I noticed a rather surprising issue here while working on removing all of the non-atomic MST code: drm_atomic_get_mst_topology_state() doesn't check the return value of drm_atomic_get_private_obj_state() and instead just passes it directly to to_dp_mst_topology_state(). This means that if we hit a deadlock or something else which would return an error code pointer, we'll likely segfault the kernel.
This is definitely another one of those fixes where I'm astonished we somehow managed never to discover this issue until now…
Signed-off-by: Lyude Paul lyude@redhat.com Fixes: a4370c777406 ("drm/atomic: Make private objs proper objects") Cc: Ville Syrjälä ville.syrjala@linux.intel.com Cc: stable@vger.kernel.org # v4.14+ --- drivers/gpu/drm/display/drm_dp_mst_topology.c | 2 +- include/drm/display/drm_dp_mst_helper.h | 2 ++ 2 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/display/drm_dp_mst_topology.c b/drivers/gpu/drm/display/drm_dp_mst_topology.c index d84673b3294b..d6e595b95f07 100644 --- a/drivers/gpu/drm/display/drm_dp_mst_topology.c +++ b/drivers/gpu/drm/display/drm_dp_mst_topology.c @@ -5468,7 +5468,7 @@ EXPORT_SYMBOL(drm_dp_mst_topology_state_funcs); struct drm_dp_mst_topology_state *drm_atomic_get_mst_topology_state(struct drm_atomic_state *state, struct drm_dp_mst_topology_mgr *mgr) { - return to_dp_mst_topology_state(drm_atomic_get_private_obj_state(state, &mgr->base)); + return to_dp_mst_topology_state_safe(drm_atomic_get_private_obj_state(state, &mgr->base)); } EXPORT_SYMBOL(drm_atomic_get_mst_topology_state);
diff --git a/include/drm/display/drm_dp_mst_helper.h b/include/drm/display/drm_dp_mst_helper.h index 10adec068b7f..fe7577e7f305 100644 --- a/include/drm/display/drm_dp_mst_helper.h +++ b/include/drm/display/drm_dp_mst_helper.h @@ -541,6 +541,8 @@ struct drm_dp_payload { };
#define to_dp_mst_topology_state(x) container_of(x, struct drm_dp_mst_topology_state, base) +#define to_dp_mst_topology_state_safe(x) \ + container_of_safe(x, struct drm_dp_mst_topology_state, base)
struct drm_dp_vcpi_allocation { struct drm_dp_mst_port *port;
On Thu, Jun 02, 2022 at 04:17:56PM -0400, Lyude Paul wrote:
I noticed a rather surprising issue here while working on removing all of the non-atomic MST code: drm_atomic_get_mst_topology_state() doesn't check the return value of drm_atomic_get_private_obj_state() and instead just passes it directly to to_dp_mst_topology_state(). This means that if we hit a deadlock or something else which would return an error code pointer, we'll likely segfault the kernel.
This is definitely another one of those fixes where I'm astonished we somehow managed never to discover this issue until now…
It has been discussed before.
struct drm_dp_mst_topology_state { struct drm_private_state base; ... }
so offsetof(base)==0.
Signed-off-by: Lyude Paul lyude@redhat.com Fixes: a4370c777406 ("drm/atomic: Make private objs proper objects") Cc: Ville Syrjälä ville.syrjala@linux.intel.com Cc: stable@vger.kernel.org # v4.14+
drivers/gpu/drm/display/drm_dp_mst_topology.c | 2 +- include/drm/display/drm_dp_mst_helper.h | 2 ++ 2 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/display/drm_dp_mst_topology.c b/drivers/gpu/drm/display/drm_dp_mst_topology.c index d84673b3294b..d6e595b95f07 100644 --- a/drivers/gpu/drm/display/drm_dp_mst_topology.c +++ b/drivers/gpu/drm/display/drm_dp_mst_topology.c @@ -5468,7 +5468,7 @@ EXPORT_SYMBOL(drm_dp_mst_topology_state_funcs); struct drm_dp_mst_topology_state *drm_atomic_get_mst_topology_state(struct drm_atomic_state *state, struct drm_dp_mst_topology_mgr *mgr) {
- return to_dp_mst_topology_state(drm_atomic_get_private_obj_state(state, &mgr->base));
- return to_dp_mst_topology_state_safe(drm_atomic_get_private_obj_state(state, &mgr->base));
} EXPORT_SYMBOL(drm_atomic_get_mst_topology_state); diff --git a/include/drm/display/drm_dp_mst_helper.h b/include/drm/display/drm_dp_mst_helper.h index 10adec068b7f..fe7577e7f305 100644 --- a/include/drm/display/drm_dp_mst_helper.h +++ b/include/drm/display/drm_dp_mst_helper.h @@ -541,6 +541,8 @@ struct drm_dp_payload { }; #define to_dp_mst_topology_state(x) container_of(x, struct drm_dp_mst_topology_state, base) +#define to_dp_mst_topology_state_safe(x) \
- container_of_safe(x, struct drm_dp_mst_topology_state, base)
Wasn't aware of container_of_safe(). I suppose no real harm in using it. Not sure why we'd even keep the non-safe version around?
Though the use of container_of_safe() everywhere won't help when "casting" the other way (&foo->base, when foo==NULL/errptr). In order to make that work for non-zero offsets we'd have to introduce a casting macro for that direction as well.
struct drm_dp_vcpi_allocation { struct drm_dp_mst_port *port; -- 2.35.3
On Thu, 2022-06-02 at 23:42 +0300, Ville Syrjälä wrote:
On Thu, Jun 02, 2022 at 04:17:56PM -0400, Lyude Paul wrote:
I noticed a rather surprising issue here while working on removing all of the non-atomic MST code: drm_atomic_get_mst_topology_state() doesn't check the return value of drm_atomic_get_private_obj_state() and instead just passes it directly to to_dp_mst_topology_state(). This means that if we hit a deadlock or something else which would return an error code pointer, we'll likely segfault the kernel.
This is definitely another one of those fixes where I'm astonished we somehow managed never to discover this issue until now…
It has been discussed before.
struct drm_dp_mst_topology_state { struct drm_private_state base; ... }
so offsetof(base)==0.
fhjdffds yes you're right, I guess we can just drop this patch then
Signed-off-by: Lyude Paul lyude@redhat.com Fixes: a4370c777406 ("drm/atomic: Make private objs proper objects") Cc: Ville Syrjälä ville.syrjala@linux.intel.com Cc: stable@vger.kernel.org # v4.14+
drivers/gpu/drm/display/drm_dp_mst_topology.c | 2 +- include/drm/display/drm_dp_mst_helper.h | 2 ++ 2 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/display/drm_dp_mst_topology.c b/drivers/gpu/drm/display/drm_dp_mst_topology.c index d84673b3294b..d6e595b95f07 100644 --- a/drivers/gpu/drm/display/drm_dp_mst_topology.c +++ b/drivers/gpu/drm/display/drm_dp_mst_topology.c @@ -5468,7 +5468,7 @@ EXPORT_SYMBOL(drm_dp_mst_topology_state_funcs); struct drm_dp_mst_topology_state *drm_atomic_get_mst_topology_state(struct drm_atomic_state *state, struct drm_dp_mst_topology_mgr *mgr) { - return to_dp_mst_topology_state(drm_atomic_get_private_obj_state(state, &mgr-
base));
+ return to_dp_mst_topology_state_safe(drm_atomic_get_private_obj_state(state, &mgr->base)); } EXPORT_SYMBOL(drm_atomic_get_mst_topology_state); diff --git a/include/drm/display/drm_dp_mst_helper.h b/include/drm/display/drm_dp_mst_helper.h index 10adec068b7f..fe7577e7f305 100644 --- a/include/drm/display/drm_dp_mst_helper.h +++ b/include/drm/display/drm_dp_mst_helper.h @@ -541,6 +541,8 @@ struct drm_dp_payload { }; #define to_dp_mst_topology_state(x) container_of(x, struct drm_dp_mst_topology_state, base) +#define to_dp_mst_topology_state_safe(x) \ + container_of_safe(x, struct drm_dp_mst_topology_state, base)
Wasn't aware of container_of_safe(). I suppose no real harm in using it. Not sure why we'd even keep the non-safe version around?
Though the use of container_of_safe() everywhere won't help when "casting" the other way (&foo->base, when foo==NULL/errptr). In order to make that work for non-zero offsets we'd have to introduce a casting macro for that direction as well.
struct drm_dp_vcpi_allocation { struct drm_dp_mst_port *port; -- 2.35.3
linux-stable-mirror@lists.linaro.org
-
Lyude Paul -
Ville Syrjälä