[Qemu-devel] [PATCH v4 4/5] target/mips: Refactor and fix COPY_U.<B|H|W> instructions

[Mateja Marjanovic <mateja.marjanovic@rt-rk.com>]

From: Mateja Marjanovic <Mateja.Marjanovic@rt-rk.com>

The old version of the helper for the COPY_U.<B|H|W> MSA instructions
has been replaced with four helpers that don't use switch, and change
the endianness of the given index, when executed on a big endian host.

Signed-off-by: Mateja Marjanovic <mateja.marjanovic@rt-rk.com>
---
 target/mips/helper.h     |  4 +++-
 target/mips/msa_helper.c | 55 +++++++++++++++++++++++++++++++-----------------
 target/mips/translate.c  | 21 +++++++++++++++++-
 3 files changed, 59 insertions(+), 21 deletions(-)

diff --git a/target/mips/helper.h b/target/mips/helper.h
index 4e49618..8b6703c 100644
--- a/target/mips/helper.h
+++ b/target/mips/helper.h
@@ -875,7 +875,6 @@ DEF_HELPER_5(msa_hsub_u_df, void, env, i32, i32, i32, i32)
 DEF_HELPER_5(msa_sldi_df, void, env, i32, i32, i32, i32)
 DEF_HELPER_5(msa_splati_df, void, env, i32, i32, i32, i32)
 
-DEF_HELPER_5(msa_copy_u_df, void, env, i32, i32, i32, i32)
 DEF_HELPER_5(msa_insert_df, void, env, i32, i32, i32, i32)
 DEF_HELPER_5(msa_insve_df, void, env, i32, i32, i32, i32)
 DEF_HELPER_3(msa_ctcmsa, void, env, tl, i32)
@@ -940,6 +939,9 @@ DEF_HELPER_4(msa_copy_s_b, void, env, i32, i32, i32)
 DEF_HELPER_4(msa_copy_s_h, void, env, i32, i32, i32)
 DEF_HELPER_4(msa_copy_s_w, void, env, i32, i32, i32)
 DEF_HELPER_4(msa_copy_s_d, void, env, i32, i32, i32)
+DEF_HELPER_4(msa_copy_u_b, void, env, i32, i32, i32)
+DEF_HELPER_4(msa_copy_u_h, void, env, i32, i32, i32)
+DEF_HELPER_4(msa_copy_u_w, void, env, i32, i32, i32)
 
 DEF_HELPER_4(msa_fclass_df, void, env, i32, i32, i32)
 DEF_HELPER_4(msa_ftrunc_s_df, void, env, i32, i32, i32)
diff --git a/target/mips/msa_helper.c b/target/mips/msa_helper.c
index 5a06579..d5bf4dc 100644
--- a/target/mips/msa_helper.c
+++ b/target/mips/msa_helper.c
@@ -1281,29 +1281,46 @@ void helper_msa_copy_s_d(CPUMIPSState *env, uint32_t rd,
     env->active_tc.gpr[rd] = (int64_t)env->active_fpu.fpr[ws].wr.d[n];
 }
 
-void helper_msa_copy_u_df(CPUMIPSState *env, uint32_t df, uint32_t rd,
-                          uint32_t ws, uint32_t n)
+void helper_msa_copy_u_b(CPUMIPSState *env, uint32_t rd,
+                         uint32_t ws, uint32_t n)
 {
-    n %= DF_ELEMENTS(df);
+    n %= 16;
+#if defined(HOST_WORDS_BIGENDIAN)
+    if (n < 8) {
+        n = 8 - n - 1;
+    } else {
+        n = 24 - n - 1;
+    }
+#endif
+    env->active_tc.gpr[rd] = (uint8_t)env->active_fpu.fpr[ws].wr.b[n];
+}
 
-    switch (df) {
-    case DF_BYTE:
-        env->active_tc.gpr[rd] = (uint8_t)env->active_fpu.fpr[ws].wr.b[n];
-        break;
-    case DF_HALF:
-        env->active_tc.gpr[rd] = (uint16_t)env->active_fpu.fpr[ws].wr.h[n];
-        break;
-    case DF_WORD:
-        env->active_tc.gpr[rd] = (uint32_t)env->active_fpu.fpr[ws].wr.w[n];
-        break;
-#ifdef TARGET_MIPS64
-    case DF_DOUBLE:
-        env->active_tc.gpr[rd] = (uint64_t)env->active_fpu.fpr[ws].wr.d[n];
-        break;
+void helper_msa_copy_u_h(CPUMIPSState *env, uint32_t rd,
+                         uint32_t ws, uint32_t n)
+{
+    n %= 8;
+#if defined(HOST_WORDS_BIGENDIAN)
+    if (n < 4) {
+        n = 4 - n - 1;
+    } else {
+        n = 12 - n - 1;
+    }
 #endif
-    default:
-        assert(0);
+    env->active_tc.gpr[rd] = (uint16_t)env->active_fpu.fpr[ws].wr.h[n];
+}
+
+void helper_msa_copy_u_w(CPUMIPSState *env, uint32_t rd,
+                         uint32_t ws, uint32_t n)
+{
+    n %= 4;
+#if defined(HOST_WORDS_BIGENDIAN)
+    if (n < 2) {
+        n = 2 - n - 1;
+    } else {
+        n = 6 - n - 1;
     }
+#endif
+    env->active_tc.gpr[rd] = (uint32_t)env->active_fpu.fpr[ws].wr.w[n];
 }
 
 void helper_msa_insert_df(CPUMIPSState *env, uint32_t df, uint32_t wd,
diff --git a/target/mips/translate.c b/target/mips/translate.c
index f2ea378..72ed0a8 100644
--- a/target/mips/translate.c
+++ b/target/mips/translate.c
@@ -29397,6 +29397,11 @@ static void gen_msa_elm_df(CPUMIPSState *env, DisasContext *ctx, uint32_t df,
             generate_exception_end(ctx, EXCP_RI);
             break;
         }
+        if ((MASK_MSA_ELM(ctx->opcode) == OPC_COPY_U_df) &&
+              (df == DF_WORD)) {
+            generate_exception_end(ctx, EXCP_RI);
+            break;
+        }
 #endif
         switch (MASK_MSA_ELM(ctx->opcode)) {
         case OPC_COPY_S_df:
@@ -29423,7 +29428,21 @@ static void gen_msa_elm_df(CPUMIPSState *env, DisasContext *ctx, uint32_t df,
             break;
         case OPC_COPY_U_df:
             if (likely(wd != 0)) {
-                gen_helper_msa_copy_u_df(cpu_env, tdf, twd, tws, tn);
+                switch (df) {
+                case DF_BYTE:
+                    gen_helper_msa_copy_u_b(cpu_env, twd, tws, tn);
+                    break;
+                case DF_HALF:
+                    gen_helper_msa_copy_u_h(cpu_env, twd, tws, tn);
+                    break;
+#if defined(TARGET_MIPS64)
+                case DF_WORD:
+                    gen_helper_msa_copy_u_w(cpu_env, twd, tws, tn);
+                    break;
+#endif
+                default:
+                    assert(0);
+                }
             }
             break;
         case OPC_INSERT_df:
-- 
2.7.4