Hexagon (target/hexagon) Move pkt_has_store_s1 to DisasContext

The pkt_has_store_s1 field is only used for bookkeeping helpers with
a load.  With recent changes that eliminate the need to free TCGv
variables, it makes more sense to make this transient.

These helpers already take the instruction slot as an argument.  We
combine the slot and pkt_has_store_s1 into a single argument called
slotval.

Suggested-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20230427230012.3800327-21-tsimpson@quicinc.com>

11 files changed, 51 insertions, 40 deletions

diff --git a/target/hexagon/cpu.h b/target/hexagon/cpu.h
index 7673f9f32d..87e457dda9 100644
--- a/target/hexagon/cpu.h
+++ b/target/hexagon/cpu.h
@@ -99,7 +99,6 @@ typedef struct CPUArchState {
     target_ulong reg_written[TOTAL_PER_THREAD_REGS];
 
     MemLog mem_log_stores[STORES_MAX];
-    target_ulong pkt_has_store_s1;
     target_ulong dczero_addr;
 
     float_status fp_status;
diff --git a/target/hexagon/gen_analyze_funcs.py b/target/hexagon/gen_analyze_funcs.py
index 36da669450..d040f67001 100755
--- a/target/hexagon/gen_analyze_funcs.py
+++ b/target/hexagon/gen_analyze_funcs.py
@@ -209,8 +209,6 @@ def gen_analyze_func(f, tag, regs, imms):
     has_generated_helper = not hex_common.skip_qemu_helper(
         tag
     ) and not hex_common.is_idef_parser_enabled(tag)
-    if has_generated_helper and "A_SCALAR_LOAD" in hex_common.attribdict[tag]:
-        f.write("    ctx->need_pkt_has_store_s1 = true;\n")
 
     ## Mark HVX instructions with generated helpers
     if (has_generated_helper and
diff --git a/target/hexagon/gen_helper_funcs.py b/target/hexagon/gen_helper_funcs.py
index e259ea3d03..39751a483c 100755
--- a/target/hexagon/gen_helper_funcs.py
+++ b/target/hexagon/gen_helper_funcs.py
@@ -303,7 +303,7 @@ def gen_helper_function(f, tag, tagregs, tagimms):
         if hex_common.need_slot(tag):
             if i > 0:
                 f.write(", ")
-            f.write("uint32_t slot")
+            f.write("uint32_t slotval")
             i += 1
         if hex_common.need_part1(tag):
             if i > 0:
@@ -331,6 +331,11 @@ def gen_helper_function(f, tag, tagregs, tagimms):
                 else:
                     print("Bad register parse: ", regtype, regid, toss, numregs)
 
+        if hex_common.need_slot(tag):
+            if "A_LOAD" in hex_common.attribdict[tag]:
+                f.write("    bool pkt_has_store_s1 = slotval & 0x1;\n")
+            f.write("    uint32_t slot = slotval >> 1;\n")
+
         if "A_FPOP" in hex_common.attribdict[tag]:
             f.write("    arch_fpop_start(env);\n")
 
diff --git a/target/hexagon/gen_tcg_funcs.py b/target/hexagon/gen_tcg_funcs.py
index 0403547387..887b1cd369 100755
--- a/target/hexagon/gen_tcg_funcs.py
+++ b/target/hexagon/gen_tcg_funcs.py
@@ -556,7 +556,7 @@ def gen_tcg_func(f, tag, regs, imms):
         if hex_common.need_part1(tag):
             f.write("    TCGv part1 = tcg_constant_tl(insn->part1);\n")
         if hex_common.need_slot(tag):
-            f.write("    TCGv slot = tcg_constant_tl(insn->slot);\n")
+            f.write("    TCGv slotval = gen_slotval(ctx);\n")
         if hex_common.need_PC(tag):
             f.write("    TCGv PC = tcg_constant_tl(ctx->pkt->pc);\n")
         if hex_common.helper_needs_next_PC(tag):
@@ -606,7 +606,7 @@ def gen_tcg_func(f, tag, regs, imms):
         if hex_common.helper_needs_next_PC(tag):
             f.write(", next_PC")
         if hex_common.need_slot(tag):
-            f.write(", slot")
+            f.write(", slotval")
         if hex_common.need_part1(tag):
             f.write(", part1")
         f.write(");\n")
diff --git a/target/hexagon/genptr.c b/target/hexagon/genptr.c
index 785778759e..361cc789d7 100644
--- a/target/hexagon/genptr.c
+++ b/target/hexagon/genptr.c
@@ -398,6 +398,14 @@ static inline void gen_store_conditional8(DisasContext *ctx,
     tcg_gen_movi_tl(hex_llsc_addr, ~0);
 }
 
+#ifndef CONFIG_HEXAGON_IDEF_PARSER
+static TCGv gen_slotval(DisasContext *ctx)
+{
+    int slotval = (ctx->pkt->pkt_has_store_s1 & 1) | (ctx->insn->slot << 1);
+    return tcg_constant_tl(slotval);
+}
+#endif
+
 void gen_store32(TCGv vaddr, TCGv src, int width, uint32_t slot)
 {
     tcg_gen_mov_tl(hex_store_addr[slot], vaddr);
diff --git a/target/hexagon/hex_common.py b/target/hexagon/hex_common.py
index 29c0508f66..011cce1a68 100755
--- a/target/hexagon/hex_common.py
+++ b/target/hexagon/hex_common.py
@@ -247,9 +247,10 @@ def is_new_val(regtype, regid, tag):
 
 def need_slot(tag):
     if (
-        ("A_CONDEXEC" in attribdict[tag] and "A_JUMP" not in attribdict[tag])
-        or "A_STORE" in attribdict[tag]
-        or "A_LOAD" in attribdict[tag]
+        "A_CVI_SCATTER" not in attribdict[tag]
+        and "A_CVI_GATHER" not in attribdict[tag]
+        and ("A_STORE" in attribdict[tag]
+             or "A_LOAD" in attribdict[tag])
     ):
         return 1
     else:
diff --git a/target/hexagon/macros.h b/target/hexagon/macros.h
index 828874f318..5308c0848e 100644
--- a/target/hexagon/macros.h
+++ b/target/hexagon/macros.h
@@ -173,14 +173,14 @@
 #define MEM_STORE8(VA, DATA, SLOT) \
     MEM_STORE8_FUNC(DATA)(cpu_env, VA, DATA, SLOT)
 #else
-#define MEM_LOAD1s(VA) ((int8_t)mem_load1(env, slot, VA))
-#define MEM_LOAD1u(VA) ((uint8_t)mem_load1(env, slot, VA))
-#define MEM_LOAD2s(VA) ((int16_t)mem_load2(env, slot, VA))
-#define MEM_LOAD2u(VA) ((uint16_t)mem_load2(env, slot, VA))
-#define MEM_LOAD4s(VA) ((int32_t)mem_load4(env, slot, VA))
-#define MEM_LOAD4u(VA) ((uint32_t)mem_load4(env, slot, VA))
-#define MEM_LOAD8s(VA) ((int64_t)mem_load8(env, slot, VA))
-#define MEM_LOAD8u(VA) ((uint64_t)mem_load8(env, slot, VA))
+#define MEM_LOAD1s(VA) ((int8_t)mem_load1(env, pkt_has_store_s1, slot, VA))
+#define MEM_LOAD1u(VA) ((uint8_t)mem_load1(env, pkt_has_store_s1, slot, VA))
+#define MEM_LOAD2s(VA) ((int16_t)mem_load2(env, pkt_has_store_s1, slot, VA))
+#define MEM_LOAD2u(VA) ((uint16_t)mem_load2(env, pkt_has_store_s1, slot, VA))
+#define MEM_LOAD4s(VA) ((int32_t)mem_load4(env, pkt_has_store_s1, slot, VA))
+#define MEM_LOAD4u(VA) ((uint32_t)mem_load4(env, pkt_has_store_s1, slot, VA))
+#define MEM_LOAD8s(VA) ((int64_t)mem_load8(env, pkt_has_store_s1, slot, VA))
+#define MEM_LOAD8u(VA) ((uint64_t)mem_load8(env, pkt_has_store_s1, slot, VA))
 
 #define MEM_STORE1(VA, DATA, SLOT) log_store32(env, VA, DATA, 1, SLOT)
 #define MEM_STORE2(VA, DATA, SLOT) log_store32(env, VA, DATA, 2, SLOT)
diff --git a/target/hexagon/op_helper.c b/target/hexagon/op_helper.c
index f9021efc7e..dfabce3123 100644
--- a/target/hexagon/op_helper.c
+++ b/target/hexagon/op_helper.c
@@ -567,41 +567,45 @@ void HELPER(probe_pkt_scalar_hvx_stores)(CPUHexagonState *env, int mask)
  * If the load is in slot 0 and there is a store in slot1 (that
  * wasn't cancelled), we have to do the store first.
  */
-static void check_noshuf(CPUHexagonState *env, uint32_t slot,
-                         target_ulong vaddr, int size)
+static void check_noshuf(CPUHexagonState *env, bool pkt_has_store_s1,
+                         uint32_t slot, target_ulong vaddr, int size)
 {
-    if (slot == 0 && env->pkt_has_store_s1 &&
+    if (slot == 0 && pkt_has_store_s1 &&
         ((env->slot_cancelled & (1 << 1)) == 0)) {
         HELPER(probe_noshuf_load)(env, vaddr, size, MMU_USER_IDX);
         HELPER(commit_store)(env, 1);
     }
 }
 
-uint8_t mem_load1(CPUHexagonState *env, uint32_t slot, target_ulong vaddr)
+uint8_t mem_load1(CPUHexagonState *env, bool pkt_has_store_s1,
+                  uint32_t slot, target_ulong vaddr)
 {
     uintptr_t ra = GETPC();
-    check_noshuf(env, slot, vaddr, 1);
+    check_noshuf(env, pkt_has_store_s1, slot, vaddr, 1);
     return cpu_ldub_data_ra(env, vaddr, ra);
 }
 
-uint16_t mem_load2(CPUHexagonState *env, uint32_t slot, target_ulong vaddr)
+uint16_t mem_load2(CPUHexagonState *env, bool pkt_has_store_s1,
+                   uint32_t slot, target_ulong vaddr)
 {
     uintptr_t ra = GETPC();
-    check_noshuf(env, slot, vaddr, 2);
+    check_noshuf(env, pkt_has_store_s1, slot, vaddr, 2);
     return cpu_lduw_data_ra(env, vaddr, ra);
 }
 
-uint32_t mem_load4(CPUHexagonState *env, uint32_t slot, target_ulong vaddr)
+uint32_t mem_load4(CPUHexagonState *env, bool pkt_has_store_s1,
+                   uint32_t slot, target_ulong vaddr)
 {
     uintptr_t ra = GETPC();
-    check_noshuf(env, slot, vaddr, 4);
+    check_noshuf(env, pkt_has_store_s1, slot, vaddr, 4);
     return cpu_ldl_data_ra(env, vaddr, ra);
 }
 
-uint64_t mem_load8(CPUHexagonState *env, uint32_t slot, target_ulong vaddr)
+uint64_t mem_load8(CPUHexagonState *env, bool pkt_has_store_s1,
+                   uint32_t slot, target_ulong vaddr)
 {
     uintptr_t ra = GETPC();
-    check_noshuf(env, slot, vaddr, 8);
+    check_noshuf(env, pkt_has_store_s1, slot, vaddr, 8);
     return cpu_ldq_data_ra(env, vaddr, ra);
 }
 
diff --git a/target/hexagon/op_helper.h b/target/hexagon/op_helper.h
index 6bd4b07849..8f3764d15e 100644
--- a/target/hexagon/op_helper.h
+++ b/target/hexagon/op_helper.h
@@ -19,10 +19,14 @@
 #define HEXAGON_OP_HELPER_H
 
 /* Misc functions */
-uint8_t mem_load1(CPUHexagonState *env, uint32_t slot, target_ulong vaddr);
-uint16_t mem_load2(CPUHexagonState *env, uint32_t slot, target_ulong vaddr);
-uint32_t mem_load4(CPUHexagonState *env, uint32_t slot, target_ulong vaddr);
-uint64_t mem_load8(CPUHexagonState *env, uint32_t slot, target_ulong vaddr);
+uint8_t mem_load1(CPUHexagonState *env, bool pkt_has_store_s1,
+                  uint32_t slot, target_ulong vaddr);
+uint16_t mem_load2(CPUHexagonState *env, bool pkt_has_store_s1,
+                   uint32_t slot, target_ulong vaddr);
+uint32_t mem_load4(CPUHexagonState *env, bool pkt_has_store_s1,
+                   uint32_t slot, target_ulong vaddr);
+uint64_t mem_load8(CPUHexagonState *env, bool pkt_has_store_s1,
+                   uint32_t slot, target_ulong vaddr);
 
 void log_store64(CPUHexagonState *env, target_ulong addr,
                  int64_t val, int width, int slot);
diff --git a/target/hexagon/translate.c b/target/hexagon/translate.c
index a585cc8cfd..ec5abd2fd0 100644
--- a/target/hexagon/translate.c
+++ b/target/hexagon/translate.c
@@ -463,7 +463,6 @@ static void mark_implicit_pred_reads(DisasContext *ctx)
 static void analyze_packet(DisasContext *ctx)
 {
     Packet *pkt = ctx->pkt;
-    ctx->need_pkt_has_store_s1 = false;
     ctx->has_hvx_helper = false;
     for (int i = 0; i < pkt->num_insns; i++) {
         Insn *insn = &pkt->insn[i];
@@ -519,10 +518,6 @@ static void gen_start_packet(DisasContext *ctx)
 
     analyze_packet(ctx);
 
-    if (ctx->need_pkt_has_store_s1) {
-        tcg_gen_movi_tl(hex_pkt_has_store_s1, pkt->pkt_has_store_s1);
-    }
-
     /*
      * pregs_written is used both in the analyze phase as well as the code
      * gen phase, so clear it again.
@@ -1207,8 +1202,6 @@ void hexagon_translate_init(void)
         offsetof(CPUHexagonState, slot_cancelled), "slot_cancelled");
     hex_branch_taken = tcg_global_mem_new(cpu_env,
         offsetof(CPUHexagonState, branch_taken), "branch_taken");
-    hex_pkt_has_store_s1 = tcg_global_mem_new(cpu_env,
-        offsetof(CPUHexagonState, pkt_has_store_s1), "pkt_has_store_s1");
     hex_dczero_addr = tcg_global_mem_new(cpu_env,
         offsetof(CPUHexagonState, dczero_addr), "dczero_addr");
     hex_llsc_addr = tcg_global_mem_new(cpu_env,
diff --git a/target/hexagon/translate.h b/target/hexagon/translate.h
index a9f1ccee24..9697b4de0e 100644
--- a/target/hexagon/translate.h
+++ b/target/hexagon/translate.h
@@ -66,7 +66,6 @@ typedef struct DisasContext {
     TCGCond branch_cond;
     target_ulong branch_dest;
     bool is_tight_loop;
-    bool need_pkt_has_store_s1;
     bool short_circuit;
     bool has_hvx_helper;
     TCGv new_value[TOTAL_PER_THREAD_REGS];