[PATCH] vfio: prevent infinite loop in vfio_mig_get_next_state() on blocked arc

[PATCH] vfio: prevent infinite loop in vfio_mig_get_next_state() on blocked arc

[Junrui Luo]

vfio_mig_get_next_state() walks vfio_from_fsm_table[] one step at a time,
looping to skip optional states the device does not support until
*next_fsm is supported. A blocked transition is encoded as
VFIO_DEVICE_STATE_ERROR, which the trailing return reports as -EINVAL.

The skip loop does not account for the ERROR sentinel.
state_flags_table[ERROR] is ~0U and vfio_from_fsm_table[ERROR][*] is
ERROR, so once *next_fsm becomes ERROR the loop condition stays true and
*next_fsm never changes. The blocked arcs STOP_COPY -> PRE_COPY and
STOP_COPY -> PRE_COPY_P2P map to ERROR yet pass the support check on a
precopy-capable device, causing the loop to spin forever while holding
the driver state mutex. This can result in a soft lockup, and a panic
with softlockup_panic set.

Terminate the skip loop on the ERROR sentinel so a blocked transition
falls through to the existing return and reports -EINVAL.

Fixes: 4db52602a607 ("vfio: Extend the device migration protocol with PRE_COPY")
Reported-by: Yuhao Jiang <danisjiang@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Junrui Luo <moonafterrain@outlook.com>
---
 drivers/vfio/vfio_main.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c
index 6222376ab6ab..5e0422014523 100644
--- a/drivers/vfio/vfio_main.c
+++ b/drivers/vfio/vfio_main.c
@@ -858,7 +858,8 @@ int vfio_mig_get_next_state(struct vfio_device *device,
 	 * logical state, as per the above comment.
 	 */
 	*next_fsm = vfio_from_fsm_table[cur_fsm][new_fsm];
-	while ((state_flags_table[*next_fsm] & device->migration_flags) !=
+	while (*next_fsm != VFIO_DEVICE_STATE_ERROR &&
+	       (state_flags_table[*next_fsm] & device->migration_flags) !=
 			state_flags_table[*next_fsm])
 		*next_fsm = vfio_from_fsm_table[*next_fsm][new_fsm];
 

---
base-commit: e43ffb69e0438cddd72aaa30898b4dc446f664f8
change-id: 20260602-fixes-f5c6a4880594

Best regards,
-- 
Junrui Luo <moonafterrain@outlook.com>

Re: [PATCH] vfio: prevent infinite loop in vfio_mig_get_next_state() on blocked arc

[Jason Gunthorpe]

On Tue, Jun 02, 2026 at 04:58:48PM +0800, Junrui Luo wrote:
> vfio_mig_get_next_state() walks vfio_from_fsm_table[] one step at a time,
> looping to skip optional states the device does not support until
> *next_fsm is supported. A blocked transition is encoded as
> VFIO_DEVICE_STATE_ERROR, which the trailing return reports as -EINVAL.
> 
> The skip loop does not account for the ERROR sentinel.
> state_flags_table[ERROR] is ~0U and vfio_from_fsm_table[ERROR][*] is
> ERROR, so once *next_fsm becomes ERROR the loop condition stays true and
> *next_fsm never changes. The blocked arcs STOP_COPY -> PRE_COPY and
> STOP_COPY -> PRE_COPY_P2P map to ERROR yet pass the support check on a
> precopy-capable device, causing the loop to spin forever while holding
> the driver state mutex. This can result in a soft lockup, and a panic
> with softlockup_panic set.
> 
> Terminate the skip loop on the ERROR sentinel so a blocked transition
> falls through to the existing return and reports -EINVAL.
> 
> Fixes: 4db52602a607 ("vfio: Extend the device migration protocol with PRE_COPY")
> Reported-by: Yuhao Jiang <danisjiang@gmail.com>
> Cc: stable@vger.kernel.org
> Signed-off-by: Junrui Luo <moonafterrain@outlook.com>
> ---
>  drivers/vfio/vfio_main.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)

Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>

Jason

RE: [PATCH] vfio: prevent infinite loop in vfio_mig_get_next_state() on blocked arc

[Tian, Kevin]

> From: Junrui Luo <moonafterrain@outlook.com>
> Sent: Tuesday, June 2, 2026 4:59 PM
> 
> vfio_mig_get_next_state() walks vfio_from_fsm_table[] one step at a time,
> looping to skip optional states the device does not support until
> *next_fsm is supported. A blocked transition is encoded as
> VFIO_DEVICE_STATE_ERROR, which the trailing return reports as -EINVAL.
> 
> The skip loop does not account for the ERROR sentinel.
> state_flags_table[ERROR] is ~0U and vfio_from_fsm_table[ERROR][*] is
> ERROR, so once *next_fsm becomes ERROR the loop condition stays true and
> *next_fsm never changes. The blocked arcs STOP_COPY -> PRE_COPY and
> STOP_COPY -> PRE_COPY_P2P map to ERROR yet pass the support check on a
> precopy-capable device, causing the loop to spin forever while holding
> the driver state mutex. This can result in a soft lockup, and a panic
> with softlockup_panic set.
> 
> Terminate the skip loop on the ERROR sentinel so a blocked transition
> falls through to the existing return and reports -EINVAL.
> 
> Fixes: 4db52602a607 ("vfio: Extend the device migration protocol with
> PRE_COPY")
> Reported-by: Yuhao Jiang <danisjiang@gmail.com>
> Cc: stable@vger.kernel.org
> Signed-off-by: Junrui Luo <moonafterrain@outlook.com>

Reviewed-by: Kevin Tian <kevin.tian@intel.com>

Re: [PATCH] vfio: prevent infinite loop in vfio_mig_get_next_state() on blocked arc

[Alex Williamson]

On Tue, 02 Jun 2026 16:58:48 +0800
Junrui Luo <moonafterrain@outlook.com> wrote:

> vfio_mig_get_next_state() walks vfio_from_fsm_table[] one step at a time,
> looping to skip optional states the device does not support until
> *next_fsm is supported. A blocked transition is encoded as
> VFIO_DEVICE_STATE_ERROR, which the trailing return reports as -EINVAL.
> 
> The skip loop does not account for the ERROR sentinel.
> state_flags_table[ERROR] is ~0U and vfio_from_fsm_table[ERROR][*] is
> ERROR, so once *next_fsm becomes ERROR the loop condition stays true and
> *next_fsm never changes. The blocked arcs STOP_COPY -> PRE_COPY and
> STOP_COPY -> PRE_COPY_P2P map to ERROR yet pass the support check on a
> precopy-capable device, causing the loop to spin forever while holding
> the driver state mutex. This can result in a soft lockup, and a panic
> with softlockup_panic set.
> 
> Terminate the skip loop on the ERROR sentinel so a blocked transition
> falls through to the existing return and reports -EINVAL.
> 
> Fixes: 4db52602a607 ("vfio: Extend the device migration protocol with PRE_COPY")
> Reported-by: Yuhao Jiang <danisjiang@gmail.com>
> Cc: stable@vger.kernel.org
> Signed-off-by: Junrui Luo <moonafterrain@outlook.com>
> ---
>  drivers/vfio/vfio_main.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c
> index 6222376ab6ab..5e0422014523 100644
> --- a/drivers/vfio/vfio_main.c
> +++ b/drivers/vfio/vfio_main.c
> @@ -858,7 +858,8 @@ int vfio_mig_get_next_state(struct vfio_device *device,
>  	 * logical state, as per the above comment.
>  	 */
>  	*next_fsm = vfio_from_fsm_table[cur_fsm][new_fsm];
> -	while ((state_flags_table[*next_fsm] & device->migration_flags) !=
> +	while (*next_fsm != VFIO_DEVICE_STATE_ERROR &&
> +	       (state_flags_table[*next_fsm] & device->migration_flags) !=
>  			state_flags_table[*next_fsm])
>  		*next_fsm = vfio_from_fsm_table[*next_fsm][new_fsm];
>  

Applied to vfio next branch for v7.2.  Thanks,

Alex