target/xtensa: fix access ring in l32ex

l32ex does memory access as all regular load/store operations at CRING
level. Fix apparent pasto from l32e that caused it to use RING instead.

This is a correctness issue, not a security issue, because in the worst
case the privilege level of memory access may be lowered, resulting in
an exception when the correct implementation would've succeeded.
In no case it would allow memory access that would've raised an
exception in the correct implementation.

Cc: qemu-stable@nongnu.org
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
(cherry picked from commit 735aa900e4bf57b777ac620bed7c88234ec4b601)
Signed-off-by: Michael Roth <michael.roth@amd.com>

diff --git a/target/xtensa/translate.c b/target/xtensa/translate.c
index 0ae4efc48a1750fe80bc95aa13a35012d90e94bd..1678b65607a7385182d063624def80735f2cfc9d 100644 (file)
--- a/target/xtensa/translate.c
+++ b/target/xtensa/translate.c
@@ -1817,7 +1817,7 @@ static void translate_l32ex(DisasContext *dc, const OpcodeArg arg[],
     tcg_gen_mov_i32(addr, arg[1].in);
     gen_load_store_alignment(dc, 2, addr, true);
     gen_check_exclusive(dc, addr, false);
-    tcg_gen_qemu_ld_i32(arg[0].out, addr, dc->ring, MO_TEUL);
+    tcg_gen_qemu_ld_i32(arg[0].out, addr, dc->cring, MO_TEUL);
     tcg_gen_mov_i32(cpu_exclusive_addr, addr);
     tcg_gen_mov_i32(cpu_exclusive_val, arg[0].out);
     tcg_temp_free(addr);