[Qemu-devel] [PATCH v2] tcg/mips: Bugfix for crash when running program with qemu-i386.

[Qemu-devel] [PATCH v2] tcg/mips: Bugfix for crash when running program with qemu-i386.

[Jiang Biao]

When running a helloworld program with qemu-i386 in linux-user
mode on Loongson 3A3000, it will crash. The reasion is wrong
assigning from base to addr_regl directly. This patch fix the
bug.

Signed-off-by: Jiang Biao <jiang.biao2@zte.com.cn>
Signed-off-by: Richard Henderson <rth@twiddle.net>
---
 tcg/mips/tcg-target.inc.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/tcg/mips/tcg-target.inc.c b/tcg/mips/tcg-target.inc.c
index 8cff9a6..cebca8e 100644
--- a/tcg/mips/tcg-target.inc.c
+++ b/tcg/mips/tcg-target.inc.c
@@ -1539,16 +1539,18 @@ static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is_64)
                         s->code_ptr, label_ptr);
 #else
     if (TCG_TARGET_REG_BITS > TARGET_LONG_BITS) {
-        tcg_out_ext32u(s, base, addr_regl);
-        addr_regl = base;
+        tcg_out_ext32u(s, TCG_TMP0, addr_regl);
+    } else {
+        tcg_out_mov(s, TCG_TYPE_PTR, TCG_TMP0, addr_regl);
     }
+
     if (guest_base == 0 && data_regl != addr_regl) {
         base = addr_regl;
     } else if (guest_base == (int16_t)guest_base) {
-        tcg_out_opc_imm(s, ALIAS_PADDI, base, addr_regl, guest_base);
+        tcg_out_opc_imm(s, ALIAS_PADDI, base, TCG_TMP0, guest_base);
     } else {
         tcg_out_movi(s, TCG_TYPE_PTR, base, guest_base);
-        tcg_out_opc_reg(s, ALIAS_PADD, base, base, addr_regl);
+        tcg_out_opc_reg(s, ALIAS_PADD, base, base, TCG_TMP0);
     }
     tcg_out_qemu_ld_direct(s, data_regl, data_regh, base, opc, is_64);
 #endif
-- 
2.7.4

Re: [Qemu-devel] [PATCH v2] tcg/mips: Bugfix for crash when running program with qemu-i386.

[Richard Henderson]

On 07/09/2017 07:49 PM, Jiang Biao wrote:
> When running a helloworld program with qemu-i386 in linux-user
> mode on Loongson 3A3000, it will crash. The reasion is wrong
> assigning from base to addr_regl directly. This patch fix the
> bug.
> 
> Signed-off-by: Jiang Biao <jiang.biao2@zte.com.cn>
> Signed-off-by: Richard Henderson <rth@twiddle.net>
> ---
>   tcg/mips/tcg-target.inc.c | 10 ++++++----
>   1 file changed, 6 insertions(+), 4 deletions(-)
> 
> diff --git a/tcg/mips/tcg-target.inc.c b/tcg/mips/tcg-target.inc.c
> index 8cff9a6..cebca8e 100644
> --- a/tcg/mips/tcg-target.inc.c
> +++ b/tcg/mips/tcg-target.inc.c
> @@ -1539,16 +1539,18 @@ static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is_64)
>                           s->code_ptr, label_ptr);
>   #else
>       if (TCG_TARGET_REG_BITS > TARGET_LONG_BITS) {
> -        tcg_out_ext32u(s, base, addr_regl);
> -        addr_regl = base;
> +        tcg_out_ext32u(s, TCG_TMP0, addr_regl);
> +    } else {
> +        tcg_out_mov(s, TCG_TYPE_PTR, TCG_TMP0, addr_regl);
>       }

No, this is exactly why the assignment to the addr_regl variable exists, so 
that you do not need to emit an extra instruction for a 64-bit guest.


r~