Message-ID: <bb0d15b3-9e68-478d-9eb0-86814e60eb6f@intel.com>
Date: Mon, 3 Feb 2025 12:17:39 +0000
From: Matthew Auld <matthew.auld@...el.com>
To: Boris Brezillon <boris.brezillon@...labora.com>,
 Liviu Dudau <liviu.dudau@....com>
Cc: Danilo Krummrich <dakr@...nel.org>, asahi@...ts.linux.dev,
 Asahi Lina <lina@...hilina.net>,
 Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>,
 Maxime Ripard <mripard@...nel.org>, Thomas Zimmermann <tzimmermann@...e.de>,
 David Airlie <airlied@...il.com>, Simona Vetter <simona@...ll.ch>,
 Frank Binns <frank.binns@...tec.com>, Matt Coster <matt.coster@...tec.com>,
 Karol Herbst <kherbst@...hat.com>, Lyude Paul <lyude@...hat.com>,
 Steven Price <steven.price@....com>,
 Lucas De Marchi <lucas.demarchi@...el.com>,
 Thomas Hellström <thomas.hellstrom@...ux.intel.com>,
 Rodrigo Vivi <rodrigo.vivi@...el.com>, dri-devel@...ts.freedesktop.org,
 linux-kernel@...r.kernel.org, nouveau@...ts.freedesktop.org,
 intel-xe@...ts.freedesktop.org, akash.goel@....com
Subject: Re: [PATCH 0/4] drm/gpuvm: Add support for single-page-filled
 mappings

On 03/02/2025 12:12, Boris Brezillon wrote:
> On Mon, 3 Feb 2025 11:23:53 +0000
> Liviu Dudau <liviu.dudau@....com> wrote:
> 
>> On Mon, Feb 03, 2025 at 10:21:53AM +0100, Boris Brezillon wrote:
>>> +Akash with whom we've been discussing adding a 'REPEAT' mode to
>>> drm_gpuvm/panthor.
>>>
>>> On Sun, 2 Feb 2025 19:53:47 +0100
>>> Danilo Krummrich <dakr@...nel.org> wrote:
>>>    
>>>> Hi Lina,
>>>>
>>>> On Sun, Feb 02, 2025 at 10:34:49PM +0900, Asahi Lina wrote:
>>>>> Some hardware requires dummy page mappings to efficiently implement
>>>>> Vulkan sparse features. These mappings consist of the same physical
>>>>> memory page, repeated for a large range of address space (e.g. 16GiB).
>>>>>
>>>>> Add support for this to drm_gpuvm. Currently, drm_gpuvm expects BO
>>>>> ranges to correspond 1:1 to virtual memory ranges that are mapped, and
>>>>> does math on the BO offset accordingly. To make single page mappings
>>>>> work, we need a way to turn off that math, keeping the BO offset always
>>>>> constant and pointing to the same page (typically BO offset 0).
>>>>>
>>>>> To make this work, we need to handle all the corner cases when these
>>>>> mappings intersect with regular mappings. The rules are simply to never
>>>>> mix or merge a "regular" mapping with a single page mapping.
>>>>>
>>>>> drm_gpuvm has support for a flags field in drm_gpuva objects. This is
>>>>> normally managed by drivers directly. We can introduce a
>>>>> DRM_GPUVA_SINGLE_PAGE flag to handle this. However, to make it work,
>>>>> sm_map and friends need to know ahead of time whether the new mapping is
>>>>> a single page mapping or not. Therefore, we need to add an argument to
>>>>> these functions so drivers can provide the flags to be filled into
>>>>> drm_gpuva.flags.
>>>>>
>>>>> These changes should not affect any existing drivers that use drm_gpuvm
>>>>> other than the API change:
>>>>>
>>>>> - imagination: Does not use flags at all
>>>>> - nouveau: Only uses drm_gpuva_invalidate(), which is only called on
>>>>>    existing drm_gpuva objects (after the map steps)
>>>>> - panthor: Does not use flags at all
>>>>> - xe: Does not use drm_gpuva_init_from_op() or
>>>>>    drm_gpuva_remap()/drm_gpuva_map() (which call it). This means that the
>>>>> flags field of the gpuva object is managed by the driver only, so these
>>>>> changes cannot clobber it.
>>>>>
>>>>> Note that the way this is implemented, drm_gpuvm does not need to know
>>>>> the GPU page size. It only has to never do math on the BO offset to meet
>>>>> the requirements.
>>>>>
>>>>> I suspect that after this change there could be some cleanup possible in
>>>>> the xe driver (which right now passes flags around in various
>>>>> driver-specific ways from the map step through to drm_gpuva objects),
>>>>> but I'll leave that to the Xe folks.
>>>>>
>>>>> Signed-off-by: Asahi Lina <lina@...hilina.net>
>>>>> ---
>>>>> Asahi Lina (4):
>>>>>        drm/gpuvm: Add a flags argument to drm_gpuvm_sm_map[_*]
>>>>>        drm/gpuvm: Plumb through flags into drm_gpuva_op_map
>>>>>        drm/gpuvm: Add DRM_GPUVA_SINGLE_PAGE flag and logic
>>>>>        drm/gpuvm: Plumb through flags into drm_gpuva_init
>>>>
>>>> Without looking into any details yet:
>>>>
>>>> This is a bit of tricky one, since we're not even close to having a user for
>>>> this new feature upstream yet, are we?
>>>
>>> Actually, we would be interesting in having this feature hooked up in
>>> panthor. One use case we have is Vulkan sparse bindings, of course. But
>>> we also have cases where we need to map a dummy page repeatedly on the
>>> FW side. The approach we've been considering is slightly different:
>>> pass a DRM_GPUVA_REPEAT_FLAG along with GEM range, so we can repeat a
>>> range of the GEM (see the below diff, which is completely untested by
>>> the way), but I think we'd be fine with this SINGLE_PAGE flag.
>>
>> Unless I've misunderstood the intent completely, it looks like Xe also wants
>> something similar although they call it CPU_ADDR_MIRROR for some reason:
>>
>> https://lore.kernel.org/r/20250129195212.745731-9-matthew.brost@intel.com
> 
> At first glance, it doesn't seem related. The Xe stuff looks more like
> an alloc-on-fault mechanism. SINGLE_PAGE is about mapping a dummy page
> repeatedly over a virtual address range so that sparse Vulkan images
> never get a fault when an unbound image region is accessed.

Yeah, the CPU_ADDR_MIRROR is for upcoming svm stuff and not related. I 
believe in xe the sparse/repeat stuff in this series is rather done 
using a special HW feature called "NULL page", which is just a bit you 
can set in the pte (can also use huge pages) that creates a special 
mapping that doesn't actually point to any real memory underneath, but 
when doing a read it will always return zeroes and any writes to it will 
be dropped by the HW.

Powered by blists - more mailing lists