v3dv: create the command buffer BO set before we init CLs

The init of the CLs doesn't create any BOs for now, but this looks
like a sensible decision to future-proof things.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkEndCommandBuffer

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: emit the render command list

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: emit scissor to render area and precompute hw color clear values

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: precompute more tiling info at framebuffer creation time

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: plug leak when destroying device

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: make v3dv_layer_offset public

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkDestroyImage and vkDestroyImageView

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: revert the decision that the command buffer takes ownership of BOs

The CLs in the command buffer will reference BOs allocated by the application
such as images and buffers, that will be destroyed by the application, so
destroying them with the command buffer won't be correct.

For now, let's just assume that the comman buffer only owns the BOs
it explicitly allocates and that other abstractions own their own
BOs and are responsible from freeing them.

In the future, we might consider a ref/unref system similar to v3d's, but
for now this should work.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: add the tile state and alloc BOs to the command buffer BO list

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: store base mip level in the image view

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: add a few more API stubs

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: make the command buffer own the command list BOs

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkCmdBeginRenderPass

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: compute tile size for framebuffer

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: make v3dv_bo_alloc allocate memory for the bo struct

This is just going to be more convenient going forward.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkGetPhysicalDeviceImageFormatProperties

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: start handling command buffer status

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkBeginCommandBuffer

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: add a concept of a command list

Just barebones for now, will expand as necessary as we start emitting
actual commands into command lists.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: create a v3dv_bo struct and reference it from v3dv_device_memory

So we have a lower level representation of a buffer object that we can
manipulate that is not tied to a Vulkan representation of memory. This
will be useful as we start allocating driver internal buffers, such as
command lists.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vk{Allocate,Free}CommandBuffers

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkCreateCommandPool

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vk{Create,Destroy}Framebuffer

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vk{Create,Destroy}RenderPass

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkCreateRenderPass

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkBindBufferMemory

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkGetBufferMemoryRequirements

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vk{Create,Destroy}Buffer

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkCreateImageView

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkBindImageMemory

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkUnmapMemory

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkMapMemory

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkFreeMemory

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkAllocateMemory

Removed MAX_MEMORY_ALLOCATION_SIZE definition because that comes from
VK_KHR_maintenance3 which we don't support yet. We can care about that
when we iplement it.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: initialize mememory heaps in the physical device

When we allocate memory we will be referring to memory types by index
so this is more convenient.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkGetImageMemoryRequirements

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: Create/DestroyShaderModule implementation

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: stubs for graphics pipeline methods

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv/debug: add v3dv_debug

Note that for the Vulkan driver we are also reusing v3d_debug.h. This
file is for debug utilities exclusive to the Vulkan driver.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv/debug: plug v3d_debug

Right now there is not a big reason/difference to implement the
utilities present at v3d_debug for the vulkan driver, so lets just
reuse it.

The other advantage is that is the debug utilities used by common
parts of the driver, like broadcom/compiler

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkCreateImage

This relies heavily in infrastructure taken from the v3d driver. We should
probably look for ways to share the code between both drivers by creating
a surface layout library that we can use from both, or at least moving
parts of the v3d driver to broadcom/common. Specifically:

We take v3d_tiling.c, which requires gallium's pipe_box type for some
helper functions that we don't quite need yet.

We copied and adapted bits of v3d_resource.c into v3dv_image.c, however,
it should be possible to look for ways to reuse the code instead of
duplicating it.

Pre-compute UIF padding into the slice setup. This is different from
what we do in v3d (we do this at cerate_surface time), but it is
more convenient for us to pre-calculate it here for all mipmap
slices.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: add stub for vkDeviceWaitIdle

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkEnumerate{Instance,Device}LayerProperties

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: add some basic support for format properties

This takes a subset of vk_format_info.h from Anvil which has some
Intel specific elements. At some point we might want to discuss
if we want to make the file reusable and move the intel bits to
some other place, but it is not a lot of code and for now this works,
so we keep going.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: add stubs for the format table and vkGetPhysicalDeviceFormatProperties

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: add dummy implementations for the packet definition generator

For now we are only interested in being able to include the header
file for format definitions, so this is enough. When we start actually
emitting packets we will need to provide proper hooks.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkGetDeviceQueue

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkCreateDevice

We will probably have to amend this as we make progress with the
implementation, but for now this is enough.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkGetPhysicalDeviceQueueFamilyProperties

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: amend vkEnumerateInstanceExtensionProperties to handle layers

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkEnumerateDeviceExtensionProperties

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkGetPhysicalDeviceFeatures

For now we don't support any optional feature, we will turn them on
as they get implemented.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkGetPhysicalDeviceMemoryProperties

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: add a comment to clarify how we should implement uuid / deviceID retrieval

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: retrieve device name from device info

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: add support to use v3d simulator

v2: use spaces on both sides of ':'

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: Implement vkGetPhysicalDeviceProperties

This is just something to get us going. We will need to tune all these
limits as we make progress with driver features.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: pretend to initialize a physical device

Just to keep us moving forward for now. Later, we should probably
revisit this after running on real hardware or after enabling
the simulator.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: memory management stubs

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkEnumeratePhysicalDevices

Not an actual implementation since this doesn't initialize any actual
physical devices just yet.

Also, this doesn't check that available decices are really compatible
with the driver for now. This is for convenience, so we can move
past this point even if we are not running on actual hardware.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkDestroyInstance

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: implement vkCreateInstance

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: add support for valgrind macros

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: add support for VK_EXT_debug_report

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

gitlab-ci: add broadcom vulkan driver

At this point we are only interested to test if it keeps building.

v2: include meson-arm64 (Eric)

Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

v3dv: add v3d vulkan driver skeleton

Initial commit, mostly a import of the minimum from anv/radv to get a
skeleton to start to work with.

In includes:
 * meson files
 * Copy & adapt entrypoints ane extensions scripts from anv (that were
   later used on radv)

This is a firt approach, but is is likely that we can remove/simplify
some things.

v2: fix copyright character at broadcom/vulkan/meson.build (Eric)
v3: no spaces inside arrays (Dylan)
v4: add gnu_symbol_visibility (detected by CI on first Merge attemp)

Reviewed-by: Eric Anholt <eric@anholt.net>
squash! v3dv: add v3d vulkan driver skeleton

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6766>

glsl: Initialize ir_to_mesa_visitor members in constructor.

Fix defects reported by Coverity Scan.

Uninitialized pointer field (UNINIT_CTOR)
uninit_member: Non-static class member ctx is not initialized in this constructor nor in any functions that it calls.
uninit_member: Non-static class member prog is not initialized in this constructor nor in any functions that it calls.
uninit_member: Non-static class member shader_program is not initialized in this constructor nor in any functions that it calls.
uninit_member: Non-static class member options is not initialized in this constructor nor in any functions that it calls.

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7071>

svga: Remove unused printf argument.

Fix defect reported by Coverity Scan.

Extra argument to printf format specifier (PRINTF_ARGS)
extra_argument: This argument was not used by the format string: tex->handle.

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Neha Bhende <bhenden@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7072>

util/xmlconfig: Initialize xmlconfig member options in constructor.

Fix defects reported by Coverity Scan.

Uninitialized pointer field (UNINIT_CTOR)
uninit_member: Non-static class member field options.info is not
initialized in this constructor nor in any functions that it calls.
uninit_member: Non-static class member field options.values is not
initialized in this constructor nor in any functions that it calls.
uninit_member: Non-static class member field options.tableSize is not
initialized in this constructor nor in any functions that it calls.

Suggested-by: Eric Engestrom <eric@engestrom.ch>
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Acked-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6929>

zink: do not report SSBOs as halfway supported

We don't support SSBOs yet, as we don't expose the
PIPE_CAP_SHADER_BUFFER_OFFSET_ALIGNMENT-cap. But the gallium
state-tracker's limit-calculation gets confused by this half-way
support, and ends up thinking we can support atomics, which we
don't support yet either.

So let's not confuse the state-tracker here, and let's introduce this
again when SSBOs are actually supported.

Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7113>

scons: fix SPIR-V -> NIR build

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Tested-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Fixes: 18f9fc919e1 ('spirv: add and use a generator id enum')
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7096>

aco: fix get_ssbo_size with a vgpr resource

The result of load_vulkan_descriptor is passed directly to get_ssbo_size.
This caused convert_pointer_to_64_bit() to skip creating a
v_readfirstlane_b32 if it was necessary.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Fixes: 05b6612b4ec ('radv: do not lower UBO/SSBO access to offsets')
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/3628
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7095>

nir/opt_uniform_atomics: don't optimize atomics twice

Applications sometimes already do this optimization themselves.

fossil-db (Navi):
Totals from 51 (0.04% of 135946) affected shaders:
CodeSize: 507484 -> 501860 (-1.11%)
Instrs: 99635 -> 98471 (-1.17%)
Cycles: 2421944 -> 2414780 (-0.30%)

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6558>

nir/opt_uniform_atomics: optimize image atomics

fossil-db (Navi):
Totals from 65 (0.05% of 135946) affected shaders:
SGPRs: 3792 -> 3784 (-0.21%)
VGPRs: 2784 -> 2716 (-2.44%)
CodeSize: 707492 -> 713080 (+0.79%)
MaxWaves: 873 -> 887 (+1.60%)
Instrs: 133376 -> 134524 (+0.86%)
Cycles: 3004772 -> 3011440 (+0.22%)

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6558>

aco: use nir_opt_uniform_atomics

Significantly improves performance of a Control compute shader. Also seems
to increase FPS at the very start of the game by ~9% (RX 580, 1080p,
medium settings, no MSAA).

fossil-db (Navi):
Totals from 315 (0.23% of 135946) affected shaders:
SGPRs: 18296 -> 18336 (+0.22%); split: -0.26%, +0.48%
VGPRs: 11856 -> 11844 (-0.10%); split: -0.81%, +0.71%
CodeSize: 2233800 -> 2457508 (+10.01%)
MaxWaves: 4506 -> 4497 (-0.20%); split: +0.04%, -0.24%
Instrs: 438766 -> 486215 (+10.81%); split: -0.00%, +10.81%
Cycles: 7880180 -> 8963340 (+13.75%); split: -0.00%, +13.75%

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6558>

nir: add pass to optimize uniform atomics

This optimizes atomics with a uniform offset so that only one atomic
operation is done in the subgroup.

For shaders which do a very large amount of atomics, this can
significantly improve performance.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6558>

nir: allow divergence information to be updated when inserting instruction

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6558>

nir: move divergence analysis options to nir_shader_compiler_options

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6558>

aco: implement last_invocation

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6558>

nir: add last_invocation intrinsic

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6558>

radv/aco,nir/lower_subgroups: don't lower elect

ACO can implement this better.

fossil-db (Navi):
Totals from 33 (0.02% of 135946) affected shaders:
SGPRs: 1736 -> 1744 (+0.46%)
VGPRs: 1680 -> 1656 (-1.43%)
CodeSize: 246160 -> 245916 (-0.10%); split: -0.14%, +0.04%
MaxWaves: 449 -> 461 (+2.67%)
Instrs: 48301 -> 48266 (-0.07%); split: -0.12%, +0.05%
Cycles: 469740 -> 469240 (-0.11%); split: -0.18%, +0.08%

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6558>

aco: implement elect

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6558>

aco: optimize more uniform reductions/scans

Uniform atomic optimization will create these.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6558>

nir: update ubo locations in nir_lower_uniforms_to_ubo

locations are important for these because they provide info about how
many block indices each ubo takes up

UBO arrays have nonzero values here. all non-array UBOs have either 0
for the base or nonzero for an io lowered block at an offset,
but only arrays need to be changed here because they're the only ones
with absolute values, whereas all the others are relative.

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6272>

glsl: fix up location setting for variables pointing to a UBO's base

while linking uniforms, we might get a variable which is the only reference
to the ubo (i.e., offset 0), as determined by its type being the UBO's
interface_type, at which point we can assign the previously-gotten
block index to this variable's location

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5831>

zink: ARB_uniform_buffer_object is now implemented, so add cap and feature doc

fixes mesa/mesa#2872

Reviewed-By: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7079>

zink: support loading any UBO

Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7079>

zink: support non-const offsets

Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7079>

zink: use nir_lower_ubo_vec4 to simplify things a bit

Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7079>

aco: implement missing nir_op_unpack_half_2x16_split_{x,y}_flush_to_zero

SPIRV->NIR emits nir_op_unpack_half_2x16_flush_to_zero instead of
nir_op_unpack_half_2x16 if the shader enables denorm flush to zero
for 16-bit floating point.

This doesn't fix anything known and CTS doesn't have tests.

Fixes: 56d9bcdded8 ("radv: enable more float_controls features")
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6939>

android: fix SPIR-V -> NIR build

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Mauro Rossi <issor.oruam@gmail.com>
Fixes: 18f9fc919e1 ('spirv: add and use a generator id enum')
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7097>

docs: fix 20.2.0 relnotes

Fixes: ddad8d9c983e04267115 ("docs: add release notes for 20.2.0")
Signed-off-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7107>

docs: fix release calendar

Fixes: 7d72110a1ceb38f7032b ("docs/release-calender: Add 20.2 stable releases")
Signed-off-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7107>

docs: fix relnotes index

Add missing `.0` to 20.2.0 and fix bad merge.

Fixes: 63f61b084a5310c58eaf ("docs: update calendar and link releases notes for 20.2")
Signed-off-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7107>

radv: add missing u_atomic.h include

Fixes: 7568c97df14f8702efcc ("radv: Use atomics to read query results.")
Signed-off-by: Eric Engestrom <eric@engestrom.ch>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7050>

radv: Fix mipmap extent adjustment on GFX9+.

With arrays we really have to use the correct size for the base
mipmap to get the right array pitch. In particular, using
surf_pitch results in pitch that is bigger than the base mipmap
and hence results in wrong pitches computed by the HW.

It seems that on GFX9 this has mostly been hidden by the epitch
provided in the descriptor but this is not something we do on
GFX10 anymore.

Now this has some draw-backs:

1. normalized coordinates don't work
2. Bounds checking uses slightly bigger bounds.

2 mostly is not an issue as we still ensure that they're within
the texture memory and not overlapping other layers/mips, but
we can't properly ignore writes.

1 is kinda dead in the water ... On the other hand I'd argue that
using normalized coords & a filter for sampling a block view of
a compressed format is extraordinarily useless.

The old method we employed already had these drawbacks for everything
except the base miplevel of the imageview.

AFAICT this is the same tradeoff AMDVLK makes and no CTS test hits
this. (once it does I think the HW is dead in the water ... Only
workaround I can think of is shader processing which is hard because
we don't know texture formats at compile time.)

I also removed the extra calculations when the image has only 1 mip
level because they ended up being a no-op in that case.

CC: mesa-stable
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/2292
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/2266
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/2483
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/2906
Gitlab: https://gitlab.freedesktop.org/mesa/mesa/-/issues/3607
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7090>

zink: enable VK_EXT_shader_stencil_export

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7099>

zink: add VK_EXT_pipeline_creation_cache_control

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7099>

zink: support VK_EXT_extended_dynamic_state

just the screen parts

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7099>