From: Danilo Krummrich <dakr@redhat.com>
To: "Christian König" <christian.koenig@amd.com>,
"Dave Airlie" <airlied@gmail.com>,
"Alex Deucher" <alexdeucher@gmail.com>
Cc: tzimmermann@suse.de, corbet@lwn.net,
nouveau@lists.freedesktop.org, dri-devel@lists.freedesktop.org,
linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org,
bskeggs@redhat.com, jason@jlekstrand.net, airlied@redhat.com
Subject: Re: [PATCH drm-next 00/14] [RFC] DRM GPUVA Manager & Nouveau VM_BIND UAPI
Date: Thu, 19 Jan 2023 05:04:32 +0100 [thread overview]
Message-ID: <6566870d-6256-8eef-5879-cb13711e4bed@redhat.com> (raw)
In-Reply-To: <7839c47e-6692-b93b-69a8-9584193cb07d@amd.com>
On 1/18/23 20:48, Christian König wrote:
> Am 18.01.23 um 20:17 schrieb Dave Airlie:
>> On Thu, 19 Jan 2023 at 02:54, Alex Deucher <alexdeucher@gmail.com> wrote:
>>> On Wed, Jan 18, 2023 at 11:50 AM Danilo Krummrich <dakr@redhat.com>
>>> wrote:
>>>>
>>>>
>>>> On 1/18/23 17:30, Alex Deucher wrote:
>>>>> On Wed, Jan 18, 2023 at 11:19 AM Danilo Krummrich <dakr@redhat.com>
>>>>> wrote:
>>>>>> On 1/18/23 16:37, Christian König wrote:
>>>>>>> Am 18.01.23 um 16:34 schrieb Danilo Krummrich:
>>>>>>>> Hi Christian,
>>>>>>>>
>>>>>>>> On 1/18/23 09:53, Christian König wrote:
>>>>>>>>> Am 18.01.23 um 07:12 schrieb Danilo Krummrich:
>>>>>>>>>> This patch series provides a new UAPI for the Nouveau driver in
>>>>>>>>>> order to
>>>>>>>>>> support Vulkan features, such as sparse bindings and sparse
>>>>>>>>>> residency.
>>>>>>>>>>
>>>>>>>>>> Furthermore, with the DRM GPUVA manager it provides a new DRM
>>>>>>>>>> core
>>>>>>>>>> feature to
>>>>>>>>>> keep track of GPU virtual address (VA) mappings in a more
>>>>>>>>>> generic way.
>>>>>>>>>>
>>>>>>>>>> The DRM GPUVA manager is indented to help drivers implement
>>>>>>>>>> userspace-manageable
>>>>>>>>>> GPU VA spaces in reference to the Vulkan API. In order to achieve
>>>>>>>>>> this goal it
>>>>>>>>>> serves the following purposes in this context.
>>>>>>>>>>
>>>>>>>>>> 1) Provide a dedicated range allocator to track GPU VA
>>>>>>>>>> allocations and
>>>>>>>>>> mappings, making use of the drm_mm range allocator.
>>>>>>>>> This means that the ranges are allocated by the kernel? If yes
>>>>>>>>> that's
>>>>>>>>> a really really bad idea.
>>>>>>>> No, it's just for keeping track of the ranges userspace has
>>>>>>>> allocated.
>>>>>>> Ok, that makes more sense.
>>>>>>>
>>>>>>> So basically you have an IOCTL which asks kernel for a free
>>>>>>> range? Or
>>>>>>> what exactly is the drm_mm used for here?
>>>>>> Not even that, userspace provides both the base address and the
>>>>>> range,
>>>>>> the kernel really just keeps track of things. Though, writing a
>>>>>> UAPI on
>>>>>> top of the GPUVA manager asking for a free range instead would be
>>>>>> possible by just adding the corresponding wrapper functions to get a
>>>>>> free hole.
>>>>>>
>>>>>> Currently, and that's what I think I read out of your question,
>>>>>> the main
>>>>>> benefit of using drm_mm over simply stuffing the entries into a
>>>>>> list or
>>>>>> something boils down to easier collision detection and iterating
>>>>>> sub-ranges of the whole VA space.
>>>>> Why not just do this in userspace? We have a range manager in
>>>>> libdrm_amdgpu that you could lift out into libdrm or some other
>>>>> helper.
>>>> The kernel still needs to keep track of the mappings within the various
>>>> VA spaces, e.g. it silently needs to unmap mappings that are backed by
>>>> BOs that get evicted and remap them once they're validated (or swapped
>>>> back in).
>>> Ok, you are just using this for maintaining the GPU VM space in the
>>> kernel.
>>>
>> Yes the idea behind having common code wrapping drm_mm for this is to
>> allow us to make the rules consistent across drivers.
>>
>> Userspace (generally Vulkan, some compute) has interfaces that pretty
>> much dictate a lot of how VMA tracking works, esp around lifetimes,
>> sparse mappings and splitting/merging underlying page tables, I'd
>> really like this to be more consistent across drivers, because already
>> I think we've seen with freedreno some divergence from amdgpu and we
>> also have i915/xe to deal with. I'd like to at least have one place
>> that we can say this is how it should work, since this is something
>> that *should* be consistent across drivers mostly, as it is more about
>> how the uapi is exposed.
>
> That's a really good idea, but the implementation with drm_mm won't work
> like that.
>
> We have Vulkan applications which use the sparse feature to create
> literally millions of mappings. That's why I have fine tuned the mapping
> structure in amdgpu down to ~80 bytes IIRC and save every CPU cycle
> possible in the handling of that.
That's a valuable information. Can you recommend such an application for
testing / benchmarking?
Your optimization effort sounds great. May it be worth thinking about
generalizing your approach by itself and stacking the drm_gpuva_manager
on top of it?
>
> A drm_mm_node is more in the range of ~200 bytes and certainly not
> suitable for this kind of job.
>
> I strongly suggest to rather use a good bunch of the amdgpu VM code as
> blueprint for the common infrastructure.
I will definitely have look.
>
> Regards,
> Christian.
>
>>
>> Dave.
>
next prev parent reply other threads:[~2023-01-19 4:28 UTC|newest]
Thread overview: 76+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-01-18 6:12 [PATCH drm-next 00/14] [RFC] DRM GPUVA Manager & Nouveau VM_BIND UAPI Danilo Krummrich
2023-01-18 6:12 ` [PATCH drm-next 01/14] drm: execution context for GEM buffers Danilo Krummrich
2023-01-18 6:12 ` [PATCH drm-next 02/14] drm/exec: fix memory leak in drm_exec_prepare_obj() Danilo Krummrich
2023-01-18 8:51 ` Christian König
2023-01-18 19:00 ` Danilo Krummrich
2023-01-18 6:12 ` [PATCH drm-next 03/14] drm: manager to keep track of GPUs VA mappings Danilo Krummrich
2023-01-19 4:14 ` Bagas Sanjaya
2023-01-23 23:23 ` Niranjana Vishwanathapura
2023-01-26 23:43 ` Matthew Brost
2023-01-27 0:24 ` Matthew Brost
2023-02-03 17:37 ` Matthew Brost
2023-02-06 13:35 ` Christian König
2023-02-06 13:46 ` Danilo Krummrich
2023-02-14 11:52 ` Danilo Krummrich
2023-01-18 6:12 ` [PATCH drm-next 04/14] drm: debugfs: provide infrastructure to dump a DRM GPU VA space Danilo Krummrich
2023-01-18 13:55 ` kernel test robot
2023-01-18 15:47 ` kernel test robot
2023-01-18 6:12 ` [PATCH drm-next 05/14] drm/nouveau: new VM_BIND uapi interfaces Danilo Krummrich
2023-01-27 1:05 ` Matthew Brost
2023-01-27 1:26 ` Danilo Krummrich
2023-01-27 7:55 ` Christian König
2023-01-27 13:12 ` Danilo Krummrich
2023-01-27 13:23 ` Christian König
2023-01-27 14:44 ` Danilo Krummrich
2023-01-27 15:17 ` Christian König
2023-01-27 20:25 ` David Airlie
2023-01-30 12:58 ` Christian König
2023-01-27 21:09 ` Danilo Krummrich
2023-01-29 18:46 ` Danilo Krummrich
2023-01-30 13:02 ` Christian König
2023-01-30 23:38 ` Danilo Krummrich
2023-02-01 8:10 ` [Nouveau] " Dave Airlie
2023-02-02 11:53 ` Christian König
2023-02-02 18:31 ` Danilo Krummrich
2023-02-06 9:48 ` Christian König
2023-02-06 13:27 ` Danilo Krummrich
2023-02-06 16:14 ` Christian König
2023-02-06 18:20 ` Danilo Krummrich
2023-02-07 9:35 ` Christian König
2023-02-07 10:50 ` Danilo Krummrich
2023-02-10 11:50 ` Christian König
2023-02-10 12:47 ` Danilo Krummrich
2023-01-27 1:43 ` Danilo Krummrich
2023-01-27 3:21 ` Matthew Brost
2023-01-27 3:33 ` Danilo Krummrich
2023-01-18 6:12 ` [PATCH drm-next 06/14] drm/nouveau: get vmm via nouveau_cli_vmm() Danilo Krummrich
2023-01-18 6:12 ` [PATCH drm-next 07/14] drm/nouveau: bo: initialize GEM GPU VA interface Danilo Krummrich
2023-01-18 6:12 ` [PATCH drm-next 08/14] drm/nouveau: move usercopy helpers to nouveau_drv.h Danilo Krummrich
2023-01-18 6:12 ` [PATCH drm-next 09/14] drm/nouveau: fence: fail to emit when fence context is killed Danilo Krummrich
2023-01-18 6:12 ` [PATCH drm-next 10/14] drm/nouveau: chan: provide nouveau_channel_kill() Danilo Krummrich
2023-01-18 6:12 ` [PATCH drm-next 11/14] drm/nouveau: nvkm/vmm: implement raw ops to manage uvmm Danilo Krummrich
2023-01-18 9:37 ` kernel test robot
2023-01-20 3:37 ` kernel test robot
2023-01-18 6:12 ` [PATCH drm-next 12/14] drm/nouveau: implement uvmm for user mode bindings Danilo Krummrich
2023-01-18 6:12 ` [PATCH drm-next 13/14] drm/nouveau: implement new VM_BIND UAPI Danilo Krummrich
2023-01-18 8:37 ` kernel test robot
2023-01-18 20:37 ` Thomas Hellström (Intel)
2023-01-19 3:44 ` Danilo Krummrich
2023-01-19 4:58 ` Matthew Brost
2023-01-19 7:32 ` Thomas Hellström (Intel)
2023-01-20 10:08 ` Boris Brezillon
2023-01-18 6:12 ` [PATCH drm-next 14/14] drm/nouveau: debugfs: implement DRM GPU VA debugfs Danilo Krummrich
2023-01-18 8:53 ` [PATCH drm-next 00/14] [RFC] DRM GPUVA Manager & Nouveau VM_BIND UAPI Christian König
2023-01-18 15:34 ` Danilo Krummrich
2023-01-18 15:37 ` Christian König
2023-01-18 16:19 ` Danilo Krummrich
2023-01-18 16:30 ` Alex Deucher
2023-01-18 16:50 ` Danilo Krummrich
2023-01-18 16:54 ` Alex Deucher
2023-01-18 19:17 ` Dave Airlie
2023-01-18 19:48 ` Christian König
2023-01-19 4:04 ` Danilo Krummrich [this message]
2023-01-19 5:23 ` Matthew Brost
2023-01-19 11:33 ` drm_gpuva_manager requirements (was Re: [PATCH drm-next 00/14] [RFC] DRM GPUVA Manager & Nouveau VM_BIND UAPI) Christian König
2023-02-06 14:48 ` [PATCH drm-next 00/14] [RFC] DRM GPUVA Manager & Nouveau VM_BIND UAPI Oded Gabbay
2023-03-16 16:39 ` Danilo Krummrich
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=6566870d-6256-8eef-5879-cb13711e4bed@redhat.com \
--to=dakr@redhat.com \
--cc=airlied@gmail.com \
--cc=airlied@redhat.com \
--cc=alexdeucher@gmail.com \
--cc=bskeggs@redhat.com \
--cc=christian.koenig@amd.com \
--cc=corbet@lwn.net \
--cc=dri-devel@lists.freedesktop.org \
--cc=jason@jlekstrand.net \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=nouveau@lists.freedesktop.org \
--cc=tzimmermann@suse.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).