Consider the flag from PPS when setting tc/beta offset.
This fixes some artifacts when decoding a hevc video,
hevc_scaling_list4.mkv from Lynne.
Signed-off-by: Hyunjun Ko <zzoon@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34782>
Those platforms requires aux map with 1MB alignment, for slab that
means that any buffer needs to have size of multiple of 1MB what
causes a lot of memory to be wasted causing it to run out of memory
when running multiple GPU applications.
Fixes: ea18572ff2 ("anv: Add support for ANV_BO_ALLOC_AUX_CCS in anv_slab_bo")
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34817>
Or current render target cache setting is to key on the binding table
index, meaning the HW associates a number in the range [0, 7] to a
RENDER_SURFACE_STATE description. If you want change the render target
0 between 2 draw calls, you need to insert a PIPE_CONTROL in between
the 2 draw calls with pb-stall + rt-flush in order to flush an writes
to a previous RENDER_SURFACE_STATE that has now becomed disassociated
with the [0, 7] number.
This PIPE_CONTROL taking care of the flush is dealt with in
cmd_buffer_maybe_flush_rt_writes(). This function diffs the current
BTI setup for render targets (first 0 to 7 BTIs) with what the next
fragment shader wants.
The issue here is we might have a render pass with 0 color attachments
and yet in 98cdb9349a we added one pointing to the render target 0,
but in the emit_binding_table() when we finally program the BTI, we
check the render pass color count and program a null surface state
instead of an actual surface state. And this leads to hangs because
the render target cache will end up with inconsistent state data.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 98cdb9349a ("anv: ensure null-rt bit in compiler isn't used when there is ds attachment")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/12955
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34603>
ANV_BO_ALLOC_MAPPED are internal allocated bos that need mmap() but as
internally we don't do any cflush() we need to make sure those are also
ANV_BO_ALLOC_HOST_COHERENT.
Checking for ANV_BO_ALLOC_HOST_CACHED could lead a cached+uncoherent
bo being allocated internally with ANV_BO_ALLOC_MAPPED.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34778>
It was failing in the first assert of anv_device_alloc_bo() because
it has ANV_BO_ALLOC_MAPPED but it don't have ANV_BO_ALLOC_HOST_COHERENT or
ANV_BO_ALLOC_HOST_CACHED(this second one is wrong and fixed in the next
patch).
LMEM is always write-combine, even SMEM on discrete GPU is always
write-back + coherent because the PCI bus protocol snooping at CPU
caches and that behavior can't be disabled.
So we can add this coherent flag without any side effects.
The ANV_BO_ALLOC_MAPPED is needed for ANV_BO_SLAB_HEAP_LMEM_SMEM
because to trigger SMEM+LMEM in anv_device_alloc_bo() we need
ANV_BO_ALLOC_MAPPED or ANV_BO_ALLOC_LOCAL_MEM_CPU_VISIBLE but the
second one is mostly used with small PCI bar discrete GPUs.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: dabb012423 ("anv: Implement anv_slab_bo and enable memory pool")
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34778>
Since RMV 1.9 pcie_family_id is checked to verify whether a
capture is supported.
Signed-off-by: Karmjit Mahil <karmjit.mahil@igalia.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34763>
Round up allocations to nearest 2MB interval if this increases
the allocation by no more than 1.33x. This reduces page count but
at the cost of extra memory consumption. Optimization only applied
to MTL(Xe KMD only)/LNL platforms, which are particularly impacted by
page misses.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: Felix DeGrood <felix.j.degrood@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33558>
BOs larger than 1MB don't go memory pool due the size but applications
tend to use a lot of VkMemory with size larger than 1MB so to reduce
the number of pages and improve performance here I'm aligning the size
of BOs larger than 1MB to 64kb, this allows 64kb pages to be used at
least on Xe KMD.
This bring substantial perfomance benefit in exchange of a small
memory waste.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33558>
MTL and newer integrated platforms has a performance gain when using
transparent huge pages, because of the fixed address requirement
we can't use slab for this case but we can change the initial pool
size to 2MB so all allocations get the transparent huge page
optimization.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33558>
I can't think in any case where that would be false, so lets drop it.
While at it, also making some variables const.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33558>
Because of the ANV_BO_ALLOC_CAPTURE flag, batch buffers were not
allowed to use memory pool.
So to workaround that here adding a new anv_bo_slab_heap heap for
cached+coherent+capture buffers with the main goal to get batch
buffer to memory pool but other buffers will as well.
For now that will only work in Xe KMD as i915 requires more changes
to support it.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33558>
The whole purpose of anv_bo_pool is to reduce the number of
gem_create/destroy calls in command buffers that is something with
a short life span.
But slab_bo/memory pool does the same with even other benefits like
doing 2MB allocations to enable THP.
So here skipping the meat of anv_bo_pool_free() to directly return
the bo to slab_bo. This change is also necessary because the way
anv_bo_pool stores freed buffers it requires that all bos has a unique
gem handle, what not true of buffer allocated by anv_slab.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Suggested-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33558>
This flag was not supported in anv_slab_bo because it is set together
with ANV_BO_ALLOC_CAPTURE and more important it has a specific VMA
range.
We can support it by adding a custom heap and allocating all bos in
the heap with all necessary flags, but because application can also
allocate those with vkAllocateMemory() here the ANV_BO_ALLOC_CAPTURE
is appended to the vkAllocateMemory() path for integrated gpu and
anv_slab_bo check if all the alloc_flags matches, because application
could choose to allocate it in a cached but not coherent memory type
for example.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33558>
This flag was not supported in anv_slab_bo because it is set together
with ANV_BO_ALLOC_CAPTURE and more important it has a specific VMA
range.
But we can easily support it by adding a custom heap with it and
allocating all bos in the heap with all necessary flags.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33558>
This changes allow us to support memory pool of bos with
ANV_BO_ALLOC_AUX_CCS set.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33558>
This is implementing the functions in anv_slab_bo and actually
enabling memory pool.
This is heavily based on Iris memory pool implementation, the main
difference is that the concept of heaps only exist in anv_slab_bo, we
have function that takes the anv_bo_alloc_flags and decides what heap
to place that bo.
Some anv_bo_alloc_flags blocks memory pool, we can relax and remove
some flags from this denied list later.
This feature can be disabled in runtime by setting
ANV_DISABLE_SLAB=true, this can help us to easily check if bugs
are due to this feature or not.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33558>
Allocating larger buffers allows KMD/HW to enable optimizations
that makes access to memory faster, also because of minimum alignment
required in some cases we allocate 4k or 64k long buffers for
usages that only needs a few bytes, wasting a lot of memory.
Memory pool takes care of both of those things and here I'm
adding the base infrastruture to implement this feature.
The next patch will implement the functions in anv_slab_bo.c, spliting
it in two to make review easier.
The idea here is take the same approach as Iris and use pb_slab.h.
In 99% of the places it will be transparent that anv_bo is actually
a slab of a larger and real anv_bo, the remaning 1% of the places are
handled here.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33558>
That will make easy to implement memory pool in the next patches as we
need to calculate the VMA aligment without the KMD alignment requirement
for memory pool.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33558>
This function will be needs in two places in the next patches.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33558>
Re-organize the configuration lists to make easier to include BFloat16
only for the Gfx125+ that support it, while keeping MTL supporting the
"lowered" configurations from pre-Gfx125.
Reviewed-by: Rohan Garg <rohan.garg@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34105>
Add debug option to show current shader type being
compiled within anv_shader_bin_create.
Signed-off-by: Michael Cheng <michael.cheng@intel.com>
Reviewed-by: Casey Bowman <casey.g.bowman@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34596>
Xe2 changed the MOCS field in few instructions, those now have a field
for the MOCS index and other the encryption enable bit but ISL returns
the combination of both aka MEMORY_OBJECT_CONTROL_STATE.
To minimize changes I have added 2 macros to extract the values
from the value returned by isl.
From all the instructions changed Mesa only make use of two, so the
other instruction will be handled in the next patch.
Cc: mesa-stable
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34592>
Not all images with DRM format modifiers use
ANV_IMAGE_MEMORY_BINDING_PRIVATE.
Reviewed-by: Nanley Chery <nanley.g.chery@intel.com>
Change-Id: Idc6bae70ec7080f96555a85dcdc0ead915b02935
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32295>
Adding a boolean "enable_64b_rt" in anv_accel_struct_header for the
interpret.py to properly decode anv_instance_leaf
Signed-off-by: Kevin Chuang <kaiwenjon23@gmail.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Sagar Ghuge <sagar.ghuge@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33047>
This change separate the encode, header, and copy shader into versions
for Xe1/2 and Xe3+, including adding compile options and handling 64bit
version of instance leaf for Xe3+.
Signed-off-by: Kevin Chuang <kaiwenjon23@gmail.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Sagar Ghuge <sagar.ghuge@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33047>
This commit moves the calculation of tri_bary out of
brw_nir_rt_load_mem_hit_from_addr(), and only do the calculation on
demand, since unorm_float_convert can be expensive. We do this for both
Xe1/2 and Xe3+ for consistency.
Signed-off-by: Kevin Chuang <kaiwenjon23@gmail.com>
Reviewed-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/33047>
Move the calculation to nir_lower_vars_to_explicit_types(). This
consolidates the check of shader_info::shared_memory_explicit_layout
in a single place instead of in all drivers.
This is motivated by SPV_KHR_untyped_pointers. Before that extension
we had essentially two modes for shared memory variables
- No layout decorations in the SPIR-V, and both internal layout and
driver location was _given by the driver_.
- Explicitly laid out, i.e. they are blocks, and decorated with Aliased.
Because they all alias, we could assign them driver location directly
to the start of the shared memory.
With the untyped pointers extension, there's a third option, to be added
by a later commit
- Explicitly laid out, i.e. they are blocks, and NOT decorated
with Aliased. Driver location is _given by the driver_. Blocks
with and without Aliased can be mixed.
The driver location of multiple blocks that don't alias depend on
alignment that is driver-specific, which we can more easily do from
the nir_lower_vars_to_explicit_types() that already has access to
a function to obtain such value.
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io> (hk)
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com> (v3dv)
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> (anv/hasvk)
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com> (panvk)
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com> (radv)
Reviewed-by: Rob Clark <robdclark@gmail.com> (tu)
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34139>
I accidentally disabled compression on CPS surfaces marked as storage or
color attachment for all platforms, when this should only be limited to
Xe.
Fixes: 80f9b6 ('anv: CPB surfaces that are used as color attachments or for stores cannot be compressed')
Signed-off-by: Rohan Garg <rohan.garg@intel.com>
Reviewed-by: Nanley Chery <nanley.g.chery@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/34297>
This patch does a couple of things to make CL integration with drivers
as seamless as possible:
- We pull in opencl-c.h and opencl-c-base.h to stop relying on system
headers.
- Parts of libcl.h are moved to new headers that are incomplete CL-safe
variants of libc headers.
- A couple of util headers are changed to remove now unnecessary
__OPENCL_VERSION__ guards and make more headers CL safe.
- Drivers now include src/compiler/libcl and use headers like
macros.h,u_math.h instead of libcl.h.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33576>
