diff options
author | Dave Airlie | 2018-07-18 21:13:14 +0200 |
---|---|---|
committer | Dave Airlie | 2018-07-18 21:27:57 +0200 |
commit | 54c88a029a0a86fe00a0ee7d2a15ee08e6d04db9 (patch) | |
tree | 57c28532f31aa88b452bc2cf947dbe4e0b1f9ddb /drivers/gpu/drm/v3d/v3d_sched.c | |
parent | Merge tag 'vmwgfx-next-4.19-3' of git://people.freedesktop.org/~thomash/linux... (diff) | |
parent | drm/client: Fix: drm_client_new: Don't require DRM to be registered (diff) | |
download | kernel-qcow2-linux-54c88a029a0a86fe00a0ee7d2a15ee08e6d04db9.tar.gz kernel-qcow2-linux-54c88a029a0a86fe00a0ee7d2a15ee08e6d04db9.tar.xz kernel-qcow2-linux-54c88a029a0a86fe00a0ee7d2a15ee08e6d04db9.zip |
Merge tag 'drm-misc-next-2018-07-11' of git://anongit.freedesktop.org/drm/drm-misc into drm-next
drm-misc-next for 4.19:
Cross-subsystem Changes:
- many dt-bindings Doc changes
Core Changes:
- Encoder clean ups (Ville Syrjälä)
- Connector Writeback improvements(Boris Brezillon)
- Fake vblank support (Boris Brezillon)
- API for in-kernel clients (Noralf Trønnes)
- improvements to the path of finding panels(Boris Brezillon)
Driver Changes:
- initial support for the virtual display driver - vkms(Haneen Mohammed and Rodrigo Siqueira)
- panel: add Rocktech RK070ER9427 LCD support (Jagan Teki)
- panel: add support for the EDT ETM0700G0EDH6 and EDT ETM0700G0BDH6(Jan Tuerk)
- panel: add DLC DLC0700YZG-1 (Philipp Zabel)
- panel: add support for BOE HV070WSA-100 (Andrzej Hajda)
- panel: add newhaven, nhd-4.3-480272ef-atxl LCD (Tomi Valkeinen)
- panel: add support for Innolux G070Y2-L01 (Christoph Fritz)
- panel: add support for DataImage SCF0700C48GGU18 (Michal Vokáč)
- panel: add support for Sharp LQ035Q7DB03 (Vladimir Zapolskiy)
- panel: p079zca: Refactor panel driver to support multiple panels (Lin Huang)
- sun4i: Add R40 display engine compatible(Jernej Skrabec)
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180712011137.GA26620@juma
Diffstat (limited to 'drivers/gpu/drm/v3d/v3d_sched.c')
-rw-r--r-- | drivers/gpu/drm/v3d/v3d_sched.c | 22 |
1 files changed, 20 insertions, 2 deletions
diff --git a/drivers/gpu/drm/v3d/v3d_sched.c b/drivers/gpu/drm/v3d/v3d_sched.c index 808bc901f567..a5501581d96b 100644 --- a/drivers/gpu/drm/v3d/v3d_sched.c +++ b/drivers/gpu/drm/v3d/v3d_sched.c @@ -14,8 +14,8 @@ * to the HW only when it has completed the last one, instead of * filling up the CT[01]Q FIFOs with jobs. Similarly, we use * v3d_job_dependency() to manage the dependency between bin and - * render, instead of having the clients submit jobs with using the - * HW's semaphores to interlock between them. + * render, instead of having the clients submit jobs using the HW's + * semaphores to interlock between them. */ #include <linux/kthread.h> @@ -153,7 +153,25 @@ v3d_job_timedout(struct drm_sched_job *sched_job) struct v3d_job *job = to_v3d_job(sched_job); struct v3d_exec_info *exec = job->exec; struct v3d_dev *v3d = exec->v3d; + enum v3d_queue job_q = job == &exec->bin ? V3D_BIN : V3D_RENDER; enum v3d_queue q; + u32 ctca = V3D_CORE_READ(0, V3D_CLE_CTNCA(job_q)); + u32 ctra = V3D_CORE_READ(0, V3D_CLE_CTNRA(job_q)); + + /* If the current address or return address have changed, then + * the GPU has probably made progress and we should delay the + * reset. This could fail if the GPU got in an infinite loop + * in the CL, but that is pretty unlikely outside of an i-g-t + * testcase. + */ + if (job->timedout_ctca != ctca || job->timedout_ctra != ctra) { + job->timedout_ctca = ctca; + job->timedout_ctra = ctra; + + schedule_delayed_work(&job->base.work_tdr, + job->base.sched->timeout); + return; + } mutex_lock(&v3d->reset_lock); |