drm/panfrost: Prevent concurrent resets

If a job times out in slot 0 while a reset is performed because a job
timed out in slot 1, the drm-sched core can get into a deadlock.

Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Signed-off-by: Rob Herring <robh@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20190418084305.45021-1-tomeu.vizoso@collabora.com

diff --git a/drivers/gpu/drm/panfrost/panfrost_device.c b/drivers/gpu/drm/panfrost/panfrost_device.c
index 91e8fb0f2b25b60e04289dea55ef07833c261d0b..970f669c6d29336ce513d2e1df12219030a3c17a 100644 (file)
--- a/drivers/gpu/drm/panfrost/panfrost_device.c
+++ b/drivers/gpu/drm/panfrost/panfrost_device.c
@@ -98,6 +98,7 @@ int panfrost_device_init(struct panfrost_device *pfdev)
        struct resource *res;
 
        mutex_init(&pfdev->sched_lock);
+       mutex_init(&pfdev->reset_lock);
        INIT_LIST_HEAD(&pfdev->scheduled_jobs);
 
        spin_lock_init(&pfdev->hwaccess_lock);

diff --git a/drivers/gpu/drm/panfrost/panfrost_device.h b/drivers/gpu/drm/panfrost/panfrost_device.h
index 1ba48d105763dccf2d331fffb874ecc14b7cdd10..56f452dfb4900ef586aacf16c78f7f5a7d64ade5 100644 (file)
--- a/drivers/gpu/drm/panfrost/panfrost_device.h
+++ b/drivers/gpu/drm/panfrost/panfrost_device.h
@@ -78,6 +78,7 @@ struct panfrost_device {
        struct list_head scheduled_jobs;
 
        struct mutex sched_lock;
+       struct mutex reset_lock;
 
        struct {
                struct devfreq *devfreq;

diff --git a/drivers/gpu/drm/panfrost/panfrost_job.c b/drivers/gpu/drm/panfrost/panfrost_job.c
index 0a7ed04f7d52b921915e382d76c7a23653143211..a5716c8fe8b39096df541b3bc4df23a5d292d380 100644 (file)
--- a/drivers/gpu/drm/panfrost/panfrost_job.c
+++ b/drivers/gpu/drm/panfrost/panfrost_job.c
@@ -384,6 +384,8 @@ static void panfrost_job_timedout(struct drm_sched_job *sched_job)
                job_read(pfdev, JS_TAIL_LO(js)),
                sched_job);
 
+       mutex_lock(&pfdev->reset_lock);
+
        for (i = 0; i < NUM_JOB_SLOTS; i++)
                drm_sched_stop(&pfdev->js->queue[i].sched);
 
@@ -406,6 +408,8 @@ static void panfrost_job_timedout(struct drm_sched_job *sched_job)
        /* restart scheduler after GPU is usable again */
        for (i = 0; i < NUM_JOB_SLOTS; i++)
                drm_sched_start(&pfdev->js->queue[i].sched, true);
+
+       mutex_unlock(&pfdev->reset_lock);
 }
 
 static const struct drm_sched_backend_ops panfrost_sched_ops = {