| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
| |
Fixes spec/arb_enhanced_layouts/execution/component-layout/vs-fs-array-dvec3.
v2: Remove nir_outputs field from fs_visitor (caught by Tim and Iago).
Cc: mesa-stable@lists.freedesktop.org
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Timothy Arceri <timothy.arceri@collabora.com>
(cherry picked from commit 59864e8e02057cc6fa0448a8af067a3cf53389da)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
brw_link.cpp:76:44: warning: unused parameter ‘shader_type’ [-Wunused-parameter]
gl_shader_stage shader_type,
^
brw_nir.c: In function ‘brw_nir_lower_vs_inputs’:
brw_nir.c:194:55: warning: unused parameter ‘devinfo’ [-Wunused-parameter]
const struct gen_device_info *devinfo,
^
brw_vec4_visitor.cpp:914:37: warning: unused parameter ‘sampler’ [-Wunused-parameter]
uint32_t sampler,
^
brw_vec4_visitor.cpp:1146:34: warning: unused parameter ‘stream_id’ [-Wunused-parameter]
vec4_visitor::gs_emit_vertex(int stream_id)
^
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
VC4 was running into a major performance regression from enabling control
flow in the glmark2 conditionals test, because of short if statements
containing an ffract.
This pass seems like it was was trying to ensure that we only flattened
IFs that should be entirely a win by guaranteeing that there would be
fewer bcsels than there were MOVs otherwise. However, if the number of
ALU ops is small, we can avoid the overhead of branching (which itself
costs cycles) and still get a win, even if it means moving real
instructions out of the THEN/ELSE blocks.
For now, just turn on aggressive flattening on vc4. i965 will need some
tuning to avoid regressions. It does looks like this may be useful to
replace freedreno code.
Improves glmark2 -b conditionals:fragment-steps=5:vertex-steps=0 from 47
fps to 95 fps on vc4.
vc4 shader-db:
total instructions in shared programs: 101282 -> 99543 (-1.72%)
instructions in affected programs: 17365 -> 15626 (-10.01%)
total uniforms in shared programs: 31295 -> 31172 (-0.39%)
uniforms in affected programs: 3580 -> 3457 (-3.44%)
total estimated cycles in shared programs: 225182 -> 223746 (-0.64%)
estimated cycles in affected programs: 26085 -> 24649 (-5.51%)
v2: Update shader-db output.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> (v1)
|
|
|
|
|
|
| |
Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
|
|
|
|
| |
Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
|
|
|
|
| |
Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
|
|
|
| |
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
| |
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Generated by:
sed -i -e 's/brw_device_info/gen_device_info/g' src/intel/**/*.c
sed -i -e 's/brw_device_info/gen_device_info/g' src/intel/**/*.h
sed -i -e 's/brw_device_info/gen_device_info/g' **/i965/*.c
sed -i -e 's/brw_device_info/gen_device_info/g' **/i965/*.cpp
sed -i -e 's/brw_device_info/gen_device_info/g' **/i965/*.h
Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The problem with the current approach is that driver output locations
are represented as a linear offset within the nir_outputs array, which
makes it rather difficult for the back-end to figure out what color
output and index some nir_intrinsic_load/store_output was meant for,
because the offset of a given output within the nir_output array is
dependent on the type and size of all previously allocated outputs.
Instead this defines the driver location of an output to be the pair
formed by its GLSL-assigned location and index (I've borrowed the
bitfield macros from brw_defines.h in order to represent the pair of
integers as a single scalar value that can be assigned to
nir_variable_data::driver_location). nir_assign_var_locations is no
longer useful for fragment outputs.
Because fragment outputs are now allocated independently rather than
within the nir_outputs array, the get_frag_output() helper becomes
necessary in order to obtain the right temporary register for a given
location-index pair.
The type_size helper passed to nir_lower_io is now type_size_dvec4
rather than type_size_vec4_times_4 so that output array offsets are
provided in terms of whole array elements rather than in terms of
scalar components (dvec4 is the largest vector type supported by the
GLSL so this will cause all individual fragment outputs to have a size
of one regardless of the type).
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This fixes the following piglit tests on gen6+:
tex-miplevel-selection textureProjGradOffset 2DRect
tex-miplevel-selection textureGradOffset 2DRect
tex-miplevel-selection textureGradOffset 2DRectShadow
tex-miplevel-selection textureProjGradOffset 2DRect_ProjVec4
tex-miplevel-selection textureProjGradOffset 2DRectShadow
Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Cc: "12.0" <mesa-dev@lists.freedesktop.org>
|
|
|
|
|
|
|
|
|
|
|
| |
Rather than trying to work out the total number of components
used at a location we simply treat all outputs as vec4s. This
removes the need for complex code looping over varyings to match
packed locations and the need for storing the total number of
components used at each location.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This eliminates the need to walk the list of input variables, recurse
into their types (via logic largely redundant with nir_lower_io), and
interpolate all possible inputs up front. The backend no longer has
to care about variables at all, which eliminates complications from
trying to pack multiple variables into the same location. Instead,
each intrinsic specifies exactly what's needed.
This should unblock Timothy's work on GL_ARB_enhanced_layouts.
Each load_interpolated_input intrinsic corresponds to PLN instructions,
while load_barycentric_at_* intrinsics correspond to pixel interpolator
messages. The pixel/centroid/sample barycentric intrinsics simply refer
to payload fields (delta_xy[]), and don't actually generate any code.
Because we use a single intrinsic for both centroid-qualified variables
and interpolateAtCentroid(), they become indistinguishable. We stop
sending pixel interpolator messages for those, and instead use the
payload provided data, which should be considerably faster.
On Broadwell:
total instructions in shared programs: 9067751 -> 9067570 (-0.00%)
instructions in affected programs: 145902 -> 145721 (-0.12%)
helped: 422
HURT: 209
total spills in shared programs: 2849 -> 2899 (1.76%)
spills in affected programs: 760 -> 810 (6.58%)
helped: 0
HURT: 10
total fills in shared programs: 3910 -> 3950 (1.02%)
fills in affected programs: 617 -> 657 (6.48%)
helped: 0
HURT: 10
LOST: 3
GAINED: 3
The differences mostly appear to be slight changes in MOVs.
v2: Use nir_shader_compiler_options::use_interpolated_input_intrinsics
flag rather than passing it directly to nir_lower_io. Use the
unreachable() macro rather than assert in one place. (Review
feedback from Chris Forbes.)
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chris Forbes <chrisforbes@google.com>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
|
|
|
|
| |
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
|
|
|
|
|
|
| |
type_size_vec4_times_4() was introduced as a fix in 8dcf807cb43383
however since 3810c1561 we can just use type_size_scalar() and
get the actual number of outputs we need.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
|
|
|
|
|
|
| |
This way it's no longer part of libi965_compiler.la since it depends on
GLSL and ARB program stuff.
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
|
|
|
|
|
| |
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
|
|
|
|
|
|
|
|
| |
Right now libglsl.la depends on libnir.la so putting it in libnir.la
adds a dependency on libglsl.la that goes the wrong direction.
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
|
|
|
|
| |
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
For cull distance GLSL will let unsized unused arrays get
into the backend, we should nuke those straight away, to
save caring about them later.
This fixes:
arb_separate_shader_objects/linker/large-number-of-unused-varyings
as a side effect (even without culling changes).
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This handles gl_FragCoord transformations and other window system vs.
user FBO coordinate system flipping by multiplying/adding uniform
values, rather than recompiles.
This is much better because we have no decent way to guess whether
the application is going to use a shader with the window system FBO
or a user FBO, much less the drawable height. This led to a lot of
recompiles in many applications.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
|
|
|
|
|
|
|
| |
Input attributes can require 2 vec4 or 1 vec4 depending on whether they
are double-precision or not.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
|
|
|
|
|
| |
Double-precision types require 1 slot in VUE for double and dvec2, and 2 slots for
anything else.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
|
|
|
| |
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
|
|
|
|
|
|
| |
Since it will gain support to lower inputs, give it a more generic name.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
|
|
|
|
|
|
|
|
|
|
| |
Going to convert this pass to parameterized lower_io_to_temporaries, and
we want the user to be able to specify whether to lower outputs or
inputs or both. The restriction of running this pass before validate
to avoid output reads no longer applies.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
|
|
|
|
|
| |
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
|
|
|
|
|
|
|
|
|
|
| |
v2: also lower trunc, ceil, floor, fract and roundEven (Iago)
v3: also lower mod for doubles (Sam)
Signed-off-by: Iago Toral Quiroga <itoral@igalia.com>
Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
|
|
|
|
|
|
|
|
|
| |
This matches the "foreach x in container" pattern found in many other
programming languages. Generated by the following regular expression:
s/nir_foreach_function(\([^,]*\),\s*\([^,]*\))/nir_foreach_function(\2, \1)/
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
|
|
|
|
|
|
|
|
|
|
|
| |
This matches the "foreach x in container" pattern found in many other
programming languages. Generated by the following regular expression:
s/nir_foreach_instr(\([^,]*\),\s*\([^,]*\))/nir_foreach_instr(\2, \1)/
and similar expressions for nir_foreach_instr_safe etc.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
|
|
|
|
| |
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
|
|
|
|
| |
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
|
|
|
|
|
|
|
| |
The old version of the pass only worked on globals and locals and always
left inputs, outputs, uniforms, etc. alone.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This makes the extra multiply visible to NIR's algebraic optimizations
(for constant reassociation) as well as constant folding. This means
that when the result of sin/cos are multiplied by an constant, we can
eliminate the extra multiply altogether, reducing the cost of the
workaround.
It also means we only have to implement it one place, rather than in
both backends.
This makes INTEL_PRECISE_TRIG=1 cost nothing on GPUTest/Volplosion,
which has a ton of sin() calls, but always multiplies them by an
immediate constant. The extra multiply gets folded away.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
|
|
|
|
|
|
|
|
|
|
| |
I want to be able to read other fields.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
|
|
|
|
|
| |
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Some opcodes need explicit bitsizes, and sometimes we need to use the
double version when constant folding.
v2: fix output type for u2f (Iago)
v3: do not change vecN opcodes to be float. The next commit will add
infrastructure to enable 64-bit integer constant folding so this is isn't
really necessary. Also, that created problems with source modifiers in
some cases (Iago)
v4 (Jason):
- do not change bcsel to work in terms of floats
- leave ldexp generic
Squashed changes to handle different bit sizes when constant
folding since otherwise we would break the build.
v2:
- Use the bit-size information from the opcode information if defined (Iago)
- Use helpers to get type size and base type of nir_alu_type enum (Sam)
- Do not fallback to sized types to guess bit-size information. (Jason)
Squashed changes in i965 and gallium/nir drivers to support sized types.
These functions should only see sized types, but we can't make that change
until we make sure that nir uses the sized versions in all the relevant places.
A later commit will address this.
Signed-off-by: Iago Toral Quiroga <itoral@igalia.com>
Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
|
|
|
|
|
| |
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
|
|
|
|
|
|
|
| |
The same code appeared in both branches; pull it above the if statement.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
|
|
|
|
|
|
|
|
| |
The caller already computes it. Now that we have stage specific
functions, it's really easy to pass this in.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
|
|
|
|
|
|
|
|
| |
The caller already computes it. Now that we have stage specific
functions, it's really easy to pass this in.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Now that each stage is directly calling brw_nir_lower_io(), and we have
per-stage helper functions, it makes sense to just call the relevant one
directly, rather than going through multiple switch statements.
This also eliminates stupid function parameters, such as the two that
only apply to vertex attributes.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
|
|
|
|
|
|
|
|
| |
These functions are both giant switch statements where most cases don't
overlap at all. Let's put the bulk of the work in per-stage helpers.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
|
|
|
|
|
|
|
|
|
| |
Most cases already call nir_lower_io explicitly for input and output
lowering. This catch all isn't very useful anymore - we can just add it
to the remaining cases.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
|
|
|
|
|
|
|
|
|
|
| |
This simplifies things. Every caller of brw_nir_lower_io() immediately
calls brw_postprocess_nir(). The only real change this will have is
that we get an extra brw_nir_optimize() call when compiling compute
shaders, but that seems fine.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We've now hit literally every case other than geometry shaders (and
compute shaders, but those are a no-op). So, let's just move geometry
shaders over too and be done with it.
The only advantage to doing this at link time was to save the expense
of running the pass on recompiles. But we're already running a lot of
passes, and the extra code complexity isn't worth it.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
The Vulkan driver wants to be able to delete fragment outputs that are
beyond key.nr_color_regions; this is a lot easier if we lower outputs at
specialization time rather than link time.
(Rationale added to commit message by Ken)
Signed-off-by: Jason Ekstrand <jason.ekstrand@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch re-implements the pre-Haswell VS attribute workarounds.
Instead of emitting shader code in the vec4 backend, we now simply
call a NIR pass to emit the necessary code.
This simplifies the vec4 backend. Beyond deleting code, it removes
the primary use of ATTR as a destination. It also eliminates the
requirement that the vec4 VS backend express the ATTR file in terms
of VERT_ATTRIB_* locations, giving us a bit more flexibility.
This approach is a little different: rather than munging the attributes
at the top, we emit code to fix them up when they're accessed. However,
we run the optimizer afterwards, so CSE should eliminate the redundant
math. It may even be able to fuse it with other calculations based on
the input value.
shader-db does not handle non-default NOS settings, so I have no
statistics about this patch.
Note that the scalar backend does not implement VS attribute
workarounds, as they are unnecessary on hardware which allows SIMD8 VS.
v2: Do one multiply for FIXED rescaling and select components from
either the original or scaled copy, rather than multiplying each
component separately (suggested by Matt Turner).
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
I don't know why, but we never hooked up this pass Eric wrote.
Otherwise, you can end up with stupid scalarized code such as:
vec4 ssa_7 = load_const (0.0, 0.0, 0.0, 0.0)
vec4 ssa_8 = ...
vec1 ssa_9 = feq ssa_8, ssa_7
vec1 ssa_10 = feq ssa_8.y, ssa_7.y
vec1 ssa_11 = feq ssa_8, ssa_7.z
vec1 ssa_12 = feq ssa_8.y, ssa_7.w
ssa_8.xyxy == <0, 0, 0, 0> should only take two feq instructions.
shader-db on Skylake:
total instructions in shared programs: 9121153 -> 9120749 (-0.00%)
instructions in affected programs: 32421 -> 32017 (-1.25%)
helped: 277
HURT: 69
total cycles in shared programs: 69003364 -> 69000912 (-0.00%)
cycles in affected programs: 899186 -> 896734 (-0.27%)
helped: 313
HURT: 403
This also prevents regressions when disabling channel expressions.
v2: Don't call opt_cse afterwards (requested by Matt). It should
happen in the optimization loop below anyway.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
|
|
|
|
|
|
| |
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Acked-by: Matt Turner <mattst88@gmail.com>
Acked-by: Jose Fonseca <jfonseca@vmware.com>
|