We've traditionally stored shader variants in a per-context hash table,
based on a key with many per-stage fields. On older hardware supported
by i965, there were potentially quite a few variants, as many features
had to be emulated in shaders, including things like texture swizzling.
However, on the modern hardware targeted by iris, our NOS dependencies
are much smaller. We almost always guess the correct state when doing
the initial precompile, and so we have maybe 1-3 variants. iris NOS
keys are also dramatically smaller (4 to 24 bytes) than i965's.
Unlike the classic world, Gallium also provides a single kind of object
for API shaders---pipe_shader_state aka iris_uncompiled_shader. We can
simply store a list of shader variants there. This makes it possible
to access shader variants across contexts, rather than compiling them
separately for each context, which better matches how the APIs work.
To look up variants, we simply walk the list and memcmp the keys.
Since the list is almost always singular (and rarely ever long),
and the keys are tiny, this should be quite low overhead.
We continue storing internally generated shaders for BLORP and
passthrough TCS in the per-context hash table, as they don't have
an associated pipe_shader_state / iris_uncompiled_shader object.
(There can also be many BLORP shaders, and the blit keys are large,
so having a hash table rather than a list makes sense there.)
Because iris_uncompiled_shaders are shared across multiple contexts,
we do require locking when accessing this list. Fortunately, this
is a per-shader lock, rather than a global one. Additionally, since
we only append variants to the list, and generate the first one at
precompile time (while only one context has the uncompiled shader),
we can assume that it is safe to access that first entry without
locking the list. This means that we only have to lock when we
have multiple variants, which is relatively uncommon.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7668>
There is a small gap of time where the currently bound uncompiled
shaders, and compiled shader variant, are out of sync. Specifically,
between pipe->bind_*_state() and the next draw.
Currently, shaders variants live entirely within a single context,
and when deleting an iris_uncompiled_shader, we check if any of its
variants are currently bound, and defer deleting those until the next
iris_update_compiled_shaders() hook runs and binds new shaders to
replace them. (This is due to the time gap between binding new
uncompiled shaders, and updating variants at draw time when we have
the required NOS in place.)
This works pretty well in a single context world. But as we move to
share compiled shader variants across multiple contexts, it breaks down.
When deleting a shader, we can't look at all contexts to see if its
variants are bound anywhere. We can't even quantify whether those
contexts will run a future draw any time soon, to update and unbind.
One fairly crazy solution would be to delete the variants anyway, and
leave the stale pointers to dead variants in place. This requires
removing any code that compares old and new variants. Today, we do
that sometimes for seeing if the old/new shaders toggled some feature.
Worse than that, though, we don't just have to avoid dereferences, we'd
have to avoid pointer comparisons. If we free a variant, and quickly
allocate a new variant, malloc may return the same pointer. If it's
for the same shader stage, we may get a new different program that has
the same pointer as a previously bound stale one, causing us to think
nothing had changed when we really needed to do updates. Again, this
is doable, but leaves the code fragile - we'd have to guard against
future patches adding such checks back in.
So, don't do that. Instead, do basic reference counting. When a
variant is bound in a context, up the reference. When it's unbound,
decrement it. When it hits zero, we know it's not bound anywhere and
is safe to delete, with no stale references. This ends up being
reasonably cheap anyway, since the atomic is usually uncontested.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7668>
Single-line version of MSVC warning suppression does not extend beyond
the #endif directive. Use push/disable/pop instead.
Also suppress 26452, which is a similar analysis warning.
This could also be fixed with constexpr if, but C++17 would be required.
Fixes: 790516db0b ("gallium/swr: fix gcc warnings")
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8093>
This would have prevented the bug that the previous commit fixes.
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Zoltán Böszörményi <zboszor@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8753>
marshal_call_after is ignored if the function is an alias of another
function. Move it to the right place.
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Reviewed-by: Zoltán Böszörményi <zboszor@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8753>
The dr_mode hack is now folded into the git tree. The uprev brings in a
shrinker fix for msm and a fix for the GPU_SET OOB messages on cheza
(possibly involved in piglit flakes).
Reviewed-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8768>
Now that we're looking at shader info system values rather than
vs_prog_data, there's no reason we have to do this when updating
the shader variants. We can simply check it when binding a new
shader from the API point of view.
Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8759>
brw_compile_vs sets the vs_prog_data fields based on the NIR program's
system values read info field. We can use that directly, enabling more
cleanups in the next patches.
Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8759>
No need to do the other checks in this case, because then we know that
we've done the UBWC clears and recursed on stencil and added deps on read
batches.
Done as a separate patch to reduce behavior changes in my upcoming move of
the batch cache to the context.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8729>
v2 (Jason Ekstrand):
- Separate the anv_gem interface from anv_queue internals
- Rework on top of the new anv_queue_family stuff
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8667>
v2 (Jason Ekstrand):
- Don't take an anv_physical_device in anv_gem_get_engine_info()
- Return the engine info from anv_gem_get_engine_info()
- Free the engine info in anv_physical_device_destroy()
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8667>
This may vary based on the newer kernel engines based contexts.
v2 (Jason Ekstrand):
- Initialize anv_queue::exec_flags in anv_queue_init
- Don't conflate this with refactors to get_reset_stats
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8667>
This is modeled on anv_memory_type and anv_memory_heap which we already
use for managing memory types. Each anv_queue_family contains some data
which is returned by vkGetPhysicalDeviceQueueFamilyProperties() verbatim
as well as some internal book-keeping bits. An array of queue families
along with a count is stored in the physical device. Each anv_queue
then contains a pointer to the anv_queue_family to which it belongs.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8667>
By moving vk_object_base_finish() to the end and putting the thread
clean-up in an if block we both better mimic anv_queue_init() and have a
more correct object destruction order. It comes at the cost of a level
of indentation but that seems to actually make the function more clear.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8667>
I don't know if this is a typo or an artifact of ancient versions of the
Vulkan API. In any case, it's wrong.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8667>
I originally wrote this several years ago to aid in app debugging. Now
that we have nice tools like RenderDoc, it's no longer needed. I don't
think anyone's really used it in 4 years or more.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8667>
I meant to do this years ago when I first added SHADER_OPCODE_SEND. At
the time, the only use for the extended descriptor was bindless handles
which were always one thing and never non-constant. However, it doesn't
actually require any extra instructions because we have to OR in ex_mlen
anyway.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8748>
we were using a system of block=array<uvec4> here, but we can really
just simplify this to block=array<uint> to make all the related code much
simpler
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8628>
this is a little hacky since we're still using unpacked layout for everything,
requiring that we "adjust" the value we pass back to the user for std430 to
be the expected value as though we were using packed layout
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8628>
this is actually crazy, but there's no other way to do it from the variable.
ideally, nir would have a separate type for atomic counters to simplify this
and then also stop mangling binding/block index during lower_buffers
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8628>
I meant to squash this down but didn't get around to it
Fixes: e79d905f5a ("zink: flag ssbo buffer resources as having pending writes on batch")
Reviewed-by: Adam Jackson <ajax@redhat.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8628>
Commits 0278d1fa323cf1f289..b688ea31fcf7e20436 added a new parameter
to set_vertex_buffers(), set_shader_images(), and set_sampler_views()
which specifies a number of trailing slots to unbind. They updated
the iris functions to do the unbinding, but didn't update the code
to mark which things are bound in the bitfields. This meant that
later code would assume those unbound slots were bound, and crash
on a NULL dereference. All that's needed is to add that slot count
when unbinding things in the bitfield.
Fixes: 0278d1fa32 ("gallium: add unbind_num_trailing_slots to set_vertex_buffers")
Fixes: 72ff66c3d7 ("gallium: add unbind_num_trailing_slots to set_shader_images")
Fixes: b688ea31fc ("gallium: add unbind_num_trailing_slots to set_sampler_views")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8758>
It was already parsed in intelInitScree2, and the results are stored in
the screen.
Fixes: d67ef48580 ("i965/screen: Allow drirc to set 'allow_rgb10_configs' again.")
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7387>