Never saw that one being trapped and could not reproduce on AMD.
But from the QEMU bug tracker, that looks like another story on
Intel processors.

Present in "whpx_failures.txt" on
https://gitlab.com/qemu-project/qemu/-/work_items/3349#note_3195249254

Signed-off-by: Mohamed Mediouni <mohamed@unpredictable.fr>
---
 target/i386/emulate/x86_emu.c | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/target/i386/emulate/x86_emu.c b/target/i386/emulate/x86_emu.c
index c2da1a133f..e6e6500ea5 100644
--- a/target/i386/emulate/x86_emu.c
+++ b/target/i386/emulate/x86_emu.c
@@ -1338,6 +1338,22 @@ static bool exec_xadd(CPUX86State *env, struct x86_decode *decode)
     return 0;
 }
 
+static bool exec_sldt(CPUX86State *env, struct x86_decode *decode)
+{
+    x86_segment_selector seg = emul_ops->read_segment_selector(env_cpu(env), R_LDTR);
+    write_val_ext(env, &decode->op[0], seg.sel, 2);
+    env->eip += decode->len;
+    return 0;
+}
+
+static bool exec_str(CPUX86State *env, struct x86_decode *decode)
+{
+    x86_segment_selector seg = emul_ops->read_segment_selector(env_cpu(env), R_TR);
+    write_val_ext(env, &decode->op[0], seg.sel, 2);
+    env->eip += decode->len;
+    return 0;
+}
+
 static struct cmd_handler {
     enum x86_decode_cmd cmd;
     bool (*handler)(CPUX86State *env, struct x86_decode *ins);
@@ -1382,6 +1398,8 @@ static struct cmd_handler {
     {X86_DECODE_CMD_MOVSX, exec_movsx},
     {X86_DECODE_CMD_XCHG, exec_xchg},
     {X86_DECODE_CMD_XADD, exec_xadd},
+    {X86_DECODE_CMD_SLDT, exec_sldt},
+    {X86_DECODE_CMD_STR, exec_str}
 };
 
 static struct cmd_handler _cmd_handler[X86_DECODE_CMD_LAST];
-- 
2.50.1 (Apple Git-155)