Mechanism for high priority scheduling in amdgpu
Quoted 2 lines · Serguei Sagalovitch
If compute queue is occupied only by you, the efficiency is equal with setting job queue to high priority I think.
The only risk is the situation when graphics will take all needed CUs. But in any case it should be very good test.
Andres/Pierre-Loup,
Did you try to do it or it is a lot of work for you?
The system will be fully loaded by the VR client application when this feature will need to be used, with hopefully both a graphics and compute job in flight using 100% of the CU capacity.
Let me try to succintly sum up requirements since you asked in the other branch:
On a fully loaded system (optimal occupancy by VR client app), we would like the VR runtime to be able to submit a task (graphics or compute, but we realize only compute might be possible for best results) and get results in a consistent amount of time. Ideally that time would be close to the time it would take to complete the same task on an otherwise idle system, but it's assumed there would be a fixed cost added to it due to winding down in-flight CUs. The quality of service provided by the feature would depend on how predictably small such a cost would be. 11ms would be a current upper limit but not really a useful number for the purpose of discussion, as the feature would be beyond useless at that point. Being able to intervene 1ms before vblank of the HMD and consistently get our task complete in time would be good.
BTW: If there is non-VR application which will use high-priority h/w queue then VR application will suffer. Any ideas how to solve it?
The intent is that this interface will require some sort of privilege that only the VR compositor would have on a well-configured VR system, higher-than-average niceness would be one idea. If you have any suggestions there also, it would be good discussion.
Thanks,
- Pierre-Loup
Sincerely yours,
Serguei Sagalovitch
Quoted 731 lines · On 2016-12-19 12:50 AM, zhoucm1 wrote:
Do you encounter the priority issue for compute queue with current driver?
If compute queue is occupied only by you, the efficiency is equal with setting job queue to high priority I think.
Regards,
David ZhouQuoted 720 lines · Andres Rodriguez
Yes, vulkan is available on all-open through the mesa radv UMD.
I'm not sure if I'm asking for too much, but if we can coordinate a similar interface in radv and amdgpu-pro at the vulkan level that would be great.
I'm not sure what that's going to be yet.
- Andres
Quoted 708 lines · On 12/19/2016 12:11 AM, zhoucm1 wrote:
Quoted 5 lines · Pierre-Loup A. Griffais
We're currently working with the open stack; I assume that a mechanism could be exposed by both open and Pro Vulkan userspace drivers and that the amdgpu kernel interface improvements we would pursue following this discussion would let both drivers take advantage of the feature, correct?
Of course.
Does open stack have Vulkan support?Regards,
David ZhouQuoted 684 lines · On 12/18/2016 07:26 PM, zhoucm1 wrote:
Quoted 682 lines
By the way, are you using all-open driver or amdgpu-pro driver?
+David Mao, who is working on our Vulkan driver.
Regards,
David ZhouQuoted 674 lines · Pierre-Loup A. Griffais
Hi Serguei,
I'm also working on the bringing up our VR runtime on top of amgpu; see replies inline.
Quoted 10 lines · Serguei
Andres,
Quoted 2 lines
For current VR workloads we have 3 separate processes running
actually:So we could have potential memory overcommit case or do you do partitioning
on your own? I would think that there is need to avoid overcomit in
VR case to
prevent any BO migration.You're entirely correct; currently the VR runtime is setting up prioritized CPU scheduling for its VR compositor, we're working on prioritized GPU scheduling and pre-emption (eg. this thread), and in the future it will make sense to do work in order to make sure that its memory allocations do not get evicted, to prevent any unwelcome additional latency in the event of needing to perform just-in-time reprojection.
Quoted 6 lines
BTW: Do you mean
__real__processes or threads?
Based on my understanding sharing BOs between different processes could introduce additional synchronization constrains. btw: I am not
sure
if we are able to share Vulkan sync. object cross-process boundary.They are different processes; it is important for the compositor that
is responsible for quality-of-service features such as consistently presenting distorted frames with the right latency, reprojection, etc,
to be separate from the main application.Currently we are using unreleased cross-process memory and semaphore extensions to fetch updated eye images from the client application, but the just-in-time reprojection discussed here does not actually have any direct interactions with cross-process resource sharing, since it's achieved by using whatever is the latest, most up-to-date eye images that have already been sent by the client application, which are already available to use without additional synchronization.
Quoted 5 lines
Quoted 4 lines
3) System compositor (we are looking at approaches to remove
this
overhead)Yes, IMHO the best is to run in "full screen mode".
Yes, we are working on mechanisms to present directly to the headset display without any intermediaries as a separate effort.
Quoted 8 lines
Quoted 2 lines
The latency is our main concern,
I would assume that this is the known problem (at least for compute usage).
It looks like that amdgpu / kernel submission is rather CPU intensive
(at least
in the default configuration).As long as it's a consistent cost, it shouldn't an issue. However, if
there's high degrees of variance then that would be troublesome and we
would need to account for the worst case.Hopefully the requirements and approach we described make sense, we're
looking forward to your feedback and suggestions.Thanks!
- Pierre-LoupQuoted 594 lines
Sincerely yours,
Serguei SagalovitchFrom: Andres Rodriguez <[email protected]>
Sent: December 16, 2016 10:00 PM
To: Sagalovitch, Serguei; [email protected]
Subject: RE: [RFC] Mechanism for high priority scheduling in amdgpuHey Serguei,
Quoted 10 lines
[Serguei] No. I mean pipe :-) as MEC define it. As far as I understand (by simplifying)
some scheduling is per pipe. I know about the current allocation scheme but I do not think
that it is ideal. I would assume that we need to switch to dynamical partition
of resources based on the workload otherwise we will have resource
conflict
between Vulkan compute and OpenCL.I agree the partitioning isn't ideal. I'm hoping we can start with a
solution that assumes that
only pipe0 has any work and the other pipes are idle (no HSA/ROCm running on the system).This should be more or less the use case we expect from VR users.
I agree the split is currently not ideal, but I'd like to consider that a separate task, because
making it dynamic is not straight forward :PQuoted 6 lines
[Serguei] Vulkan works via amdgpu (kernel submissions) so amdkfd will be not
involved. I would assume that in the case of VR we will have one main
application ("console" mode(?)) so we could temporally "ignore" OpenCL/ROCm needs when VR is running.Correct, this is why we want to enable the high priority compute queue through
libdrm-amdgpu, so that we can expose it through Vulkan later.For current VR workloads we have 3 separate processes running actually:
1) Game process
2) VR Compositor (this is the process that will require high
priority queue)
3) System compositor (we are looking at approaches to remove
this
overhead)For now I think it is okay to assume no OpenCL/ROCm running simultaneously, but
I would also like to be able to address this case in the future (cross-pipe priorities).Quoted 3 lines
[Serguei] The problem with pre-emption of graphics task: (a) it may take time so
latency may sufferThe latency is our main concern, we want something that is predictable. A good
illustration of what the reprojection scheduling looks like can be found here:
https://community.amd.com/servlet/JiveServlet/showImage/38-1310-104754/pastedImage_3.pngQuoted 4 lines
(b) to preempt we need to have different "context" - we want to guarantee that submissions from the same context will be executed
in order.This is okay, as the reprojection work doesn't have dependencies on the game context, and it
even happens in a separate process.Quoted 3 lines
BTW: (a) Do you want "preempt" and later resume or do you want "preempt" and
"cancel/abort"Preempt the game with the compositor task and then resume it.
Quoted 3 lines
(b) Vulkan is generic API and could be used for graphics as well as
for plain compute tasks (VK_QUEUE_COMPUTE_BIT).Yeah, the plan is to use vulkan compute. But if you figure out a way
for us to get
a guaranteed execution time using vulkan graphics, then I'll take you
out for a beer :)Regards,
Andres
________________________________________
From: Sagalovitch, Serguei [[email protected]]
Sent: Friday, December 16, 2016 9:13 PM
To: Andres Rodriguez; [email protected]
Subject: Re: [RFC] Mechanism for high priority scheduling in amdgpuHi Andres,
Please see inline (as [Serguei])
Sincerely yours,
Serguei SagalovitchFrom: Andres Rodriguez <[email protected]>
Sent: December 16, 2016 8:29 PM
To: Sagalovitch, Serguei; [email protected]
Subject: RE: [RFC] Mechanism for high priority scheduling in amdgpuHi Serguei,
Thanks for the feedback. Answers inline as [AR].
Regards,
Andres________________________________________
From: Sagalovitch, Serguei [[email protected]]
Sent: Friday, December 16, 2016 8:15 PM
To: Andres Rodriguez; [email protected]
Subject: Re: [RFC] Mechanism for high priority scheduling in amdgpuAndres,
Quick comments:
1) To minimize "bubbles", etc. we need to "force" CU assignments/binding
to high-priority queue when it will be in use and "free" them later
(we do not want forever take CUs from e.g. graphic task to degrade graphics
performance).Otherwise we could have scenario when long graphics task (or low-priority
compute) will took all (extra) CUs and high--priority will wait for needed resources.
It will not be visible on "NOP " but only when you submit "real" compute task
so I would recommend not to use "NOP" packets at all for testing.It (CU assignment) could be relatively easy done when everything is going via kernel
(e.g. as part of frame submission) but I must admit that I am not sure
about the best way for user level submissions (amdkfd).[AR] I wasn't aware of this part of the programming sequence. Thanks
for the heads up!
Is this similar to the CU masking programming? [Serguei] Yes. To simplify: the problem is that "scheduler" when deciding which
queue to run will check if there is enough resources and if not then
it will begin
to check other queues with lower priority.2) I would recommend to dedicate the whole pipe to high-priority queue and have
nothing their except it.[AR] I'm guessing in this context you mean pipe = queue? (as opposed
to the MEC definition
of pipe, which is a grouping of queues). I say this because amdgpu only has access to 1 pipe,
and the rest are statically partitioned for amdkfd usage.[Serguei] No. I mean pipe :-) as MEC define it. As far as I understand (by simplifying)
some scheduling is per pipe. I know about the current allocation scheme but I do not think
that it is ideal. I would assume that we need to switch to dynamical partition
of resources based on the workload otherwise we will have resource conflict
between Vulkan compute and OpenCL.BTW: Which user level API do you want to use for compute: Vulkan or OpenCL?
[AR] Vulkan
[Serguei] Vulkan works via amdgpu (kernel submissions) so amdkfd will
be not
involved. I would assume that in the case of VR we will have one main
application ("console" mode(?)) so we could temporally "ignore" OpenCL/ROCm needs when VR is running.Quoted 2 lines
we will not be able to provide a solution compatible with GFX
worloads.I assume that you are talking about graphics? Am I right?
[AR] Yeah, my understanding is that pre-empting the currently running
graphics job and scheduling in
something else using mid-buffer pre-emption has some cases where it doesn't work well. But if with
polaris10 it starts working well, it might be a better solution for us (because the whole reprojection
work uses the vulkan graphics stack at the moment, and porting it to
compute is not trivial).[Serguei] The problem with pre-emption of graphics task: (a) it may
take time so
latency may suffer (b) to preempt we need to have different "context"
- we want
to guarantee that submissions from the same context will be executed
in order.
BTW: (a) Do you want "preempt" and later resume or do you want "preempt" and
"cancel/abort"? (b) Vulkan is generic API and could be used for graphics as well as for plain compute tasks (VK_QUEUE_COMPUTE_BIT).Sincerely yours,
Serguei SagalovitchFrom: amd-gfx <[email protected]> on behalf of Andres Rodriguez <[email protected]>
Sent: December 16, 2016 6:15 PM
To: [email protected]
Subject: [RFC] Mechanism for high priority scheduling in amdgpuHi Everyone,
This RFC is also available as a gist here:
https://gist.github.com/lostgoat/7000432cd6864265dbc2c3ab93204249[RFC] Mechanism for high priority scheduling in amdgpu gist.github.com
[RFC] Mechanism for high priority scheduling in amdgpu[RFC] Mechanism for high priority scheduling in amdgpu gist.github.com
[RFC] Mechanism for high priority scheduling in amdgpu[RFC] Mechanism for high priority scheduling in amdgpu gist.github.com
[RFC] Mechanism for high priority scheduling in amdgpuWe are interested in feedback for a mechanism to effectively schedule
high
priority VR reprojection tasks (also referred to as time-warping) for
Polaris10
running on the amdgpu kernel driver.Brief context:
--------------The main objective of reprojection is to avoid motion sickness for VR
users in
scenarios where the game or application would fail to finish rendering a new
frame in time for the next VBLANK. When this happens, the user's head
movements
are not reflected on the Head Mounted Display (HMD) for the duration
of an
extra frame. This extended mismatch between the inner ear and the eyes may
cause the user to experience motion sickness.The VR compositor deals with this problem by fabricating a new frame
using the
user's updated head position in combination with the previous frames.
This
avoids a prolonged mismatch between the HMD output and the inner ear.Because of the adverse effects on the user, we require high confidence that the
reprojection task will complete before the VBLANK interval. Even if the GFX pipe
is currently full of work from the game/application (which is most likely the case).For more details and illustrations, please refer to the following document:
https://community.amd.com/community/gaming/blog/2016/03/28/asynchronous-shaders-evolvedGaming: Asynchronous Shaders Evolved | Community community.amd.com
One of the most exciting new developments in GPU technology over the
past year has been the adoption of asynchronous shaders, which can make more efficient use of ...Gaming: Asynchronous Shaders Evolved | Community community.amd.com
One of the most exciting new developments in GPU technology over the
past year has been the adoption of asynchronous shaders, which can make more efficient use of ...Gaming: Asynchronous Shaders Evolved | Community community.amd.com
One of the most exciting new developments in GPU technology over the
past year has been the adoption of asynchronous shaders, which can make more efficient use of ...Requirements:
-------------The mechanism must expose the following functionaility:
* Job round trip time must be predictable, from submission to
fence signal* The mechanism must support compute workloads.
Goals:
------* The mechanism should provide low submission latencies
Test: submitting a NOP packet through the mechanism on busy hardware
should
be equivalent to submitting a NOP on idle hardware.Nice to have:
-------------* The mechanism should also support GFX workloads.
My understanding is that with the current hardware capabilities in Polaris10 we
will not be able to provide a solution compatible with GFX worloads.But I would love to hear otherwise. So if anyone has an idea, approach or
suggestion that will also be compatible with the GFX ring, please let
us know
about it.* The above guarantees should also be respected by amdkfd
workloadsWould be good to have for consistency, but not strictly necessary as
users running
games are not traditionally running HPC workloads in the background.Proposed approach:
------------------Similar to the windows driver, we could expose a high priority compute queue to
userspace.Submissions to this compute queue will be scheduled with high priority, and may
acquire hardware resources previously in use by other queues.This can be achieved by taking advantage of the 'priority' field in the HQDs
and could be programmed by amdgpu or the amdgpu scheduler. The relevant
register fields are:
*mmCP_HQD_PIPE_PRIORITY
*mmCP_HQD_QUEUE_PRIORITYImplementation approach 1 - static partitioning:
------------------------------------------------The amdgpu driver currently controls 8 compute queues from pipe0. We can
statically partition these as follows:
* 7x regular
* 1x high priorityThe relevant priorities can be set so that submissions to the high priority
ring will starve the other compute rings and the GFX ring.The amdgpu scheduler will only place jobs into the high priority rings if the
context is marked as high priority. And a corresponding priority should be
added to keep track of this information:
*AMD_SCHED_PRIORITY_KERNEL
* ->AMD_SCHED_PRIORITY_HIGH
*AMD_SCHED_PRIORITY_NORMALThe user will request a high priority context by setting an appropriate flag
indrm_amdgpu_ctx_in(AMDGPU_CTX_HIGH_PRIORITYor similar):
https://github.com/torvalds/linux/blob/master/include/uapi/drm/amdgpu_drm.h#L163The setting is in a per context level so that we can:
* Maintain a consistent FIFO ordering of all submissions to a
context
* Create high priority and non-high priority contexts in the
same
processImplementation approach 2 - dynamic priority programming:
---------------------------------------------------------Similar to the above, but instead of programming the priorities and
amdgpu_init()time, the SW scheduler will reprogram the queue priorities
dynamically when scheduling a task.This would involve having a hardware specific callback from the scheduler to
set the appropriate queue priority: set_priority(int ring, int index,
int priority)During this callback we would have to grab the SRBM mutex to perform
the appropriate
HW programming, and I'm not really sure if that is something we should be doing from
the scheduler.On the positive side, this approach would allow us to program a range of
priorities for jobs instead of a single "high priority" value", achieving
something similar to the niceness API available for CPU scheduling.I'm not sure if this flexibility is something that we would need for
our use
case, but it might be useful in other scenarios (multiple users sharing compute
time on a server).This approach would require a new int field in
drm_amdgpu_ctx_in, or
repurposing
of the flags field.Known current obstacles:
------------------------The SQ is currently programmed to disregard the HQD priorities, and instead it picks
jobs at random. Settings from the shader itself are also disregarded
as this is
considered a privileged field.Effectively we can get our compute wavefront launched ASAP, but we might not get the
time we need on the SQ.The current programming would have to be changed to allow priority propagation
from the HQD into the SQ.Generic approach for all HW IPs:
--------------------------------For consistency purposes, the high priority context can be enabled for all HW IPs
with support of the SW scheduler. This will function similarly to the
currentAMD_SCHED_PRIORITY_KERNELpriority, where the job can jump ahead of anything not
commited to the HW queue.The benefits of requesting a high priority context for a non-compute
queue will
be lesser (e.g. up to 10s of wait time if a GFX command is stuck in front of
you), but having the API in place will allow us to easily improve the
implementation
in the future as new features become available in new hardware.Future steps:
-------------Once we have an approach settled, I can take care of the implementation.
Also, once the interface is mostly decided, we can start thinking about
exposing the high priority queue through radv.Request for feedback:
---------------------We aren't married to any of the approaches outlined above. Our goal is to
obtain a mechanism that will allow us to complete the reprojection job within a
predictable amount of time. So if anyone anyone has any suggestions for
improvements or alternative strategies we are more than happy to hear
them.If any of the technical information above is also incorrect, feel free to point
out my misunderstandings.Looking forward to hearing from you.
Regards,
Andres_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfxamd-gfx Info Page - lists.freedesktop.org lists.freedesktop.org
To see the collection of prior postings to the list, visit the amd-gfx Archives. Using amd-gfx: To post a message to all the list members, send email ...amd-gfx Info Page - lists.freedesktop.org lists.freedesktop.org
To see the collection of prior postings to the list, visit the amd-gfx Archives. Using amd-gfx: To post a message to all the list members, send email ..._______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfxQuoted 6 lines
Quoted 5 lines
Quoted 4 lines
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
Sincerely yours,
Serguei Sagalovitch
_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx