drm/amd/display: Fix deadlock when flushing irq
authorMikita Lipski <mikita.lipski@amd.com>
Wed, 10 Jan 2018 15:01:38 +0000 (10:01 -0500)
committerAlex Deucher <alexander.deucher@amd.com>
Tue, 15 May 2018 18:44:06 +0000 (13:44 -0500)
Lock irq table when reading a work in queue,
unlock to flush the work, lock again till all tasks
are cleared

Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Harry Wentland <Harry.Wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c

index 490017d..4be21bf 100644 (file)
@@ -329,14 +329,15 @@ void amdgpu_dm_irq_fini(struct amdgpu_device *adev)
 {
        int src;
        struct irq_list_head *lh;
+       unsigned long irq_table_flags;
        DRM_DEBUG_KMS("DM_IRQ: releasing resources.\n");
-
        for (src = 0; src < DAL_IRQ_SOURCES_NUMBER; src++) {
-
+               DM_IRQ_TABLE_LOCK(adev, irq_table_flags);
                /* The handler was removed from the table,
                 * it means it is safe to flush all the 'work'
                 * (because no code can schedule a new one). */
                lh = &adev->dm.irq_handler_list_low_tab[src];
+               DM_IRQ_TABLE_UNLOCK(adev, irq_table_flags);
                flush_work(&lh->work);
        }
 }