Vulkan says you can do things like image resolves or blits on transfer
queues, but D3D only allows literal copies. We could try to emulate
a Vulkan transfer-only queue backed by multiple D3D queues, where we
use the copy queue when possible but fall back to compute when needed,
but let's wait until there's a good reason to do that...
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/23218>
Forcing bindless on is nice for apps that don't use EXT_descriptor_indexing,
but for the CTS, whenever EXT_descriptor_indexing is supported, it's used.
To be able to more thoroughly test the not-bindless path, add a debug flag
to turn it off.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/23218>
Vulkan's concept of dedicated resources is dangerously close to
D3D12's concept of a "committed" resource, where the memory and
resource are inextricably tied. This is a minor optimization,
but will start to be more important going forward with external
memory.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/22879>
These are unneeded. Events can't be used to indefinitely stall work
like you can with a semaphore or timeline semaphore. The signals
still need to happen so that execution will modify the state that
can be polled from the CPU though.
Fixes dEQP-VK.synchronization.basic.event.single_submit_multi_command_buffer
Fixes: 04fa6c71 ("dzn: Batch command lists together")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/22842>
In theory, ResolveSubresourceRegion should be able to resolve just
the depth or just the stencil. In practice, WARP had bugs, which
means that was never tested, so just do it via blits.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/22836>
Previously, custom buffer descriptors were owned by a descriptor set. Now,
custom buffer descriptors are owned by the buffer. Additionally, we respect
the app-provided sizes when they're smaller than the buffer size, even if
robustness is not enabled, so that size queries work correctly.
This new design fixes several issues:
* Descriptor set copies were broken when they involved custom descriptors,
because the original descriptor set owned the lifetime of the custom
descriptor, the new one was just borrowing it. If those lifetimes didn't
line up, problems would arise.
* A single buffer with the same sub-view placed in multiplel descriptor sets
would allocate multiple slots, when it only really needed one.
* Custom buffer descriptors now lower the base offset to 0 to allow merging
multiple overlapping (ending at the same upper bound) descriptors. Since
the shader is already doing an offset add, making it nonzero is free.
* Dynamic buffer descriptors were incorrect before. The size passed into the
descriptor set is supposed to be the size from the *dynamic* offset, not the
size from the static offset. By allocating/populating the descriptor when
placed into the set, it prevented larger offsets from working correctly. This
buffer-owned design prevents cmdbufs from having to own lifetime of custom
descriptors.
Fixes dEQP-VK.ssbo.unsized_array_length.float_offset_explicit_size
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/22639>
There's bad interactions with dynamic buffers at this point:
* Perf issues due to allocating and freeing the buffer to store indices/offsets
* Large dynamic uniform buffer offsets (above 65K) cause out-of-bounds reads
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/22371>
Cache coherent UMA implies that the GPU is reading data through the
CPU caches. Using write-combined CPU pages for such a system would
be bad, since the GPU would then be reading uncached data. One
example of such a system is WARP. This significantly improves WARP's
performance for some apps (including the CTS).
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/22225>
When operating in "bindless" mode, the device will own 2 descriptor
heaps, one for views, and one for samplers. Every time a view is
created (image view, buffer view), a slot is allocated for it out
of the device view heap for each usage type (sampled vs storage).
Then, in a future change, descriptor sets will just contain view/
sampler indices instead of actual descriptors. Instead of copying
these to a cmdbuf-owned descriptor heap, we can directly bind the
descriptor set as a buffer. We'll also modify shaders to perform
an indirection and index into the device heap.
Buffers also get views set up on creation. In a perfect world, we
could just put addresses/sizes in the descriptor set, but DXIL
doesn't support loading from addresses, we need descriptors. When
robust buffer access is disabled *or* descriptor set buffer views
reference the remainder of the buffer, we can just re-use a view
from the buffer and use an offset.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21913>
A future change is going to add descriptor heaps *to* the dzn_device,
and having 3x ID3D12Device pointers in a single object just seems
wrong. All of the callers already had a device, so just pass it
along where needed.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21913>
