[PATCH] target/i386: Add support save/load exception error_code

[PATCH] target/i386: Add support save/load exception error_code

[Wang Xin via]

From: WangXin <wangxinxin.wang@huawei.com>

For now, qemu save/load CPU exception info(such as exception_nr and
has_error_code), while the exception error_code is ignored. This will
cause the dest hypervisor reinject a vCPU exception with error_code(0),
potentially causing a guest kernel panic.

For instance, if src VM stopped with an user-mode write #PF (error_code 6),
the dest hypervisor will reinject an #PF with error_code(0) when vCPU resume,
then guest kernel panic as:
  BUG: unable to handle page fault for address: 00007f80319cb010
  #PF: supervisor read access in user mode
  #PF: error_code(0x0000) - not-present page
  RIP: 0033:0x40115d

To fix it, support save/load exception error_code.

Signed-off-by: Xin Wang <wangxinxin.wang@huawei.com>
Signed-off-by: Tuo Xie <xietuo@huawei.com>

diff --git a/target/i386/machine.c b/target/i386/machine.c
index dd2dac1d44..45b7cea80a 100644
--- a/target/i386/machine.c
+++ b/target/i386/machine.c
@@ -462,6 +462,24 @@ static const VMStateDescription vmstate_exception_info = {
     }
 };
 
+static bool cpu_errcode_needed(void *opaque)
+{
+    X86CPU *cpu = opaque;
+
+    return cpu->env.has_error_code != 0;
+}
+
+static const VMStateDescription vmstate_error_code = {
+    .name = "cpu/error_code",
+    .version_id = 1,
+    .minimum_version_id = 1,
+    .needed = cpu_errcode_needed,
+    .fields = (const VMStateField[]) {
+        VMSTATE_INT32(env.error_code, X86CPU),
+        VMSTATE_END_OF_LIST()
+    }
+};
+
 /* Poll control MSR enabled by default */
 static bool poll_control_msr_needed(void *opaque)
 {
@@ -1746,6 +1764,7 @@ const VMStateDescription vmstate_x86_cpu = {
     },
     .subsections = (const VMStateDescription * const []) {
         &vmstate_exception_info,
+        &vmstate_error_code,
         &vmstate_async_pf_msr,
         &vmstate_async_pf_int_msr,
         &vmstate_pv_eoi_msr,
-- 
2.43.0

Re: [PATCH] target/i386: Add support save/load exception error_code

[Paolo Bonzini]

Queued, thanks.  But please let me know what's the correct signed off by chain;
either Tuo created the patch, and then you should have "--author" for him and
his SoB first; or you did, and then Tuo's SoB is unnecessary.

Paolo

RE: [PATCH] target/i386: Add support save/load exception error_code

[Wangxin (Alexander) via]

> 
> Queued, thanks.  But please let me know what's the correct signed off by chain;
> either Tuo created the patch, and then you should have "--author" for him and
> his SoB first; or you did, and then Tuo's SoB is unnecessary.
> 
> Paolo
> 

The author is me, should I resend this patch?

Thanks,
Xin

Re: [PATCH] target/i386: Add support save/load exception error_code

[Zhao Liu]

On Tue, Aug 19, 2025 at 10:58:34PM +0800, Wang Xin wrote:
> Date: Tue, 19 Aug 2025 22:58:34 +0800
> From: Wang Xin <wangxinxin.wang@huawei.com>
> Subject: [PATCH] target/i386: Add support save/load exception error_code
> X-Mailer: git-send-email 2.50.1.windows.1
> 
> From: WangXin <wangxinxin.wang@huawei.com>
> 
> For now, qemu save/load CPU exception info(such as exception_nr and
> has_error_code), while the exception error_code is ignored. This will
> cause the dest hypervisor reinject a vCPU exception with error_code(0),
> potentially causing a guest kernel panic.
> 
> For instance, if src VM stopped with an user-mode write #PF (error_code 6),
> the dest hypervisor will reinject an #PF with error_code(0) when vCPU resume,
> then guest kernel panic as:
>   BUG: unable to handle page fault for address: 00007f80319cb010
>   #PF: supervisor read access in user mode
>   #PF: error_code(0x0000) - not-present page
>   RIP: 0033:0x40115d
> 
> To fix it, support save/load exception error_code.
> 
> Signed-off-by: Xin Wang <wangxinxin.wang@huawei.com>
> Signed-off-by: Tuo Xie <xietuo@huawei.com>

Reviewed-by: Zhao Liu <zhao1.liu@intel.com>

Re: [PATCH] target/i386: Add support save/load exception error_code

[zhao1.liu]

On Wed, Aug 20, 2025 at 02:18:21AM +0000, Wangxin (Alexander) wrote:
> Date: Wed, 20 Aug 2025 02:18:21 +0000
> From: "Wangxin (Alexander)" <wangxinxin.wang@huawei.com>
> Subject: RE: [PATCH] target/i386: Add support save/load exception error_code
> 
> > 
> > Queued, thanks.  But please let me know what's the correct signed off by chain;
> > either Tuo created the patch, and then you should have "--author" for him and
> > his SoB first; or you did, and then Tuo's SoB is unnecessary.
> > 
> > Paolo
> > 
> 
> The author is me,

But if Tuo co-worked with you, his SoB is also necessary to keep:

"git commits will usually be expected to have a ``Signed-off-by``
 line for each contributor involved in creation of the patch."

In this case, “Co-developed-by” is not necessary for now.

I think the reason of confusion is, as the person sending the patch,
you'd have put your SoB on the last line.

Thanks,
Zhao