The following tests now pass:
dEQP-VK.texture.filtering.2d.formats.a4b4g4r4_unorm.nearest
dEQP-VK.texture.filtering.2d.formats.a4b4g4r4_unorm.linear
dEQP-VK.texture.filtering.2d.formats.a4b4g4r4_unorm.nearest_mipmap_nearest
dEQP-VK.texture.filtering.2d.formats.a4b4g4r4_unorm.linear_mipmap_nearest
dEQP-VK.texture.filtering.2d.formats.a4b4g4r4_unorm.nearest_mipmap_linear
dEQP-VK.texture.filtering.2d.formats.a4b4g4r4_unorm.linear_mipmap_linear
The corresponding tests for a4r4g4b4_unorm are still passing.
Closes: https://gitlab.freedesktop.org/nouveau/mesa/-/issues/36
Signed-off-by: Rebecca Mckeever <rebecca.mckeever@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24326>
The following tests now pass:
dEQP-VK.pipeline.monolithic.sampler.view_type.2d_unnormalized.format.b4g4r4a4_unorm_pack16.*
Signed-off-by: Rebecca Mckeever <rebecca.mckeever@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24326>
Implement nvk_CmdDispatchBase() and then delete nvk_CmdDispatch() so that
the vk_common verion with be used instead. This will avoid repeated code.
Update nvk_CmdDispatchIndirect() so that it sets the three base_group
fields of the root descriptor table to 0.
Signed-off-by: Rebecca Mckeever <rebecca.mckeever@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24326>
It's unused because Vulkan doesn't support compute dispatches with an
unknown workgroup size. If we ever do, we can add it back in. In case
it ever comes back, add an unreachable() to nvk_nir_lower_descriptors().
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24326>
Currently, any attempt to use gl_PointCoord (GLSL) or PointCoord (SPIR-V)
results in an assert in codegen:
ERROR: unknown nir_intrinsic_op load_point_coord
This patch addresses the error by converting nir_intrinsic_load_point_coord
system value into VARYING_SLOT_PNTC. The following tests now pass:
dEQP-VK.glsl.builtin_var.simple.pointcoord
dEQP-VK.glsl.builtin_var.simple.pointcoord_uniform_frag
dEQP-VK.glsl.builtin_var.simple.pointcoord_uniform_vert
Closes: https://gitlab.freedesktop.org/nouveau/mesa/-/issues/33
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24326>
It seems for Vulkan we always need square points with shader PointCoords
just like OpenGL point sprites.
From vulkan spec's point rasterization:
"Point rasterization produces a fragment for each fragment area group of
framebuffer pixels with one or more sample points that intersect a region
centered at the point’s (xf,yf).
This region is a square with side equal to the current point size.
... (xf,yf) is the exact, unrounded framebuffer coordinate of the vertex
for the point"
From OpenGL compatibility spec:
Basic point rasterization:
"If point sprites are enabled, then point rasterization produces a
fragment for each framebuffer pixel whose center lies inside a square
centered at the point’s (xw, yw), with side length equal to the current
point size.
... and xw and yw are the exact, unrounded window coordinates of the
vertex for the point"
Also in OpenGL multisampling leads to circle points as we see in the spec:
Point multisample rasterization:
"This region is a circle having diameter equal to the current point width
if POINT_SPRITE is disabled, or a square with side equal to the current
point width if POINT_SPRITE is enabled."
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24326>
Vulkan's gl_InstanceIndex is different than OpenGL gl_InstanceID.
For nvk, gl_InstanceIndex is lowered as gl_BaseInstance + gl_InstanceID in nir
code. This means we need to supply base instance to the vertex shader.
We load the value at mme draw time to a root constant, as it seems there
is no existing system value corresponding to this info.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/24326>
