aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rwxr-xr-xconfigure1
-rw-r--r--target/xtensa/cpu.h3
-rw-r--r--target/xtensa/translate.c46
3 files changed, 35 insertions, 15 deletions
diff --git a/configure b/configure
index af72fc852e..c0cb9e6498 100755
--- a/configure
+++ b/configure
@@ -6840,6 +6840,7 @@ case "$target_name" in
;;
xtensa|xtensaeb)
TARGET_ARCH=xtensa
+ mttcg="yes"
;;
*)
error_exit "Unsupported target CPU"
diff --git a/target/xtensa/cpu.h b/target/xtensa/cpu.h
index 255cc9e08e..d9d3b33a70 100644
--- a/target/xtensa/cpu.h
+++ b/target/xtensa/cpu.h
@@ -31,6 +31,9 @@
#define ALIGNED_ONLY
#define TARGET_LONG_BITS 32
+/* Xtensa processors have a weak memory model */
+#define TCG_GUEST_DEFAULT_MO (0)
+
#define CPUArchState struct CPUXtensaState
#include "qemu-common.h"
diff --git a/target/xtensa/translate.c b/target/xtensa/translate.c
index 5969d7c3cd..c06d30d179 100644
--- a/target/xtensa/translate.c
+++ b/target/xtensa/translate.c
@@ -1664,9 +1664,15 @@ static void translate_ldst(DisasContext *dc, const uint32_t arg[],
gen_load_store_alignment(dc, par[0] & MO_SIZE, addr, par[1]);
}
if (par[2]) {
+ if (par[1]) {
+ tcg_gen_mb(TCG_BAR_STRL | TCG_MO_ALL);
+ }
tcg_gen_qemu_st_tl(cpu_R[arg[0]], addr, dc->cring, par[0]);
} else {
tcg_gen_qemu_ld_tl(cpu_R[arg[0]], addr, dc->cring, par[0]);
+ if (par[1]) {
+ tcg_gen_mb(TCG_BAR_LDAQ | TCG_MO_ALL);
+ }
}
tcg_temp_free(addr);
}
@@ -1823,6 +1829,12 @@ static void translate_mac16(DisasContext *dc, const uint32_t arg[],
}
}
+static void translate_memw(DisasContext *dc, const uint32_t arg[],
+ const uint32_t par[])
+{
+ tcg_gen_mb(TCG_BAR_SC | TCG_MO_ALL);
+}
+
static void translate_minmax(DisasContext *dc, const uint32_t arg[],
const uint32_t par[])
{
@@ -2193,29 +2205,33 @@ static void translate_setb_expstate(DisasContext *dc, const uint32_t arg[],
tcg_gen_ori_i32(cpu_UR[EXPSTATE], cpu_UR[EXPSTATE], 1u << arg[0]);
}
+#ifdef CONFIG_USER_ONLY
+static void gen_check_atomctl(DisasContext *dc, TCGv_i32 addr)
+{
+}
+#else
+static void gen_check_atomctl(DisasContext *dc, TCGv_i32 addr)
+{
+ TCGv_i32 tpc = tcg_const_i32(dc->pc);
+
+ gen_helper_check_atomctl(cpu_env, tpc, addr);
+ tcg_temp_free(tpc);
+}
+#endif
+
static void translate_s32c1i(DisasContext *dc, const uint32_t arg[],
const uint32_t par[])
{
if (gen_window_check2(dc, arg[0], arg[1])) {
- TCGLabel *label = gen_new_label();
TCGv_i32 tmp = tcg_temp_local_new_i32();
TCGv_i32 addr = tcg_temp_local_new_i32();
- TCGv_i32 tpc;
tcg_gen_mov_i32(tmp, cpu_R[arg[0]]);
tcg_gen_addi_i32(addr, cpu_R[arg[1]], arg[2]);
gen_load_store_alignment(dc, 2, addr, true);
-
- tpc = tcg_const_i32(dc->pc);
- gen_helper_check_atomctl(cpu_env, tpc, addr);
- tcg_gen_qemu_ld32u(cpu_R[arg[0]], addr, dc->cring);
- tcg_gen_brcond_i32(TCG_COND_NE, cpu_R[arg[0]],
- cpu_SR[SCOMPARE1], label);
-
- tcg_gen_qemu_st32(tmp, addr, dc->cring);
-
- gen_set_label(label);
- tcg_temp_free(tpc);
+ gen_check_atomctl(dc, addr);
+ tcg_gen_atomic_cmpxchg_i32(cpu_R[arg[0]], addr, cpu_SR[SCOMPARE1],
+ tmp, dc->cring, MO_32);
tcg_temp_free(addr);
tcg_temp_free(tmp);
}
@@ -2828,7 +2844,7 @@ static const XtensaOpcodeOps core_ops[] = {
.translate = translate_extui,
}, {
.name = "extw",
- .translate = translate_nop,
+ .translate = translate_memw,
}, {
.name = "hwwdtlba",
.translate = translate_ill,
@@ -2945,7 +2961,7 @@ static const XtensaOpcodeOps core_ops[] = {
.par = (const uint32_t[]){TCG_COND_GEU},
}, {
.name = "memw",
- .translate = translate_nop,
+ .translate = translate_memw,
}, {
.name = "min",
.translate = translate_minmax,