On Thu, 8 Jul 2021 14:10:45 +0200
Christian König wrote:
> >> --- a/drivers/gpu/drm/panfrost/panfrost_job.c
> >> +++ b/drivers/gpu/drm/panfrost/panfrost_job.c
> >> @@ -254,6 +254,9 @@ static int panfrost_acquire_object_fences(struct
> >> panfrost_job *job)
> >>return
lcome, as usual.
>
> Cheers, Daniel
>
> Daniel Vetter (18):
> drm/sched: Split drm_sched_job_init
> drm/sched: Barriers are needed for entity->last_scheduled
> drm/sched: Add dependency tracking
> drm/sched: drop entity parameter from drm_sched_push_job
> drm/s
On Thu, 11 Mar 2021 12:16:33 +
Steven Price wrote:
> Also the current code completely ignores PANFROST_BO_REF_READ. So either
> that should be defined as 0, or even better we support 3 modes:
>
> * Exclusive ('write' access)
> * Shared ('read' access)
> * No fence - ensures the BO is
On Fri, 12 Mar 2021 09:37:49 -0600
Jason Ekstrand wrote:
> On Fri, Mar 12, 2021 at 1:31 AM Boris Brezillon
> wrote:
> >
> > On Thu, 11 Mar 2021 12:11:48 -0600
> > Jason Ekstrand wrote:
> >
> > > > > > > > 2/ Queued
On Fri, 12 Mar 2021 19:25:13 +0100
Boris Brezillon wrote:
> > So where does this leave us? Well, it depends on your submit model
> > and exactly how you handle pipeline barriers that sync between
> > engines. If you're taking option 3 above and doing two comma
discussing the first issue with Steve or Robin a while back,
but we never hit it before (now we do :)).
The last patch is a perf improvement: no need to re-enable hardware
interrupts if we know the threaded irq handler will be woken up right
away.
Regards,
Boris
Boris Brezillon (3):
drm/panfrost
heap allocations")
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_mmu.c | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_mmu.c
b/drivers/gpu/drm/panfrost/panfrost_mmu.c
index 904d63450862..21e552d1ac71 100644
---
t for GPU heap allocations")
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_mmu.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_mmu.c
b/drivers/gpu/drm/panfrost/panfrost_mmu.c
index 7c1b3481b785..904d63450
Doing a hw-irq -> threaded-irq round-trip is counter-productive, stay
in the threaded irq handler as long as we can.
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_mmu.c | 7 +++
1 file changed, 7 insertions(+)
diff --git a/drivers/gpu/drm/panfrost/panfrost_mmu.c
On Mon, 1 Feb 2021 12:13:49 +
Steven Price wrote:
> On 01/02/2021 08:21, Boris Brezillon wrote:
> > Doing a hw-irq -> threaded-irq round-trip is counter-productive, stay
> > in the threaded irq handler as long as we can.
> >
> > Signed-off-by: Boris Brezillon
On Mon, 1 Feb 2021 13:24:00 +
Steven Price wrote:
> On 01/02/2021 12:59, Boris Brezillon wrote:
> > On Mon, 1 Feb 2021 12:13:49 +
> > Steven Price wrote:
> >
> >> On 01/02/2021 08:21, Boris Brezillon wrote:
> >>> Doing a hw-irq -> thread
handling loop to avoid a goto
Boris Brezillon (3):
drm/panfrost: Clear MMU irqs before handling the fault
drm/panfrost: Don't try to map pages that are already mapped
drm/panfrost: Stay in the threaded MMU IRQ handler until we've handled
all IRQs
drivers/gpu/drm/panfrost/panf
t for GPU heap allocations")
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_mmu.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_mmu.c
b/drivers/gpu/drm/panfrost/panfrost_m
Doing a hw-irq -> threaded-irq round-trip is counter-productive, stay
in the threaded irq handler as long as we can.
v2:
* Rework the loop to avoid a goto
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_mmu.c | 26 +
1 file changed, 14 inserti
heap allocations")
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_mmu.c | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_mmu.c
b/drivers/gpu/drm/panfrost/panfrost_mmu.c
index 90
On Fri, 5 Feb 2021 12:17:54 +0100
Boris Brezillon wrote:
> Hello,
>
> Here are 2 fixes and one improvement for the page fault handling. Those
> bugs were found while working on indirect draw supports which requires
> the allocation of a big heap buffer for varyings, and t
> >>> fs_reclaim shrinker annotations, which means you don't have lockdep
> > >>> checks. I think at least, would need some deadlock and testing.
> > >>
> > >> The big problem with this sort of method for triggering the shrinkers is
> >
I'm just trying to collect feedback. I don't intend to get
those patches merged until we have a userspace user, but I thought
starting the discussion early would be a good thing.
Feel free to suggest other approaches.
Regards,
Boris
Boris Brezillon (7):
drm/panfrost: Pass a job
So we can re-use it from elsewhere.
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_drv.c | 52 ++---
1 file changed, 29 insertions(+), 23 deletions(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_drv.c
b/drivers/gpu/drm/panfrost/panfrost_drv.c
index
This way we can re-use the standard drm_gem_fence_array_add_implicit()
helper and simplify the panfrost_job_dependency() logic.
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_drv.c | 42 +++---
drivers/gpu/drm/panfrost/panfrost_job.c | 57
Jobs reading from the same BO should not be serialized. Add access flags
so we can relax the implicit dependencies in that case. We force RW
access for now to keep the behavior unchanged, but a new SUBMIT ioctl
taking explicit access flags will be introduced.
Signed-off-by: Boris Brezillon
We now have a new ioctl that allows submitting multiple jobs at once
(among other things) and we support timelined syncobjs. Bump the
minor version number to reflect those changes.
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_drv.c | 3 ++-
1 file changed, 2 insertions
Now that we have a new SUBMIT ioctl dealing with timelined syncojbs we
can advertise the feature.
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_drv.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_drv.c
b/drivers
So we don't have to change the prototype if we extend the function.
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_job.c | 22 --
1 file changed, 8 insertions(+), 14 deletions(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_job.c
b/drivers/gp
This should help limit the number of ioctls when submitting multiple
jobs. The new ioctl also supports syncobj timelines and BO access flags.
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_drv.c | 303
include/uapi/drm/panfrost_drm.h | 79
Hi Steven,
On Thu, 11 Mar 2021 12:16:33 +
Steven Price wrote:
> On 11/03/2021 09:25, Boris Brezillon wrote:
> > Hello,
> >
> > I've been playing with Vulkan lately and struggled quite a bit to
> > implement VkQueueSubmit with the submit ioctl we have. There
Hi Jason,
On Thu, 11 Mar 2021 10:58:46 -0600
Jason Ekstrand wrote:
> Hi all,
>
> Dropping in where I may or may not be wanted to feel free to ignore. : -)
I'm glad you decided to chime in. :-)
> > > > 2/ Queued jobs might be executed out-of-order (unless they have
> > > > explicit/impli
On Thu, 11 Mar 2021 12:11:48 -0600
Jason Ekstrand wrote:
> > > > > > 2/ Queued jobs might be executed out-of-order (unless they have
> > > > > > explicit/implicit deps between them), and Vulkan asks that the
> > > > > > out
> > > > > > fence be signaled when all jobs are done. Timeline s
Make sure all bo->base.pages entries are either NULL or pointing to a
valid page before calling drm_gem_shmem_put_pages().
Reported-by: Tomeu Vizoso
Cc:
Fixes: 187d2929206e ("drm/panfrost: Add support for GPU heap allocations")
Signed-off-by: Boris Brezillon
---
drivers/gpu
On Thu, 3 Dec 2020 18:20:48 +0530
Nikhil Devshatwar wrote:
> input_bus_flags are specified in drm_bridge_timings (legacy) as well
> as drm_bridge_state->input_bus_cfg.flags
>
> The flags from the timings will be deprecated. Bridges are supposed
> to validate and set the bridge state flags from a
On Tue, 1 Dec 2020 17:48:27 +0530
Nikhil Devshatwar wrote:
> input_bus_flags are specified in drm_bridge_timings (legacy) as well
> as drm_bridge_state->input_bus_cfg.flags
>
> The flags from the timings will be deprecated. Bridges are supposed
> to validate and set the bridge state flags from a
On Tue, 1 Dec 2020 17:48:26 +0530
Nikhil Devshatwar wrote:
> With new connector model, mhdp bridge will not create the connector and
> SoC driver will rely on format negotiation to setup the encoder format.
>
> Support minimal format negotiations hooks in the drm_bridge_funcs.
> Complete format
On Tue, 1 Dec 2020 17:48:27 +0530
Nikhil Devshatwar wrote:
> input_bus_flags are specified in drm_bridge_timings (legacy) as well
> as drm_bridge_state->input_bus_cfg.flags
>
> The flags from the timings will be deprecated. Bridges are supposed
> to validate and set the bridge state flags from a
On Tue, 1 Dec 2020 17:48:28 +0530
Nikhil Devshatwar wrote:
> Remove the old code to iterate over the bridge chain, as this is
> already done by the framework.
> The bridge state should have the negotiated bus format and flags.
> Use these from the bridge's state.
> If the bridge does not support
On Fri, 4 Dec 2020 12:56:27 +0200
Tomi Valkeinen wrote:
> Hi Boris,
>
> On 04/12/2020 12:50, Boris Brezillon wrote:
> > On Tue, 1 Dec 2020 17:48:28 +0530
> > Nikhil Devshatwar wrote:
> >
> >> Remove the old code to iterate over the bridge chain, as this i
On Fri, 4 Dec 2020 13:47:05 +0200
Tomi Valkeinen wrote:
> On 04/12/2020 13:12, Boris Brezillon wrote:
>
> >>> That'd be even better if you implement the bridge interface instead of
> >>> the encoder one so we can get rid of the encoder_{helper}_funcs an
The documentation is a bit vague and doesn't really describe what the
->timedout_job() is expected to do. Let's add a few more details.
v5:
* New patch
Suggested-by: Daniel Vetter
Signed-off-by: Boris Brezillon
---
include/drm/gpu_scheduler.h | 14 ++
1 file changed,
see each commit for a detailed
changelog) and collect R-b/A-b tags
Regards,
Boris
Boris Brezillon (15):
drm/sched: Document what the timedout_job method should do
drm/sched: Allow using a dedicated workqueue for the timeout/fault tdr
drm/panfrost: Make ->run_job() return an ERR
If the fence creation fail, we can return the error pointer directly.
The core will update the fence error accordingly.
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
Reviewed-by: Alyssa Rosenzweig
---
drivers/gpu/drm/panfrost/panfrost_job.c | 2 +-
1 file changed, 1 insertion(+), 1
() method
v3:
* New patch
v4:
* Actually use the timeout_wq to queue the timeout work
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
Reviewed-by: Lucas Stach
Cc: Qiang Yu
Cc: Emma Anholt
Cc: Alex Deucher
Cc: "Christian König"
---
drivers/gpu/drm/amd/amdgpu/amdgpu_fe
executing and
cancel_work_sync() will wait for the handler to return. Checking the
reset pending status should avoid spurious resets
v3:
* New patch
Suggested-by: Daniel Vetter
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_device.h | 6 +
Expose a helper to trigger a GPU reset so we can easily trigger reset
operations outside the job timeout handler.
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
Reviewed-by: Alyssa Rosenzweig
---
drivers/gpu/drm/panfrost/panfrost_device.h | 8
drivers/gpu/drm/panfrost
Things are unlikely to resolve until we reset the GPU. Let's not wait
for other faults/timeout to happen to trigger this reset.
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_mmu.c | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
This should avoid switching to interrupt context when the GPU is under
heavy use.
v3:
* Don't take the job_lock in panfrost_job_handle_irq()
Signed-off-by: Boris Brezillon
Acked-by: Alyssa Rosenzweig
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_job.c
Currently unused. We'll add it back if we need per-GPU definitions.
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_device.c | 2 +-
drivers/gpu/drm/panfrost/panfrost_device.h | 2 +-
drivers/gpu/drm/panfrost/panfrost_gpu.c| 2 +-
driver
tring table declaration
in the same patch
v3:
* Drop the error field
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
Reviewed-by: Alyssa Rosenzweig
---
drivers/gpu/drm/panfrost/panfrost_device.c | 130 +
drivers/gpu/drm/panfrost/panfrost_device.h | 69 +++
2 fi
If we don't do that, we have to wait for the job timeout to expire
before the fault jobs gets killed.
v3:
* Make sure the AS is re-enabled when new jobs are submitted to the
context
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_device.h
taking the job_lock (not needed since
this lock is never taken from an interrupt context)
v3:
* Set fence error to ECANCELED when a TERMINATED exception is received
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_device.h | 7
drivers/gpu/drm/panfrost/panfrost_job
If we can recover from a fault without a reset there's no reason to
issue one.
v3:
* Drop the mention of Valhall requiring a reset on JOB_BUS_FAULT
* Set the fence error to -EINVAL instead of having per-exception
error codes
Signed-off-by: Boris Brezillon
Reviewed-by: Steven
lid active state
* Only enable the second slot on GPUs supporting jobchain disambiguation
* Split interrupt handling in sub-functions
Signed-off-by: Steven Price
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_device.h | 2 +-
drivers/gpu/drm/panfrost/panfrost_job.
Exception types will be defined as an enum.
v4:
* Fix typo in the commit message
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
Reviewed-by: Alyssa Rosenzweig
---
drivers/gpu/drm/panfrost/panfrost_regs.h | 3 ---
1 file changed, 3 deletions(-)
diff --git a/drivers/gpu/drm/panfrost
igger timeout value (1ms -> 100ms) to be on
the safe side.
v5:
* New patch
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_mmu.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_mmu.c
b/drivers/gpu/drm/panfrost/panfro
et is issued gets tricky if we keep job interrupts active.
Let's prepare for that and mask+flush job IRQs before issuing a reset.
v4:
* Add a comment explaining why we WARN_ON(!job) in the irq handler
* Keep taking the job_lock when evicting stalled jobs
v3:
* New patch
Signed-off-by: Boris
On Tue, 29 Jun 2021 10:50:36 +0200
Daniel Vetter wrote:
> On Tue, Jun 29, 2021 at 09:34:56AM +0200, Boris Brezillon wrote:
> > Mali Midgard/Bifrost GPUs have 3 hardware queues but only a global GPU
> > reset. This leads to extra complexity when we need to synchronize timeout
>
Hi Christian,
On Tue, 29 Jun 2021 13:03:58 +0200
Christian König wrote:
> Am 29.06.21 um 09:34 schrieb Boris Brezillon:
> > Mali Midgard/Bifrost GPUs have 3 hardware queues but only a global GPU
> > reset. This leads to extra complexity when we need to synchronize timeout
>
On Tue, 29 Jun 2021 09:35:03 +0200
Boris Brezillon wrote:
> @@ -379,57 +370,72 @@ void panfrost_job_enable_interrupts(struct
> panfrost_device *pfdev)
> job_write(pfdev, JOB_INT_MASK, irq_mask);
> }
>
> -static bool panfrost_scheduler_stop(struct panfrost_
The documentation is a bit vague and doesn't really describe what the
->timedout_job() is expected to do. Let's add a few more details.
v5:
* New patch
Suggested-by: Daniel Vetter
Signed-off-by: Boris Brezillon
Reviewed-by: Daniel Vetter
---
include/drm/gpu_scheduler.h | 14
e header
Regards,
Boris
Boris Brezillon (15):
drm/sched: Document what the timedout_job method should do
drm/sched: Allow using a dedicated workqueue for the timeout/fault tdr
drm/panfrost: Make ->run_job() return an ERR_PTR() when appropriate
drm/panfrost: Get rid of
() method
v3:
* New patch
v4:
* Actually use the timeout_wq to queue the timeout work
Suggested-by: Daniel Vetter
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
Reviewed-by: Lucas Stach
Acked-by: Daniel Vetter
Acked-by: Christian König
Cc: Qiang Yu
Cc: Emma Anholt
Cc: Alex Deucher
Exception types will be defined as an enum.
v4:
* Fix typo in the commit message
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
Reviewed-by: Alyssa Rosenzweig
---
drivers/gpu/drm/panfrost/panfrost_regs.h | 3 ---
1 file changed, 3 deletions(-)
diff --git a/drivers/gpu/drm/panfrost
If the fence creation fail, we can return the error pointer directly.
The core will update the fence error accordingly.
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
Reviewed-by: Alyssa Rosenzweig
---
drivers/gpu/drm/panfrost/panfrost_job.c | 2 +-
1 file changed, 1 insertion(+), 1
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_device.h | 6 +-
drivers/gpu/drm/panfrost/panfrost_job.c| 186 -
2 files changed, 69 insertions(+), 123 deletions(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_device.h
b/
Currently unused. We'll add it back if we need per-GPU definitions.
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_device.c | 2 +-
drivers/gpu/drm/panfrost/panfrost_device.h | 2 +-
drivers/gpu/drm/panfrost/panfrost_gpu.c| 2 +-
driver
Expose a helper to trigger a GPU reset so we can easily trigger reset
operations outside the job timeout handler.
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
Reviewed-by: Alyssa Rosenzweig
---
drivers/gpu/drm/panfrost/panfrost_device.h | 8
drivers/gpu/drm/panfrost
This should avoid switching to interrupt context when the GPU is under
heavy use.
v3:
* Don't take the job_lock in panfrost_job_handle_irq()
Signed-off-by: Boris Brezillon
Acked-by: Alyssa Rosenzweig
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_job.c
If we can recover from a fault without a reset there's no reason to
issue one.
v3:
* Drop the mention of Valhall requiring a reset on JOB_BUS_FAULT
* Set the fence error to -EINVAL instead of having per-exception
error codes
Signed-off-by: Boris Brezillon
Reviewed-by: Steven
If we don't do that, we have to wait for the job timeout to expire
before the fault jobs gets killed.
v3:
* Make sure the AS is re-enabled when new jobs are submitted to the
context
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_device.h
tring table declaration
in the same patch
v3:
* Drop the error field
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
Reviewed-by: Alyssa Rosenzweig
---
drivers/gpu/drm/panfrost/panfrost_device.c | 130 +
drivers/gpu/drm/panfrost/panfrost_device.h | 69 +++
2 fi
lid active state
* Only enable the second slot on GPUs supporting jobchain disambiguation
* Split interrupt handling in sub-functions
Signed-off-by: Steven Price
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_device.h | 2 +-
drivers/gpu/drm/panfrost/panfrost_job.
et is issued gets tricky if we keep job interrupts active.
Let's prepare for that and mask+flush job IRQs before issuing a reset.
v4:
* Add a comment explaining why we WARN_ON(!job) in the irq handler
* Keep taking the job_lock when evicting stalled jobs
v3:
* New patch
Signed-off-by: Boris
igger timeout value (1ms -> 100ms) to be on
the safe side.
v5:
* New patch
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_mmu.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_mmu.c
b/drivers/gpu/drm/panfrost/panfro
taking the job_lock (not needed since
this lock is never taken from an interrupt context)
v3:
* Set fence error to ECANCELED when a TERMINATED exception is received
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_device.h | 7
drivers/gpu/drm/panfrost/panfrost_job
Things are unlikely to resolve until we reset the GPU. Let's not wait
for other faults/timeout to happen to trigger this reset.
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_mmu.c | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
On Wed, 30 Jun 2021 08:27:35 +0200
Boris Brezillon wrote:
> Hello,
>
> Bunch of improvements to make the panfrost driver more robust and allow
> queuing jobs at the HW level.
>
> Changes in v6:
> * Collected acks/reviews
> * Got rid of the cancel_delayed_work()
So we can re-use it from elsewhere.
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_drv.c | 52 ++---
1 file changed, 29 insertions(+), 23 deletions(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_drv.c
b/drivers/gpu/drm/panfrost/panfrost_drv.c
index
this new ioctl is welcome, in particular, do you
think other things are missing/would be nice to have for Vulkan?
Regards,
Boris
P.S.: basic igt tests for these new ioctls re available there [1]
[1]https://gitlab.freedesktop.org/bbrezillon/igt-gpu-tools/-/tree/panfrost-batch-submit
Boris
So we don't have to change the prototype if we extend the function.
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_job.c | 22 --
1 file changed, 8 insertions(+), 14 deletions(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_job.c
b/drivers/gp
Needed to keep VkQueues isolated from each other.
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/Makefile | 3 +-
drivers/gpu/drm/panfrost/panfrost_device.h| 2 +-
drivers/gpu/drm/panfrost/panfrost_drv.c | 69 --
drivers/gpu/drm/panfrost
This should help limit the number of ioctls when submitting multiple
jobs. The new ioctl also supports syncobj timelines and BO access flags.
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_drv.c | 305
drivers/gpu/drm/panfrost/panfrost_job.c | 3
Jobs reading from the same BO should not be serialized. Add access
flags so we can relax the implicit dependencies in that case. We force
exclusive access for now to keep the behavior unchanged, but a new
SUBMIT ioctl taking explicit access flags will be introduced.
Signed-off-by: Boris Brezillon
Now that we have a new SUBMIT ioctl dealing with timelined syncojbs we
can advertise the feature.
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_drv.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_drv.c
b/drivers
We now have a new ioctl that allows submitting multiple jobs at once
(among other things) and we support timelined syncobjs. Bump the
minor version number to reflect those changes.
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/panfrost_drv.c | 4 +++-
1 file changed, 3 insertions
On Fri, 2 Jul 2021 10:56:29 +0100
Steven Price wrote:
> On 01/07/2021 10:12, Boris Brezillon wrote:
> > Needed to keep VkQueues isolated from each other.
> >
> > Signed-off-by: Boris Brezillon
>
> My Vulkan knowledge is limited so I'm not sure whether this is
On Fri, 2 Jul 2021 11:08:58 +0100
Steven Price wrote:
> On 01/07/2021 10:12, Boris Brezillon wrote:
> > Needed to keep VkQueues isolated from each other.
>
> One more comment I noticed when I tried this out:
>
> [...]
> > +struct panfrost_submitqueue *
> > +p
On Fri, 2 Jul 2021 11:58:34 +0100
Steven Price wrote:
> On 02/07/2021 11:52, Boris Brezillon wrote:
> > On Fri, 2 Jul 2021 11:08:58 +0100
> > Steven Price wrote:
> >
> >> On 01/07/2021 10:12, Boris Brezillon wrote:
> >>> Needed to keep VkQueues i
On Fri, 2 Jul 2021 09:58:06 -0400
Alyssa Rosenzweig wrote:
> > > My Vulkan knowledge is limited so I'm not sure whether this is the right
> > > approach or not. In particular is it correct that an application can
> > > create a high priority queue which could affect other (normal priority)
> > >
3:
* Fix a deadlock in the submitqueue logic
* Limit the number of submitqueue per context to 16
Boris Brezillon (7):
drm/panfrost: Pass a job to panfrost_{acquire,attach}_object_fences()
drm/panfrost: Move the mappings collection out of
panfrost_lookup_bos()
drm/panfrost: Add BO access fl
So we don't have to change the prototype if we extend the function.
v3:
* Fix subject
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_job.c | 22 --
1 file changed, 8 insertions(+), 14 deletions(-)
diff --git a/driver
So we can re-use it from elsewhere.
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_drv.c | 52 ++---
1 file changed, 29 insertions(+), 23 deletions(-)
diff --git a/drivers/gpu/drm/panfrost/panfrost_drv.c
b/drivers/gpu/drm
Jobs reading from the same BO should not be serialized. Add access
flags so we can relax the implicit dependencies in that case. We force
exclusive access for now to keep the behavior unchanged, but a new
SUBMIT ioctl taking explicit access flags will be introduced.
Signed-off-by: Boris Brezillon
This should help limit the number of ioctls when submitting multiple
jobs. The new ioctl also supports syncobj timelines and BO access flags.
v3:
* Re-use panfrost_get_job_bos() and panfrost_get_job_in_syncs() in the
old submit path
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost
We now have a new ioctl that allows submitting multiple jobs at once
(among other things) and we support timelined syncobjs. Bump the
minor version number to reflect those changes.
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_drv.c | 4 +++-
1
Now that we have a new SUBMIT ioctl dealing with timelined syncojbs we
can advertise the feature.
Signed-off-by: Boris Brezillon
Reviewed-by: Steven Price
---
drivers/gpu/drm/panfrost/panfrost_drv.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/panfrost
Needed to keep VkQueues isolated from each other.
v3:
* Limit the number of submitqueue per context to 16
* Fix a deadlock
Signed-off-by: Boris Brezillon
---
drivers/gpu/drm/panfrost/Makefile | 3 +-
drivers/gpu/drm/panfrost/panfrost_device.h| 2 +-
drivers/gpu/drm/panfrost
On Fri, 2 Jul 2021 11:13:16 -0400
Alyssa Rosenzweig wrote:
> ```
> > +/* Syncobj reference passed at job submission time to encode explicit
> > + * input/output fences.
> > + */
> > +struct drm_panfrost_syncobj_ref {
> > + __u32 handle;
> > + __u32 pad;
> > + __u64 point;
> > +};
> ```
>
On Fri, 2 Jul 2021 16:05:30 +0100
Steven Price wrote:
> On 02/07/2021 15:32, Boris Brezillon wrote:
> > Needed to keep VkQueues isolated from each other.
> >
> > v3:
> > * Limit the number of submitqueue per context to 16
> > * Fix a deadlock
> >
> &
On Fri, 2 Jul 2021 17:49:10 +0200
Boris Brezillon wrote:
> On Fri, 2 Jul 2021 16:05:30 +0100
> Steven Price wrote:
>
> > On 02/07/2021 15:32, Boris Brezillon wrote:
> > > Needed to keep VkQueues isolated from each other.
> > >
> > > v3:
> > &g
On Fri, 2 Jul 2021 12:49:55 -0400
Alyssa Rosenzweig wrote:
> > > Why is there padding instead of putting point first?
> >
> > We can move the point field first, but we need to keep the explicit
> > padding: the struct has to be 64bit aligned because of the __u64 field
> > (which the compiler t
On Fri, 2 Jul 2021 12:49:55 -0400
Alyssa Rosenzweig wrote:
> > > ```
> > > > #define PANFROST_BO_REF_EXCLUSIVE 0x1
> > > > +#define PANFROST_BO_REF_NO_IMPLICIT_DEP0x2
> > > ```
> > >
> > > This seems logically backwards. NO_IMPLICIT_DEP makes sense if we're
> > > trying to ke
ubmitqueue per context to 16
*** BLURB HERE ***
Boris Brezillon (7):
drm/panfrost: Pass a job to panfrost_{acquire,attach}_object_fences()
drm/panfrost: Move the mappings collection out of
panfrost_lookup_bos()
drm/panfrost: Add BO access flags to relax dependencies between jobs
drm/panfr
1 - 100 of 2442 matches
Mail list logo
