Having to always pull the physical device from the instance has been
annoying for almost as long as the driver has existed. It also won't
work in a world where we ever have more than one physical device. This
commit adds a new field called "physical" to anv_device and switches
every location where we use device->instance->physicalDevice to use the
new field instead.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3461>
All VkFoo structs are typedef'd to not need the struct keyword. Leaving
it in there is just extra characters and breaks Vulkan's aliasing when
stuff gets promoted to core versions. It's better to just never use
struct for VkFoo.
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
We already have a mechanism for specifying that we want a fixed address
provided by the driver internals. We're about to let the client start
specifying addresses in some very special scenarios as well so we want
to pass this through to the allocation function.
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Instead of dirtying all graphics or all compute based on binding point,
we're now much more careful. We first check to see if the actual
descriptor set changed and then only dirty the stages used by that
descriptor set. For dynamic offsets, we keep a bitfield per-stage of
which offsets are actually used in that stage and we only dirty push
constants and descriptors if that stage has dynamic offsets AND those
offsets actually change.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
the ASYNC flag, in particular, has the potential to help performance
because it means less sync tracking in the kernel.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
With inline uniform blocks descriptor, the meaning of descriptorCount
is a number of bytes to copy into the descriptor. Don't try to use
that size as an index into the descriptor table.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 43f40dc7cb ("anv: Implement VK_EXT_inline_uniform_block")
Gitlab: https://gitlab.freedesktop.org/mesa/mesa/issues/1195
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Later generations support bindless for samplers, images, and buffers and
thus per-stage descriptors are not limited by the binding table size.
However, gen8 doesn't support bindless images and thus needs to report a
lower per-stage limit so that all combinations of descriptors that fit
within the advertised limits are reported as supported by
vkGetDescriptorSetLayoutSupport.
Fixes test dEQP-VK.api.maintenance3_check.descriptor_set
Fixes: 79fb0d27f3 ("anv: Implement SSBOs bindings with GPU addresses in the descriptor BO")
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
In a descriptor set inline uniform blocks don't use up any bindings.
However, the presence of any inline uniform blocks doed require the
use of the descriptor buffer, which takes up one binding.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
When immutable samplers are set we call write_image_view with a NULL
image view. This causes issues on IVB where we have to fake texture
swizzling.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=110999
Fixes: d2aa65eb18 "anv: Emulate texture swizzle in the shader when..."
If descriptorType is VK_DESCRIPTOR_TYPE_STORAGE_IMAGE
or VK_DESCRIPTOR_TYPE_INPUT_ATTACHMENT, the imageView member of each
element of pImageInfo must have been created with the identity swizzle.
Fixes: d2aa65eb
Signed-off-by: Danylo Piliaiev <danylo.piliaiev@globallogic.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Now that we have the descriptor buffer mechanism, emulated texture
swizzle can be implemented in a very non-invasive way. Previous
attempts all tried to extend the push constant based image param
mechanism which was gross. This could, in theory, be done much faster
with a magic back-end instruction which does indirect MOVs but Vulkan on
IVB is already so slow this isn't going to matter much.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104355
Cc: "19.1" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
In 105002bd2d, we fixed a memory leak bug where we weren't properly
destroying descriptor when destroying/resetting a descriptor pool.
However, the only real leak that happened was that we we take a
reference to the descriptor set layout in the descriptor set and we
weren't dropping our reference. Everything else in the descriptor set
is tied to the pool itself and doesn't need to be freed on a per-set
basis. This commit changes the destroy/reset functions to only bother
walking the list of sets to unref the layouts and otherwise we just
assume that the whole-pool destroy/reset takes care of the rest.
Now that we're doing more non-trivial things with descriptor sets such
as allocating things with util_vma_heap, per-set destruction is starting
to show up on perf traces. This takes reset back to where it's supposed
to be as a cheap whole-pool operation.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
In c520f4dec9, we chose to align the sizes of descriptor set buffers to
32 bytes. We have to align the descriptor set buffer to 32B so that
it's valid for using with push constants. We align the size as well so
we don't leave lots of holes with util_vma_heap_alloc. Unfortunately,
we were only aligning it for alloc and not for free so we were still
creating piles of holes when we delete descriptor sets. This causes
terrible perf for the allocator once we've deleted piles of descriptor
sets.
This commit reworks the code so that we align the descriptor set buffer
size to 32B for both alloc and free. The result is that it takes the
new crucible vkResetDescriptorPool from 104.567719 to 2.898354 seconds.
Fixes: c520f4dec9 "anv: Add a concept of a descriptor buffer"
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=110497
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Instead of aligning and then taking inline uniforms into account, we
need to take inline uniforms into account and then align to a page.
Otherwise, we may not be aligned to a page and allocation may fail.
Fixes: 43f40dc7cb "anv: Implement VK_EXT_inline_uniform_block"
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
anv_descriptor_pool_free_set is called on the clean-up path of
anv_descriptor_set_create and the set may not have been added to the
pool's list of sets yet. While we're here, we move adding it to that
list into set_create for symmetry.
Fixes: 105002bd2d "anv: destroy descriptor sets when pool gets..."
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Previously, we were storing the per-binding create info pointer in the
immutable_samplers field temporarily so that we can switch the order in
which we walk the loop. However, now that we have multiple arrays of
structs to walk, it makes more sense to store an index of some sort.
Because we want to leave immutable_samplers as NULL for undefined
bindings, we store index + 1 and then subtract one later.
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
I missed this on the first go round. The bindingCount field of
VkDescriptorSetLayoutBindingFlagsCreateInfoEXT is allowed to be zero
which means the flags array is ignored.
Fixes: d6c9bd6e01 "anv: Put binding flags in descriptor set layouts"
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Now that everything is in place to do bindless for all resource types
except input attachments and UBOs, VK_EXT_descriptor_indexing is
"trivial".
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
This commit changes anv to put bindless handles and sampler pointers
into the descriptor buffer and use those instead of bindful when we run
out of binding table space. This "spilling" of descriptors allows to to
advertise an almost unbounded number of images and samplers.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Instead of setting it manually, call the helper. When setting
descriptor sets becomes more complicated than just setting some struct
values, this will keep immutable sampler handling correct.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
This commit adds a new way for ANV to do SSBO bindings by just passing a
GPU address in through the descriptor buffer and using the A64 messages
to access the GPU address directly. This means that our variable
pointers are now "real" pointers instead of a vec2(BTI, offset) pair.
This carries a few of advantages:
1. It lets us support a virtually unbounded number of SSBO bindings.
2. It lets us implement VK_KHR_shader_atomic_int64 which we couldn't
implement before because those atomic messages are only available
in the bindless A64 form.
3. It's way better than messing around with bindless handles for SSBOs
which is the only other option for VK_EXT_descriptor_indexing.
4. It's more future looking, maybe? At the least, this is what NVIDIA
does (they don't have binding based SSBOs at all). This doesn't a
priori mean it's better, it just means it's probably not terrible.
The big disadvantage, of course, is that we have to start doing our own
bounds checking for robustBufferAccess again have to push in dynamic
offsets.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
If the number of surfaces or samplers exceeds what we can put in a
table, we will want to spill out to bindless. There is no bindless
support yet but this gets us the basic framework that will be used by
later commits.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
This also fixes a bug where we mis-calculate maximum binding table sizes
and may return true in vkGetDescriptorSetLayoutSupport even for sets too
large to fit in a binding table.
Fixes: ddc4069122 "anv: Implement VK_KHR_maintenance3"
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
This is really where they belong; not push constants. The one downside
here is that we can't push them anymore for compute shaders. However,
that's a general problem and we should figure out how to push descriptor
sets for compute shaders. This lets us bump MAX_IMAGES to 64 on BDW and
earlier platforms because we no longer have to worry about push constant
overhead limits.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
pool->next and pool->free_list were reset before their usage in
anv_descriptor_pool_free_set
Fixes: 775aabdd "anv: destroy descriptor sets when pool gets reset"
Signed-off-by: Danylo Piliaiev <danylo.piliaiev@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
As stated in Vulkan spec:
"Resetting a descriptor pool recycles all of the resources from all
of the descriptor sets allocated from the descriptor pool back to
the descriptor pool, and the descriptor sets are implicitly freed."
This fixes dEQP-VK.api.descriptor_pool.*
Fixes: 14f6275c92 "anv/descriptor_set: add reference counting for..."
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Tested-by: Clayton Craft <clayton.a.craft@intel.com>
Fixes following leak:
==21853== 32 bytes in 1 blocks are definitely lost in loss record 2 of 20
==21853== at 0x483AB1A: calloc (vg_replace_malloc.c:762)
==21853== by 0x4C4DD7F: util_vma_heap_free (vma.c:221)
==21853== by 0x4C4D647: util_vma_heap_init (vma.c:46)
==21853== by 0x4957B9F: anv_CreateDescriptorPool (anv_descriptor_set.c:578)
Fixes: c520f4dec9 ("anv: Add a concept of a descriptor buffer")
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Patch maintains a list of sets in the pool and destroys possible
remaining sets when pool is destroyed.
As stated in Vulkan spec:
"When a pool is destroyed, all descriptor sets allocated from
the pool are implicitly freed and become invalid."
This fixes memory leaks spotted with valgrind:
==19622== 96 bytes in 1 blocks are definitely lost in loss record 2 of 3
==19622== at 0x483880B: malloc (vg_replace_malloc.c:309)
==19622== by 0x495B67E: default_alloc_func (anv_device.c:547)
==19622== by 0x4955E05: vk_alloc (vk_alloc.h:36)
==19622== by 0x4956A8F: anv_multialloc_alloc (anv_private.h:538)
==19622== by 0x4956A8F: anv_CreateDescriptorSetLayout (anv_descriptor_set.c:217)
Fixes: 14f6275c92 ("anv/descriptor_set: add reference counting for descriptor set layouts")
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
This buffer goes along side the CPU data structure and may contain
pointers, bindless handles, or any other descriptor information.
Currently, all descriptors are size zero and nothing goes in the buffer
but this commit sets up the framework we will need later.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
The descriptor set layout code in our driver has undergone many changes
over the years. Some of the fields which were once essential are now
useless or nearly so. The has_dynamic_offsets field was completely
unused accept for the code to set and hash it. The per-stage indices
were only being used to determine if a particular binding had images,
samplers, etc. The fact that it's per-stage also doesn't matter because
that binding should never be accessed by a shader of the wrong stage.
This commit deletes a pile of cruft and replaces it all with a
descriptive bitfield which states what a particular descriptor contains.
This merely describes the data available and doesn't necessarily dictate
how it will be lowered in anv_nir_apply_pipeline_layout.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
This is what we're actually storing in the descriptor set and consuming
when we bind surface states. This commit renames image_count to
image_param_count a few places and moves the decision to not count image
params on gen9+ into anv_descriptor_set.c when we build the layout.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Make them all take a device followed by a set. This is consistent
with how the actual Vulkan entrypoint parameters are laid out.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
This commit just puts the free list code together as part of the pool
instead of having it inlined into the descriptor set create code.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
We were accidentally not counting those surfaces
Fixes: ddc4069122 "anv: Implement VK_KHR_maintenance3"
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
This commit makes three changes. One is to only walk the descriptors once
and set bind map sizes at the same time as filling out the entries. The
second is to make the pass additive so that we can put stuff in the bind
map before applying the pipeline layout. Third, we switch to using
designated initializers.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Previously, we just hashed the entire descriptor set layout verbatim.
This meant that a bunch of extra stuff such as pointers and reference
counts made its way into the cache. It also meant that we weren't
properly hashing in the Y'CbCr conversion information information from
bound immutable samplers.
Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Instead of storing a BO and offset separately, use an anv_address. This
changes anv_fill_buffer_surface_state to use anv_address and we now call
anv_address_physical and pass that into ISL.
Reviewed-by: Scott D Phillips <scott.d.phillips@intel.com>