1,TDR will kickout guilty job if it hang exceed the threshold
of the given one from kernel paramter "job_hang_limit", that
way a bad command stream will not infinitly cause GPU hang.
by default this threshold is 1 so a job will be kicked out
after it hang.
2,if a job timeout TDR routine will not reset all sched/ring,
instead if will only reset on the givn one which is indicated
by @job of amdgpu_sriov_gpu_reset, that way we don't need to
reset and recover each sched/ring if we already know which job
cause GPU hang.
3,unblock sriov_gpu_reset for AI family.
V2:
1:put kickout guilty job after sched parked.
2:since parking scheduler prior to kickout already occupies a
while, we can do last check on the in question job before
doing hw_reset.
TODO:
1:when a job is considered as guilty, we should mark some flag
in its fence status flag, and let UMD side aware that this
fence signaling is not due to job complete but job hang.
2:if gpu reset cause all video memory lost, we need introduce
a new policy to implement TDR, like drop all jobs not yet
signaled, and all IOCTL on this device will return ERROR
DEVICE_LOST.
this will be implemented later.
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
because we don't want to do sriov-gpu-reset under certain
cases, so just split those two funtion and don't invoke
sr-iov one from bare-metal one.
V2:
remove debugfs_gpu_reset routine on SRIOV case.
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
KIQ is used for interaction between driver and
CP, and not exposed to outside client, as such it
doesn't need to be handled by GPU scheduler.
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Signed-off-by: Xiangliang Yu <Xiangliang.Yu@amd.com>
Signed-off-by: Trigger Huang <trigger.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
-----BEGIN PGP SIGNATURE-----
iQEcBAABAgAGBQJYHmoCAAoJEHm+PkMAQRiG7RMIAI2i7Y5hpL5yCxK5AFaL4u/G
KxXfp1B1UanUTgjOmd7zGqtDYcFX9t7GTTUFixQ7/9Opr4PD9qbnatoDGSc3xjbT
msDgA1B78F1/Q3kHWfeGq32MihQ4mj5NwUCo+igUcUvvWG7mHgzErj/Nh5RoobQX
p/izdpTbrw3GX6xXB8olbG7XWHaVye/+TT3q6+gmgm8I/QEujcLeGoycE0zlhPN8
FG/JX76At/+ZM2Py7Oxo3k+oKL9CHrtOQYDp/wN0uslV5eYvvkZz0/M1HMOGZt+c
gZU5jzM17K7C4Nzo06WAuBU9wUBGc25m+cPicLlOmljnzfU+f50SKaDjZq3p7QI=
=2KUF
-----END PGP SIGNATURE-----
Backmerge tag 'v4.9-rc4' into drm-next
Linux 4.9-rc4
This is needed for nouveau development.
To free fences, call_rcu() is used, which calls amdgpu_fence_free()
after a grace period. During teardown, there is no guarantee all
callbacks have finished, so amdgpu_fence_slab may be destroyed before
all fences have been freed. If we are lucky, this results in some slab
warnings, if not, we get a crash in one of rcu threads because callback
is called after amdgpu has already been unloaded.
Fix it with a rcu_barrier().
Fixes: b44135351a ("drm/amdgpu: RCU protected amdgpu_fence_release")
Acked-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Grazvydas Ignotas <notasas@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Right now it's possible to trigger fence_drv.fences[] dereference after
the array has been freed. While the real problem is elsewhere, this still
results in confusing errors that depend on how the freed memory was
reused (I've seen "kernel tried to execute NX-protected page"), it's
better to clear them and get NULL dereference so that it's obvious what's
going wrong.
Signed-off-by: Grazvydas Ignotas <notasas@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
A little fallout from "drm/amdgpu: sanitize fence numbers", we
sometimes need to signal all fences in the ring.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Looks like the VCE block sometimes still sends nonsense
fence numbers on startup.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Using wrong counter for walking fences. Fixes
a crash when unloading the driver.
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This avoids problems with multiple GPUs. For example,
if the first GPU failed before amdgpu_fence_init() was
called, amdgpu_fence_slab_ref is still 0 and it will
get decremented in amdgpu_fence_driver_fini(). This
will lead to a crash during init of the second GPU since
amdgpu_fence_slab_ref is not 0.
v2: add functions for init/exit instead of
moving the variables into the driver.
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The rcu_assign_pointer() ensures that the initialization of a structure
is carried out before storing a pointer to that structre. It is always
safe to use RCU_INIT_POINTER() to NULL a pointer, instead of
rcu_assign_pointer().
This results in slightly smaller/faster code.
The following semantic patch was used:
<smpl>
@@
@@
- rcu_assign_pointer
+ RCU_INIT_POINTER
(..., NULL)
</smpl>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Muhammad Falak R Wani <falakreyaz@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
we introduced vmid fence, so one hw submission could produce two fences.
Signed-off-by: Chunming Zhou <David1.Zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
All these are compile time constand and the
drm_debugfs_create/remove_files functions take a const
pointer argument.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Nils Wallménius <nils.wallmenius@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
since the rcu slot was initialized to be num_hw_submission,
if command submission doesn't use scheduler, this limitation
will be invalid like uvd test.
Signed-off-by: Chunming Zhou <David1.Zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We don't need to extend them to 64bits any more, so avoid the extra overhead.
v2: update commit message.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
It's just overhead to check the fence value
when we signal them directly anyway.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Amdgpu doesn't support using scratch registers for fences any more.
So we won't see values like 0xdeadbeef as fence value any more.
v2: reschedule timer even if no change detected
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Because of the scheduler we need to signal all fences immediately
anyway, so try to avoid the waitqueue overhead.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Just wait for last fence instead of waiting for the sequence manually.
v2: don't use amdgpu_sched_jobs for the mask
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Just keep all HW fences in a RCU protected array as a
first step to replace the wait queue.
v2: update commit message, move fixes into separate patch.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Make this a parameter instead of using the global variable directly.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Fences must be freed RCU protected, otherwise the reservation_object_*_rcu()
functions can run into problems.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
The comment about the loop counter was never valid, even when you have
multiple threads this loop only runs as long as the sequence increases.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
No need to have that in the header file any more.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Not used any more.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Try to avoid using the hardware specific fences even more.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Not used any more since we now always use the sheduler.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
wait_event() never returns before the fence was signaled.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
It's not needed any more because all access goes through the scheduler now.
v2: Update commit message.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Ported from similar code in radeon.
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Ken Wang <Qingqing.Wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Not used any more without semaphores
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Non-scheduler code is longer supported.
v2: agd: rebased on upstream
Signed-off-by: Chunming Zhou <David1.Zhou@amd.com>
Reviewed-by: Ken Wang <Qingqing.Wang@amd.com>
Reviewed-by: Monk Liu <monk.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Change-Id: I5ad8dd156ccf27a6f18004aa0a215a0925b6e67b
Signed-off-by: Chunming Zhou <David1.Zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Less overhead than a work item and also adds proper cleanup handling.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Mostly unused and replaced by the common trace points.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
And also add some missing function documentation. No functional change.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Interrupts are notorious unreliable, enable the fallback at
a couple of more places.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Just move the remaining users to fence_put/get.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
No need to duplicate the functionality any more.
v2: fix handling if no fence is available.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com> (v1)
amdgpu_fence_default_wait isn't needed any more the default wait does the same
thing and amdgpu_test_signaled is dead as well.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
It didn't worked to well anyway.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Change-Id: I67e987db0efdca28faa80b332b75571192130d33
Signed-off-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: David Zhou <david1.zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Embed the scheduler into the ring structure instead of allocating it.
Use the ring name directly instead of the id.
v2: rebased, whitespace cleanup
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Reviewed-by: Chunming Zhou<david1.zhou@amd.com>
Move the fence related stuff into amdgpu_fence.c
v2: rework commit message, cause this is actually not a bug
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou<david1.zhou@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Just to be consistent with the other members.
v2: rename the ring member as well.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com> (v1)
Reviewed-by: Chunming Zhou<david1.zhou@amd.com>
That isn't used any more.
v2: rebase
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
