[Mesa-dev] GBM and the Device Memory Allocator Proposals

[Rob Clark]

On Wed, Dec 6, 2017 at 6:25 AM, Nicolai Hähnle <nhaehnle at gmail.com> wrote:
> On 06.12.2017 08:07, James Jones wrote:
> [snip]
>
>>>>>> So lets say you have a setup where both display and GPU supported
>>>>>> FOO/tiled, but only GPU supported compressed (FOO/CC) and cached
>>>>>> (FOO/cached).  But the GPU supported the following transitions:
>>>>>>
>>>>>>     trans_a: FOO/CC -> null
>>>>>>     trans_b: FOO/cached -> null
>>>>>>
>>>>>> Then the sets for each device (in order of preference):
>>>>>>
>>>>>> GPU:
>>>>>>     1: caps(FOO/tiled, FOO/CC, FOO/cached); constraints(alignment=32k)
>>>>>>     2: caps(FOO/tiled, FOO/CC); constraints(alignment=32k)
>>>>>>     3: caps(FOO/tiled); constraints(alignment=32k)
>>>>>>
>>>>>> Display:
>>>>>>     1: caps(FOO/tiled); constraints(alignment=64k)
>>>>>>
>>>>>> Merged Result:
>>>>>>     1: caps(FOO/tiled, FOO/CC, FOO/cached);
>>>>>> constraints(alignment=64k);
>>>>>>        transition(GPU->display: trans_a, trans_b; display->GPU: none)
>>>>>>     2: caps(FOO/tiled, FOO/CC); constraints(alignment=64k);
>>>>>>        transition(GPU->display: trans_a; display->GPU: none)
>>>>>>     3: caps(FOO/tiled); constraints(alignment=64k);
>>>>>>        transition(GPU->display: none; display->GPU: none)
>>>>>
>>>>>
>>>>>
>>>>> We definitely don't want to expose a way of getting uncached rendering
>>>>> surfaces for radeonsi. I mean, I think we are supposed to be able to
>>>>> program
>>>>> our hardware so that the backend bypasses all caches, but (a) nobody
>>>>> validates that and (b) it's basically suicide in terms of performance.
>>>>> Let's
>>>>> build fewer footguns :)
>>>>
>>>>
>>>> sure, this was just a hypothetical example.  But to take this case as
>>>> another example, if you didn't want to expose uncached rendering (or
>>>> cached w/ cache flushes after each draw), you would exclude the entry
>>>> from the GPU set which didn't have FOO/cached (I'm adding back a
>>>> cached but not CC config just to make it interesting), and end up
>>>> with:
>>>>
>>>>     trans_a: FOO/CC -> null
>>>>     trans_b: FOO/cached -> null
>>>>
>>>> GPU:
>>>>    1: caps(FOO/tiled, FOO/CC, FOO/cached); constraints(alignment=32k)
>>>>    2: caps(FOO/tiled, FOO/cached); constraints(alignment=32k)
>>>>
>>>> Display:
>>>>    1: caps(FOO/tiled); constraints(alignment=64k)
>>>>
>>>> Merged Result:
>>>>    1: caps(FOO/tiled, FOO/CC, FOO/cached); constraints(alignment=64k);
>>>>       transition(GPU->display: trans_a, trans_b; display->GPU: none)
>>>>    2: caps(FOO/tiled, FOO/cached); constraints(alignment=64k);
>>>>       transition(GPU->display: trans_b; display->GPU: none)
>>>>
>>>> So there isn't anything in the result set that doesn't have GPU cache,
>>>> and the cache-flush transition is always in the set of required
>>>> transitions going from GPU -> display
>>>>
>>>> Hmm, I guess this does require the concept of a required cap..
>>>
>>>
>>> Which we already introduced to the allocator API when we realized we
>>> would need them as we were prototyping.
>>
>>
>> Note I also posed the question of whether things like cached (and
>> similarly compression, since I view compression as roughly an equivalent
>> mechanism to a cache) in one of the open issues on my XDC 2017 slides
>> because of this very problem of over-pruning it causes.  It's on slide 15,
>> as "No device-local capabilities".  You'll have to listen to my coverage of
>> it in the recorded presentation for that slide to make any sense, but it's
>> the same thing Nicolai has laid out here.
>>
>> As I continued working through our prototype driver support, I found I
>> didn't actually need to include cached or compressed as capabilities: The
>> GPU just applies them as needed and the usage transitions make it
>> transparent to the non-GPU engines.  That does mean the GPU driver currently
>> needs to be the one to realize the allocation from the capability set to get
>> optimal behavior.  We could fix that by reworking our driver though.  At
>> this point, not including device-local properties like on-device caching in
>> capabilities seems like the right solution to me.  I'm curious whether this
>> applies universally though, or if other hardware doesn't fit the
>> "compression and stuff all behaves like a cache" idiom.
>
>
> Compression is a part of the memory layout for us: framebuffer compression
> uses an additional "meta surface". At the most basic level, an allocation
> with loss-less compression support is by necessity bigger than an allocation
> without.
>
> We can allocate this meta surface separately, but then we're forced to
> decompress when passing the surface around (e.g. to a compositor.)
>

side note:  I think this is pretty typical.. although afaict for
adreno at least, when you start getting into sampling from things with
multiple layers/levels, the meta surface needs to be interleaved with
the "main" surface, so it can't really be allocated after the fact.

Also for depth buffer, there is potentially an additional meta buffer
for lrz.  (Although cross-device depth buffer sharing seems a bit...
strange.)

> Consider also the example I gave elsewhere, where a cross-vendor tiling
> layout is combined with vendor-specific compression:
>
> Device 1, rendering: caps(BASE/foo-tiling, VND1/compression)
> Device 2, sampling/scanout: caps(BASE/foo-tiling, VND2/compression)
>
> Some more thoughts on caching or "device-local" properties below.
>
>
> [snip]
>
>>> I think I like the idea of having transitions being part of the
>>> per-device/engine cap sets, so that such information can be used upon
>>> merging to know which capabilities may remain or have to be dropped.
>>>
>>> I think James's proposal for usage transitions was intended to work
>>> with flows like:
>>>
>>>    1. App gets GPU caps for RENDER usage
>>>    2. App allocates GPU memory using a layout from (1)
>>>    3. App now decides it wants use the buffer for SCANOUT
>>>    4. App queries usage transition metadata from RENDER to SCANOUT,
>>>       given the current memory layout.
>>>    5. Do the transition and hand the buffer off to display
>>
>>
>> No, all usages the app intends to transition to must be specified up front
>> when initially querying caps in the model I assumed.  The app then specifies
>> some subset (up to the full set) of the specified usages as a src and dst
>> when querying transition metadata.
>>
>>> The problem I see with this is that it isn't guaranteed that there will
>>> be a chain of transitions for the buffer to be usable by display.
>>
>>
>> I hadn't thought hard about it, but my initial thoughts were that it would
>> be required that the driver support transitioning to any single usage given
>> the capabilities returned.  However, transitioning to multiple usages (E.g.,
>> to simultaneously rendering and scanning out) could fail to produce a valid
>> transition, in which case the app would have to fall back to a copy in that
>> case, or avoid that simultaneous usage combination in some other way.
>>
>>> Adding transition metadata to the original capability sets, and using
>>> that information when merging could give us a compatible memory layout
>>> that would be usable by both GPU and display.
>>>
>>> I'll look into extending the current merging logic to also take into
>>> account transitions.
>>
>>
>> Yes, it'll be good to see whether this can be made to work.  I agree Rob's
>> example outcomes above are ideal, but it's not clear to me how to code up
>> such an algorithm.  This also all seems unnecessary if "device local"
>> capabilities aren't needed, as posited above.
>>
>>>> although maybe the user doesn't need to know every possible transition
>>>> between devices once you have more than two devices..
>>>
>>>
>>> We should be able to infer how buffers are going to be moved around
>>> from the list of usages, shouldn't we?
>>>
>>> Maybe we are missing some bits of information there, but I think the
>>> allocator should be able to know what transitions the app will care
>>> about and provide only those.
>>
>>
>> The allocator only knows the requested union of all usages currently. The
>> number of possible transitions grows combinatorially for every usage
>> requested I believe.  I expect there will be cases where ~10 usages are
>> specified, so generating all possible transitions all the time may be
>> excessive, when the app will probably generally only care about 2 or 3
>> states, and in practice, there will probably only actually be 2 or 3
>> different underlying possible combinations of operations.
>
>
> Exactly. So I wonder if we can't just "cut through the bullshit" somehow?
>
> I'm looking for something that would also eliminate another part of the
> design that makes me uncomfortable: the metadata for transitions. This makes
> me uncomfortable for a number of reasons. Who computes the metadata? How is
> the representation of the metadata? With cross-device usages (which is the
> whole point of the exercise), this quickly becomes infeasible.
>
> So instead as a thought experiment, let's just use what we already have:
> capabilities and constraints (or properties/attributes).
>
> I kind of already outlined this with the long example in my email here
> https://lists.freedesktop.org/archives/mesa-dev/2017-December/179055.html
>
> Let me try to summarize the transition algorithm. Its inputs are:
> - the current (source) capability set
> - the desired new usages
> - the capability sets associated with these usages, as queried when the
> surface was allocated
>
> Steps of the algorithm:
>
> 1. Compute the merged capability set for the new usages (the destination
> capability set).
> 2. Compute the transition capability set, which is the merger of the source
> and destination sets.
> 3. Determine whether a "release" transition is required on the source
> device(s):
> 3a. For global properties, a transition is required if the source capability
> set is a superset of the transition set.
> 3b. For device-local properties, a transition is required if there is some
> destination device for which the device-local properties are a subset of the
> source set.
> 4. Determine whether an "acquire" transition is required on the destination
> device(s) in a similar way.
>
> Finally, execute the transitions using corresponding APIs, where the APIs
> simply receive the computed capability sets.
>
> For example, release transitions would receive the source capability set
> (and perhaps the source usages), the transition capability set, and the set
> difference of device-local capabilities, and nothing else.
>
> The point is that all steps of the algorithm can be implemented in a
> device-agnostic way in libdevicealloc, without calling into any
> device/driver callbacks.
>
> I'm pretty sure this or something like it can be made to work. We need to
> think through a lot of example cases, but at least we'll have thought them
> through, which is better than relying on some opaque metadata thing and then
> finding out later that there are some new cross-device cases where things
> don't work out because the piece of (presumably device-specific driver) code
> that computes the metadata isn't aware of them.
>
>
> [snip]
>>
>> One final note:  When I initially wrote up the capability merging logic, I
>> treated "layout" as a sort of "special" capability, basically like Nicolai
>> originally outlined above.  Miguel suggested I add the "required" bit
>> instead to generalize things, and it ended up working out much cleaner.
>> Besides the layout, there is at least one other obvious candidate for a
>> "required" capability that became obvious as soon as I started coding up the
>> prototype driver: memory location.  It might seem like memory location is a
>> simple device-agnostic constraint rather than a capability, but it's
>> actually too complicated (we need more memory locations than "device" and
>> "host").  It has to be vendor specific, and hence fits in better as a
>> capability.
>
>
> Could you give more concrete examples of what you'd like to see, and why
> having this as constraints is insufficient?
>
>
>> I think if possible, we should try to keep the design generalized to as
>> few types of objects and special cases as possible.  The more we can
>> generalize the solutions to our existing problem set, the better the
>> mechanism should hold up as we apply it to new and unknown problems as they
>> arise.
>
>
> I'm coming around to the fact that those things should perhaps live in a
> single list/array, but I still don't like the term "capability".
>
> I admit it's a bit of bike-shedding, but I'm starting to think it would be
> better to go with the generic term "property" or "attribute", and then add
> flags/adjectives to that based on how merging should work.
>
> This would include the constraints as well -- it seems arbitrary to me that
> those would be singled out into their own list.

I'm not picky about the names, but I think "constraint" has to be
something well specified, if we are to merge constraints.  If you
don't know what "pitch alignment" is how do you merge the pitch
constraints from two different devices?  This was the reason for the
separate list.

BR,
-R

> Basically, the underlying principle is that a good API would have either one
> list that includes all the properties, or one list per merging-behavior. And
> I think one single list is easier on the API consumer and easier to extend.
>
> Cheers,
> Nicolai
>
>
> --
> Lerne, wie die Welt wirklich ist,
> Aber vergiss niemals, wie sie sein sollte.