target/loongarch: Implement xvsrln xvsran

This patch includes:
- XVSRLN.{B.H/H.W/W.D};
- XVSRAN.{B.H/H.W/W.D};
- XVSRLNI.{B.H/H.W/W.D/D.Q};
- XVSRANI.{B.H/H.W/W.D/D.Q}.

Signed-off-by: Song Gao <gaosong@loongson.cn>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20230914022645.1151356-39-gaosong@loongson.cn>

diff --git a/target/loongarch/disas.c b/target/loongarch/disas.c
index bc5eb82b494c2fc5730018de9e5524b8e9011395..28e5e16eb2123ca7eac93a8e31f4a6ec4883fbed 100644 (file)
--- a/target/loongarch/disas.c
+++ b/target/loongarch/disas.c
@@ -2104,6 +2104,22 @@ INSN_LASX(xvsrari_h,         vv_i)
 INSN_LASX(xvsrari_w,         vv_i)
 INSN_LASX(xvsrari_d,         vv_i)
 
+INSN_LASX(xvsrln_b_h,        vvv)
+INSN_LASX(xvsrln_h_w,        vvv)
+INSN_LASX(xvsrln_w_d,        vvv)
+INSN_LASX(xvsran_b_h,        vvv)
+INSN_LASX(xvsran_h_w,        vvv)
+INSN_LASX(xvsran_w_d,        vvv)
+
+INSN_LASX(xvsrlni_b_h,       vv_i)
+INSN_LASX(xvsrlni_h_w,       vv_i)
+INSN_LASX(xvsrlni_w_d,       vv_i)
+INSN_LASX(xvsrlni_d_q,       vv_i)
+INSN_LASX(xvsrani_b_h,       vv_i)
+INSN_LASX(xvsrani_h_w,       vv_i)
+INSN_LASX(xvsrani_w_d,       vv_i)
+INSN_LASX(xvsrani_d_q,       vv_i)
+
 INSN_LASX(xvreplgr2vr_b,     vr)
 INSN_LASX(xvreplgr2vr_h,     vr)
 INSN_LASX(xvreplgr2vr_w,     vr)

diff --git a/target/loongarch/insn_trans/trans_vec.c.inc b/target/loongarch/insn_trans/trans_vec.c.inc
index 8be8670259d2a060b40d4f7eb358eeed94d9e5f9..51a6846c1580f4b726709af6ba4377e407d46a5a 100644 (file)
--- a/target/loongarch/insn_trans/trans_vec.c.inc
+++ b/target/loongarch/insn_trans/trans_vec.c.inc
@@ -3771,6 +3771,12 @@ TRANS(vsrln_w_d, LSX, gen_vvv, gen_helper_vsrln_w_d)
 TRANS(vsran_b_h, LSX, gen_vvv, gen_helper_vsran_b_h)
 TRANS(vsran_h_w, LSX, gen_vvv, gen_helper_vsran_h_w)
 TRANS(vsran_w_d, LSX, gen_vvv, gen_helper_vsran_w_d)
+TRANS(xvsrln_b_h, LASX, gen_xxx, gen_helper_vsrln_b_h)
+TRANS(xvsrln_h_w, LASX, gen_xxx, gen_helper_vsrln_h_w)
+TRANS(xvsrln_w_d, LASX, gen_xxx, gen_helper_vsrln_w_d)
+TRANS(xvsran_b_h, LASX, gen_xxx, gen_helper_vsran_b_h)
+TRANS(xvsran_h_w, LASX, gen_xxx, gen_helper_vsran_h_w)
+TRANS(xvsran_w_d, LASX, gen_xxx, gen_helper_vsran_w_d)
 
 TRANS(vsrlni_b_h, LSX, gen_vv_i, gen_helper_vsrlni_b_h)
 TRANS(vsrlni_h_w, LSX, gen_vv_i, gen_helper_vsrlni_h_w)
@@ -3780,6 +3786,14 @@ TRANS(vsrani_b_h, LSX, gen_vv_i, gen_helper_vsrani_b_h)
 TRANS(vsrani_h_w, LSX, gen_vv_i, gen_helper_vsrani_h_w)
 TRANS(vsrani_w_d, LSX, gen_vv_i, gen_helper_vsrani_w_d)
 TRANS(vsrani_d_q, LSX, gen_vv_i, gen_helper_vsrani_d_q)
+TRANS(xvsrlni_b_h, LASX, gen_xx_i, gen_helper_vsrlni_b_h)
+TRANS(xvsrlni_h_w, LASX, gen_xx_i, gen_helper_vsrlni_h_w)
+TRANS(xvsrlni_w_d, LASX, gen_xx_i, gen_helper_vsrlni_w_d)
+TRANS(xvsrlni_d_q, LASX, gen_xx_i, gen_helper_vsrlni_d_q)
+TRANS(xvsrani_b_h, LASX, gen_xx_i, gen_helper_vsrani_b_h)
+TRANS(xvsrani_h_w, LASX, gen_xx_i, gen_helper_vsrani_h_w)
+TRANS(xvsrani_w_d, LASX, gen_xx_i, gen_helper_vsrani_w_d)
+TRANS(xvsrani_d_q, LASX, gen_xx_i, gen_helper_vsrani_d_q)
 
 TRANS(vsrlrn_b_h, LSX, gen_vvv, gen_helper_vsrlrn_b_h)
 TRANS(vsrlrn_h_w, LSX, gen_vvv, gen_helper_vsrlrn_h_w)

diff --git a/target/loongarch/insns.decode b/target/loongarch/insns.decode
index ca0951e1cc6e067c78c4b852e0d2418a20d8f753..204dcfa0756a571b85bb5b1a90a2771880e739be 100644 (file)
--- a/target/loongarch/insns.decode
+++ b/target/loongarch/insns.decode
@@ -1678,6 +1678,22 @@ xvsrari_h        0111 01101010 10000 1 .... ..... .....   @vv_ui4
 xvsrari_w        0111 01101010 10001 ..... ..... .....    @vv_ui5
 xvsrari_d        0111 01101010 1001 ...... ..... .....    @vv_ui6
 
+xvsrln_b_h       0111 01001111 01001 ..... ..... .....    @vvv
+xvsrln_h_w       0111 01001111 01010 ..... ..... .....    @vvv
+xvsrln_w_d       0111 01001111 01011 ..... ..... .....    @vvv
+xvsran_b_h       0111 01001111 01101 ..... ..... .....    @vvv
+xvsran_h_w       0111 01001111 01110 ..... ..... .....    @vvv
+xvsran_w_d       0111 01001111 01111 ..... ..... .....    @vvv
+
+xvsrlni_b_h      0111 01110100 00000 1 .... ..... .....   @vv_ui4
+xvsrlni_h_w      0111 01110100 00001 ..... ..... .....    @vv_ui5
+xvsrlni_w_d      0111 01110100 0001 ...... ..... .....    @vv_ui6
+xvsrlni_d_q      0111 01110100 001 ....... ..... .....    @vv_ui7
+xvsrani_b_h      0111 01110101 10000 1 .... ..... .....   @vv_ui4
+xvsrani_h_w      0111 01110101 10001 ..... ..... .....    @vv_ui5
+xvsrani_w_d      0111 01110101 1001 ...... ..... .....    @vv_ui6
+xvsrani_d_q      0111 01110101 101 ....... ..... .....    @vv_ui7
+
 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/vec_helper.c b/target/loongarch/vec_helper.c
index bb30d24b8938f0989b448312c7be48d367432b9c..8c405ce32bf075aa80703f3a362fc23a43c1902c 100644 (file)
--- a/target/loongarch/vec_helper.c
+++ b/target/loongarch/vec_helper.c
@@ -1109,105 +1109,121 @@ VSRARI(vsrari_d, 64, D)
 
 #define R_SHIFT(a, b) (a >> b)
 
-#define VSRLN(NAME, BIT, T, E1, E2)                             \
-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;                                      \
-                                                                \
-    for (i = 0; i < LSX_LEN/BIT; i++) {                         \
-        Vd->E1(i) = R_SHIFT((T)Vj->E2(i),((T)Vk->E2(i)) % BIT); \
-    }                                                           \
-    Vd->D(1) = 0;                                               \
+#define VSRLN(NAME, BIT, E1, E2)                                          \
+void HELPER(NAME)(void *vd, void *vj, void *vk, uint32_t desc)            \
+{                                                                         \
+    int i, j, ofs;                                                        \
+    VReg *Vd = (VReg *)vd;                                                \
+    VReg *Vj = (VReg *)vj;                                                \
+    VReg *Vk = (VReg *)vk;                                                \
+    int oprsz = simd_oprsz(desc);                                         \
+                                                                          \
+    ofs = LSX_LEN / BIT;                                                  \
+    for (i = 0; i < oprsz / 16; i++) {                                    \
+        for (j = 0; j < ofs; j++) {                                       \
+            Vd->E1(j + ofs * 2 * i) = R_SHIFT(Vj->E2(j + ofs * i),        \
+                                              Vk->E2(j + ofs * i) % BIT); \
+        }                                                                 \
+        Vd->D(2 * i + 1) = 0;                                             \
+    }                                                                     \
 }
 
-VSRLN(vsrln_b_h, 16, uint16_t, B, H)
-VSRLN(vsrln_h_w, 32, uint32_t, H, W)
-VSRLN(vsrln_w_d, 64, uint64_t, W, D)
+VSRLN(vsrln_b_h, 16, B, UH)
+VSRLN(vsrln_h_w, 32, H, UW)
+VSRLN(vsrln_w_d, 64, W, UD)
 
-#define VSRAN(NAME, BIT, T, E1, E2)                            \
-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;                                     \
-                                                               \
-    for (i = 0; i < LSX_LEN/BIT; i++) {                        \
-        Vd->E1(i) = R_SHIFT(Vj->E2(i), ((T)Vk->E2(i)) % BIT);  \
-    }                                                          \
-    Vd->D(1) = 0;                                              \
+#define VSRAN(NAME, BIT, E1, E2, E3)                                      \
+void HELPER(NAME)(void *vd, void *vj, void *vk, uint32_t desc)            \
+{                                                                         \
+    int i, j, ofs;                                                        \
+    VReg *Vd = (VReg *)vd;                                                \
+    VReg *Vj = (VReg *)vj;                                                \
+    VReg *Vk = (VReg *)vk;                                                \
+    int oprsz = simd_oprsz(desc);                                         \
+                                                                          \
+    ofs = LSX_LEN / BIT;                                                  \
+    for (i = 0; i < oprsz / 16; i++) {                                    \
+        for (j = 0; j < ofs; j++) {                                       \
+            Vd->E1(j + ofs * 2 * i) = R_SHIFT(Vj->E2(j + ofs * i),        \
+                                              Vk->E3(j + ofs * i) % BIT); \
+        }                                                                 \
+        Vd->D(2 * i + 1) = 0;                                             \
+    }                                                                     \
 }
 
-VSRAN(vsran_b_h, 16, uint16_t, B, H)
-VSRAN(vsran_h_w, 32, uint32_t, H, W)
-VSRAN(vsran_w_d, 64, uint64_t, W, D)
+VSRAN(vsran_b_h, 16, B, H, UH)
+VSRAN(vsran_h_w, 32, H, W, UW)
+VSRAN(vsran_w_d, 64, W, D, UD)
 
-#define VSRLNI(NAME, BIT, T, E1, E2)                               \
-void HELPER(NAME)(void *vd, void *vj, uint64_t imm, uint32_t desc) \
-{                                                                  \
-    int i, max;                                                    \
-    VReg temp;                                                     \
-    VReg *Vd = (VReg *)vd;                                         \
-    VReg *Vj = (VReg *)vj;                                         \
-                                                                   \
-    temp.D(0) = 0;                                                 \
-    temp.D(1) = 0;                                                 \
-    max = LSX_LEN/BIT;                                             \
-    for (i = 0; i < max; i++) {                                    \
-        temp.E1(i) = R_SHIFT((T)Vj->E2(i), imm);                   \
-        temp.E1(i + max) = R_SHIFT((T)Vd->E2(i), imm);             \
-    }                                                              \
-    *Vd = temp;                                                    \
+#define VSRLNI(NAME, BIT, E1, E2)                                         \
+void HELPER(NAME)(void *vd, void *vj, uint64_t imm, uint32_t desc)        \
+{                                                                         \
+    int i, j, ofs;                                                        \
+    VReg temp = {};                                                       \
+    VReg *Vd = (VReg *)vd;                                                \
+    VReg *Vj = (VReg *)vj;                                                \
+    int oprsz = simd_oprsz(desc);                                         \
+                                                                          \
+    ofs = LSX_LEN / BIT;                                                  \
+    for (i = 0; i < oprsz / 16; i++) {                                    \
+        for (j = 0; j < ofs; j++) {                                       \
+            temp.E1(j + ofs * 2 * i) = R_SHIFT(Vj->E2(j + ofs * i), imm); \
+            temp.E1(j + ofs * (2 * i + 1)) = R_SHIFT(Vd->E2(j + ofs * i), \
+                                                     imm);                \
+        }                                                                 \
+    }                                                                     \
+    *Vd = temp;                                                           \
 }
 
 void HELPER(vsrlni_d_q)(void *vd, void *vj, uint64_t imm, uint32_t desc)
 {
-    VReg temp;
+    int i;
+    VReg temp = {};
     VReg *Vd = (VReg *)vd;
     VReg *Vj = (VReg *)vj;
 
-    temp.D(0) = 0;
-    temp.D(1) = 0;
-    temp.D(0) = int128_getlo(int128_urshift(Vj->Q(0), imm % 128));
-    temp.D(1) = int128_getlo(int128_urshift(Vd->Q(0), imm % 128));
+    for (i = 0; i < 2; i++) {
+        temp.D(2 * i) = int128_getlo(int128_urshift(Vj->Q(i), imm % 128));
+        temp.D(2 * i +1) = int128_getlo(int128_urshift(Vd->Q(i), imm % 128));
+    }
     *Vd = temp;
 }
 
-VSRLNI(vsrlni_b_h, 16, uint16_t, B, H)
-VSRLNI(vsrlni_h_w, 32, uint32_t, H, W)
-VSRLNI(vsrlni_w_d, 64, uint64_t, W, D)
+VSRLNI(vsrlni_b_h, 16, B, UH)
+VSRLNI(vsrlni_h_w, 32, H, UW)
+VSRLNI(vsrlni_w_d, 64, W, UD)
 
-#define VSRANI(NAME, BIT, E1, E2)                                  \
-void HELPER(NAME)(void *vd, void *vj, uint64_t imm, uint32_t desc) \
-{                                                                  \
-    int i, max;                                                    \
-    VReg temp;                                                     \
-    VReg *Vd = (VReg *)vd;                                         \
-    VReg *Vj = (VReg *)vj;                                         \
-                                                                   \
-    temp.D(0) = 0;                                                 \
-    temp.D(1) = 0;                                                 \
-    max = LSX_LEN/BIT;                                             \
-    for (i = 0; i < max; i++) {                                    \
-        temp.E1(i) = R_SHIFT(Vj->E2(i), imm);                      \
-        temp.E1(i + max) = R_SHIFT(Vd->E2(i), imm);                \
-    }                                                              \
-    *Vd = temp;                                                    \
+#define VSRANI(NAME, BIT, E1, E2)                                         \
+void HELPER(NAME)(void *vd, void *vj, uint64_t imm, uint32_t desc)        \
+{                                                                         \
+    int i, j, ofs;                                                        \
+    VReg temp = {};                                                       \
+    VReg *Vd = (VReg *)vd;                                                \
+    VReg *Vj = (VReg *)vj;                                                \
+    int oprsz = simd_oprsz(desc);                                         \
+                                                                          \
+    ofs = LSX_LEN / BIT;                                                  \
+    for (i = 0; i < oprsz / 16; i++) {                                    \
+        for (j = 0; j < ofs; j++) {                                       \
+            temp.E1(j + ofs * 2 * i) = R_SHIFT(Vj->E2(j + ofs * i), imm); \
+            temp.E1(j + ofs * (2 * i + 1)) = R_SHIFT(Vd->E2(j + ofs * i), \
+                                                     imm);                \
+        }                                                                 \
+    }                                                                     \
+    *Vd = temp;                                                           \
 }
 
 void HELPER(vsrani_d_q)(void *vd, void *vj, uint64_t imm, uint32_t desc)
 {
-    VReg temp;
+    int i;
+    VReg temp = {};
     VReg *Vd = (VReg *)vd;
     VReg *Vj = (VReg *)vj;
 
-    temp.D(0) = 0;
-    temp.D(1) = 0;
-    temp.D(0) = int128_getlo(int128_rshift(Vj->Q(0), imm % 128));
-    temp.D(1) = int128_getlo(int128_rshift(Vd->Q(0), imm % 128));
+    for (i = 0; i < 2; i++) {
+        temp.D(2 * i) = int128_getlo(int128_rshift(Vj->Q(i), imm % 128));
+        temp.D(2 * i + 1) = int128_getlo(int128_rshift(Vd->Q(i), imm % 128));
+    }
     *Vd = temp;
 }