Noticed renderdoc complaining about our size :
RDOC 692028: [18:08:18] vk_core.cpp(2272) - Warning -
VkPhysicalDeviceDescriptorBufferPropertiesEXT.imageCaptureReplayDescriptorDataSizeis too large at 32
(must be <= 16), can't support capture of VK_EXT_descriptor_buffer
Since we only need 2 pointers (main + private), we can shrink this to
16bytes. The 1/2 planes have a relative offset from the base.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38625>
If the library_path is just a basename like `libvulkan_lvp.so`, then we
can share the same JSON manifest like `lvp_icd.json` between all of the
architectures, like we already do for Vulkan layers. The library will
be looked up in the dynamic linker's default search path in this case,
and in practice will be found in `${libdir}`. This is how the Mesa's
EGL driver and Vulkan layers work, how Mesa is packaged in Debian 13,
and also how the Nvidia proprietary driver works; it makes installation
simpler for distros, especially on multiarch systems like Debian and
the freedesktop.org SDK.
However, if we want a separate manifest per architecture in order to
be able to write the full path into it, we still need per-architecture
filename disambiguation like `lvp_icd.x86_64.json`.
We presumably still want a separate per architecture on Windows, because
the concept of a single monolithic `${libdir}` is less common there, and
it can also be helpful during development when setting `$VK_DRIVER_FILES`
to force the use of a specific driver installed in a non-default location.
Use the following parameter to passed to vk_icd_gen:
'--icd-lib-path', vulkan_icd_lib_path,
'--icd-filename', icd_file_name,
output : 'virtio_icd.' + vulkan_manifest_suffix,
and the output is passed by '--out', '@OUTPUT@',
so we can detect vulkan_manifest_per_architecture from the --out parameter in script.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/13745
Signed-off-by: Simon McVittie <smcv@collabora.com>
Co-authored-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Mel Henning <mhenning@darkrefraction.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37314>
Without these fixes, H.265 streams using long-term references would
fail to decode correctly as the decoder wouldn't distinguish between
short-term and long-term reference frames.
Fixes: 896f95a37e ("vulkan/video: fix h265 decoding with LT enabled.")
Signed-off-by: Hyunjun Ko <zzoon@igalia.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38571>
The libvulkan_intel does not need the decoder when buildtype=release
where the debugging is disabled.
However, the decoder implementation is decided by the dep_expat
which may be turned on by like -Dtools=intel and the binary size
of libvulkan_intel increase unexpectedly.
This change creates the stub dependency and decide the exact
decoder dependency of libvulkan_intel by the buildtype.
Test: meson setup builddir -D build-tests=true -Dbuildtype=release --reconfigure && ninja -C builddir && cd builddir && meson test
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: Andy Hsu <hwandy@google.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38569>
This is the last level layer of emission, we want the tracking to be
added above that, so that when flushing of previously accumulated
reasons happens, another pointless reason isn't added.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38542>
Sadly this probably won't change anything in terms of perf as the CCS
engine has a bunch of other restrictions.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 243c01c703 ("anv/iris: implement Wa_18040903259")
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com>
Reviewed-by: Sagar Ghuge <sagar.ghuge@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38484>
Saves unncessary PC and stall during encode phase.
Thanks to Felix for pointing out that CCS always needs a CS stall once
we add a pipe control, that will kill the performance for BVH
construction.
Signed-off-by: Sagar Ghuge <sagar.ghuge@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38513>
Up to 4 reasons can be saved and displayed. Previously, we were
only displaying one reason for Perfetto.
Co-authored-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38500>
Our backend compiler explains the limits as :
32 bytes for the patch header (tessellation factors)
480 bytes for per-patch varyings (a varying component is 4 bytes and
gl_MaxTessPatchComponents = 120)
16384 bytes for per-vertex varyings (a varying component is 4 bytes,
gl_MaxPatchVertices = 32 and
gl_MaxTessControlOutputComponents = 128)
In all that's :
* 32 patches * 128 components (counting tessellation factors)
* 32 vertices * 128 components
8192 total components.
I'm not sure why the limit was set so low, maybe leftover from older platforms?
Bump the limit to something like competition.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38523>
Trying to cut down apply_pipeline_layout a bit and also allowing some
reuse for a new extension.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Acked-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38495>
Rewrite ANV's encode.comp, the final intel-specific raytracing shader
used for bvh-build. Performance is greatly improved for this shader
by adding the following features:
1) Find children early. All threads speculative find their children
before they know if they are valid (not collapsed). This makes more
work overall but reduces latency for propagating valid nodes from
root to leaves. Nodes find out if they are valid faster if all nodes
know who their children are upfront.
2) Hoist code used for intra-thread communication. Communicate
to children as soon as possible, minimizing wait time for later
threads.
3) Multithread encoding. Still launching 1 simd lane per node, same
as before, but encoding of nodes and children are parallelized across
multiple lanes. This works well because most nodes are collapsed
without any encode work required.
4) Hash globalID. Reduce chance that the thread processing a node
will also need to process node's children, which was found to
degrade performance, particularly for root node processing.
Measured RT game speedups:
* Hitman3 +48%
* F1'22 +10%
* Indiana Jones +8%
* GravityMark +2.5%
Reviewed-by: Sagar Ghuge <sagar.ghuge@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36937>
Rather than adding another boolean to optionally lower PLS vars, pass
the types we want to lowers through a nir_variable_mode bitmask.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Eric R. Smith <eric.smith@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37110>
The VMA of VkDeviceMemory has to accomodate all the resources that can
be bound to it. For sparse images it's 64KiB alignment, for other
tiled images it's 4KiB. But we also have a workaround that requires a
64KiB alignment for Tile4 images.
The initial version of the slab allocator missed the 4KiB alignment.
This fix adds the workaround handling too.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: dabb012423 ("anv: Implement anv_slab_bo and enable memory pool")
Reviewed-by: Nanley Chery <nanley.g.chery@intel.com>
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38480>
This helps for debug when wanting to check which pipeline mode the
driver has selected for a given section of a frame.
Signed-off-by: Casey Bowman <casey.g.bowman@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38317>
These vertex fetch flushes aren't required in gen9+ because the display
driver will take care of this invalidation on QueueSubmit. So let's remove it.
Signed-off-by: Casey Bowman <casey.g.bowman@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38317>
Use mesa_logi_v(..) in sparse_debug(..).
Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38190>
Also ensure the printfs are read even if the device is lost or ran
into a fault.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Ivan Briano <ivan.briano@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38358>
This is done by grep ALIGN( to align(
docs,*.xml,blake3 is excluded
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38365>
When the input is size_t, cast the input to uint32_t, as the output is expect uint32_t not size_t
Signed-off-by: Yonggang Luo <luoyonggang@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38365>
We add a bunch of new helpers to avoid the need to touch >parent_instr,
including the full set of:
* nir_def_is_*
* nir_def_as_*_or_null
* nir_def_as_* [assumes the right instr type]
* nir_src_is_*
* nir_src_as_*
* nir_scalar_is_*
* nir_scalar_as_*
Plus nir_def_instr() where there's no more suitable helper.
Also an existing helper is renamed to unify all the names, while we're
churning the tree:
* nir_src_as_alu_instr -> nir_src_as_alu
..and then we port the tree to use the helpers as much as possible, using
nir_def_instr() where that does not work.
Acked-by: Marek Olšák <maraeo@gmail.com>
---
To eliminate nir_def::parent_instr we need to churn the tree anyway, so I'm
taking this opportunity to clean up a lot of NIR patterns.
Co-authored-by: Konstantin Seurer <konstantin.seurer@gmail.com>
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38313>
This is what happens when you leave MR unreviewed for months.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: d39e443ef8 ("anv: add infrastructure for common vk_pipeline")
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38400>
Use tristate for the aligned setting, otherwise it is always
first disabled which contributes to the condition if we set the
new stride active.
v2: set ByteStride in dword units and take secondary cmdbuf
in to account (Lionel)
Cc: mesa-stable
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Tested-by: Nataraj Deshpande <nataraj.deshpande@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38349>
Drivers already have to track this workaround, so remove the logic
from Blorp and let the driver manage this.
Also in Anv don't accumulate this workaround, emit it directly in
place right after COMPUTE_WALKER. Accumulating can be problematic when
you want to dispatch concurrent compute shaders that do not need any
cache flush interaction (typical example with the internal
simple_shader framework).
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 3e0ad0176b ("anv: Emit state cache invalidation after every compute dispatch")
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com>
Reviewed-by: Sagar Ghuge <sagar.ghuge@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38306>
CmdWriteAccelerationStructuresPropertiesKHR writes the data with MI
commands, we no longer dispatch shaders to write the properties.
As a result, we don't need to flush untyped cache.
Fixes: f0e18c475b ("intel: remove GRL/intel-clc")
Signed-off-by: Sagar Ghuge <sagar.ghuge@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38291>
Refactor the intermediate buffer copy path to use a generic callback
approach, making the code more maintainable and easier to extend with
new format conversions.
The core copy_intermediate() function is now format-agnostic, accepting
a conversion callback that handles the actual data transformation. This
moves format-specific logic (RGB<->RGBA conversion and ASTC
decompression) into dedicated callback functions, making the conversion
path explicit at each call site rather than hidden inside the copy
function.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37691>
Instead of allocating a buffer for the entire RGB->RGBA conversion
process. Just allocate a smaller buffer that is the size of a tile and
do the conversion one tile at a time.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37691>
The needs_temp_copy() function was incorrectly identifying some
depth/stencil formats as needing RGB<->RGBA conversion.
VK_FORMAT_D32_SFLOAT_S8_UINT maps to PIPE_FORMAT_Z32_FLOAT_S8X24_UINT,
which has 3 channels (F32 depth, UP8 stencil, X24 padding). The
component count check (== 3) was matching this as an RGB color format,
causing depth/stencil images to incorrectly use the RGB conversion path.
Add an explicit vk_format_is_depth_or_stencil() check before the
component count test to ensure depth/stencil formats always use the
direct copy path.
Fixes: f97b51186f ("anv: intermediate RGB <-> RGBX copy for HIC")
Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37691>
The code in question multiplies `uint32_t`s together and assigns them to
a `uint64_t`. It seems rather unlikely at there would be an overflow,
but we might as well do the cast.
CID: 1649587
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/38289>
Our exposed limits say we shouldn't be able to, but let's add an assert
in case something changes, and to help Coverity out.
CID: 1662103
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37583>
