[PATCH] drm/radeon: fix virtual memory locking in case of reset
Alex Deucher
alexdeucher at gmail.com
Thu Aug 2 09:27:33 PDT 2012
On Thu, Aug 2, 2012 at 12:05 PM, <j.glisse at gmail.com> wrote:
> From: Jerome Glisse <jglisse at redhat.com>
>
> Lock/unlock mutex in proper order to avoid deadlock in case
> of GPU reset triggered from VM code path.
>
> Cc: stable at vger.kernel.org [3.5]
> Signed-off-by: Jerome Glisse <jglisse at redhat.com>
Reviewed-by: Alex Deucher <alexander.deucher at amd.com>
> ---
> drivers/gpu/drm/radeon/radeon_gart.c | 11 +++++++++--
> 1 file changed, 9 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c
> index b372005..7eabb59 100644
> --- a/drivers/gpu/drm/radeon/radeon_gart.c
> +++ b/drivers/gpu/drm/radeon/radeon_gart.c
> @@ -508,14 +508,19 @@ static void radeon_vm_unbind_locked(struct radeon_device *rdev,
> while (vm->fence) {
> int r;
> r = radeon_fence_wait(vm->fence, false);
> - if (r)
> + if (r) {
> DRM_ERROR("error while waiting for fence: %d\n", r);
> + }
> if (r == -EDEADLK) {
> + /* release mutex and lock in right order */
> mutex_unlock(&rdev->vm_manager.lock);
> + mutex_unlock(&vm->mutex);
> r = radeon_gpu_reset(rdev);
> mutex_lock(&rdev->vm_manager.lock);
> - if (!r)
> + mutex_lock(&vm->mutex);
> + if (!r) {
> continue;
> + }
> }
> break;
> }
> @@ -551,7 +556,9 @@ void radeon_vm_manager_fini(struct radeon_device *rdev)
> mutex_lock(&rdev->vm_manager.lock);
> /* unbind all active vm */
> list_for_each_entry_safe(vm, tmp, &rdev->vm_manager.lru_vm, list) {
> + mutex_lock(&vm->mutex);
> radeon_vm_unbind_locked(rdev, vm);
> + mutex_unlock(&vm->mutex);
> }
> rdev->vm_manager.funcs->fini(rdev);
> mutex_unlock(&rdev->vm_manager.lock);
> --
> 1.7.10.4
>
> _______________________________________________
> dri-devel mailing list
> dri-devel at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/dri-devel
More information about the dri-devel
mailing list