_32, do_vavg_u)
+TRANS(xvavg_du, LASX, gvec_vvv, 32, MO_64, do_vavg_u)
+TRANS(xvavgr_b, LASX, gvec_vvv, 32, MO_8, do_vavgr_s)
+TRANS(xvavgr_h, LASX, gvec_vvv, 32, MO_16, do_vavgr_s)
+TRANS(xvavgr_w, LASX, gvec_vvv, 32, MO_32, do_vavgr_s)
+TRANS(xvavgr_d, LASX, gvec_vvv, 32, MO_64, do_vavgr_s)
+TRANS(xvavgr_bu, LASX, gvec_vvv, 32, MO_8, do_vavgr_u)
+TRANS(xvavgr_hu, LASX, gvec_vvv, 32, MO_16, do_vavgr_u)
+TRANS(xvavgr_wu, LASX, gvec_vvv, 32, MO_32, do_vavgr_u)
+TRANS(xvavgr_du, LASX, gvec_vvv, 32, MO_64, do_vavgr_u)
+
TRANS(xvreplgr2vr_b, LASX, gvec_dup, 32, MO_8)
TRANS(xvreplgr2vr_h, LASX, gvec_dup, 32, MO_16)
TRANS(xvreplgr2vr_w, LASX, gvec_dup, 32, MO_32)
--
2.39.1
- [PATCH v4 06/48] target/loongarch: Implement xvreplgr2vr, (continued)
- [PATCH v4 06/48] target/loongarch: Implement xvreplgr2vr, Song Gao, 2023/08/30
- [PATCH v4 01/48] target/loongarch: Add LASX data support, Song Gao, 2023/08/30
- [PATCH v4 09/48] target/loongarch: Implement xvsadd/xvssub, Song Gao, 2023/08/30
- [PATCH v4 07/48] target/loongarch: Implement xvaddi/xvsubi, Song Gao, 2023/08/30
- [PATCH v4 08/48] target/loongarch: Implement xvneg, Song Gao, 2023/08/30
- [PATCH v4 02/48] target/loongarch: meson.build support build LASX, Song Gao, 2023/08/30
- [PATCH v4 05/48] target/loongarch: Implement xvadd/xvsub, Song Gao, 2023/08/30
- [PATCH v4 13/48] target/loongarch: Implement xavg/xvagr,
Song Gao <=
- [PATCH v4 15/48] target/loongarch: Implement xvadda, Song Gao, 2023/08/30
- [PATCH v4 11/48] target/loongarch: Implement xvhaddw/xvhsubw, Song Gao, 2023/08/30
- [PATCH v4 16/48] target/loongarch: Implement xvmax/xvmin, Song Gao, 2023/08/30
- [PATCH v4 19/48] target/loongarch; Implement xvdiv/xvmod, Song Gao, 2023/08/30