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>
v2: use a ternary (Jason)
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
This really hacky commit adds a bit size to registers and SSA values. It
also adds rules in the validator to validate that they do the right things.
It's still an open question as to whether or not we want a bit_size in
nir_alu_instr or if we just want to let it inherit from the destination.
I'm inclined to just let it inherit from the destination. A similar
question needs to be asked about intrinsics.
v2 (Connor):
- Relax validation: comparisons have explicit destination sizes
and implicit source sizes.
v3 (Sam):
- Use helpers to get size and base types of nir_alu_type enum.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
As nir_alu_type has now embedded the data size, the check for the
instruction's output type (to see if a boolean resolve is required)
should ignore the data size part.
Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
v2: Fix size/type mask to properly handle 8-bit types.
v3: Add helpers to get the bitsize and base type of a
nir_alu_type enum.
Signed-off-by: Juan A. Suarez Romero <jasuarez@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>
This allows us to first generate atomic operations for shared
variables using these opcodes, and then later we can lower those to
the shared atomics intrinsics with nir_lower_io.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Previously we were receiving shared variable accesses via a lowered
intrinsic function from glsl. This change allows us to send in
variables instead. For example, when converting from SPIR-V.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
This code in brw_set_dest adjusts the execution size of any instruction
with a dst.width < 8. However, we don't want to do this with instructions
operating on doubles, since these will have a width of 4, but still
need an execution size of 8 (for SIMD8). Unfortunately, we can't just check
the size of the operands involved to detect if we are doing an operation on
doubles, because we can have instructions that do operations on double
operands interpreted as UD, operating on any of its 2 32-bit components.
Previous commits have made it so we never emit instructions with a horizontal
width of 4 that don't have the correct execution size set for gen6+, so
we can skip it in this case, avoiding the conflicts with fp64 requirements.
Expanding the same fix to other hardware generations requires many more
changes but since we are not targetting fp64 support on them
wer don't really care for now.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
v2 (Topi):
- No need to set the execsize for the indirect send message,
the next patch will handle that.
- Set the execution size explicitly instead of taking it from
the width of the dst that we set before.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Previously, we gave our internal clear/blit shaders actual GL handles
and stored them in the shader/program hash table. We used ordinary
GL API entrypoints to work with them.
We thought this shouldn't be a problem because GL doesn't allow
applications to invent their own names for shaders or programs.
GL allocates all names via glCreateShader and glCreateProgram.
However, having them in the hash table is a bit risky: if a broken
application guesses the name of our shaders or programs, it could
alter them, potentially screwing up future meta operations.
Also, test cases can observe the programs in the hash table. Running
a single dEQP process that executes the following test list:
dEQP-GLES3.functional.negative_api.buffer.clear
dEQP-GLES3.functional.negative_api.shader.compile_shader
dEQP-GLES3.functional.negative_api.shader.delete_shader
would result in the last two tests breaking. The compile_shader test
calls glCompileShader(9) straight away, and since it hasn't even created
any shaders or programs, it expects to get a GL_INVALID_VALUE error
because there's no such name. However, because the clear test ran
first, it created Meta programs, so an object named "9" did exist.
This patch reworks Meta to work with gl_shader and gl_shader_program
pointers directly. These internal programs have bogus names, and are
never stored in the hash tables, so they're invisible to applications.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94485
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
This will allow me to use them directly from Meta, bypassing the
versions that work with GL integer handles.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Timothy Arceri <timothy.arceri@collabora.com>
In half the callers, we already have a pointer, and don't need
to look it up again. This will also help with upcoming meta work.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Timothy Arceri <timothy.arceri@collabora.com>
In half the callers, we already have a pointer, and don't need
to look it up again. This will also help with upcoming meta work.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Timothy Arceri <timothy.arceri@collabora.com>
Now that the field exists in the instruction, we can make discards less
special. As a bonus, that means that we should be able to merge some more
.sf instructions together when we get around to that.
This causes some scheduling changes, as it allows tlb_color_reads to be
delayed past the discard condition setup. Since the tlb_color_read ends
up later, this may mean performance improvements, but I haven't tested.
total instructions in shared programs: 78114 -> 78035 (-0.10%)
instructions in affected programs: 1922 -> 1843 (-4.11%)
total estimated cycles in shared programs: 234318 -> 234329 (0.00%)
estimated cycles in affected programs: 8200 -> 8211 (0.13%)
The register allocator doesn't really do anything about the temp, so it
doesn't seem like it should matter. However, the scheduler would think
that a new def is being created.
This doesn't change anything yet, but it avoids a bunch of regressions in
the next commit.
This is cleaner than using glBindAttribLocation().
Not all drivers support the extension, but I don't think those drivers
use GLSL in the first place. Apparently some Meta shaders already use
GL_ARB_explicit_attrib_location, so I think it should be okay.
Honestly, I'm not sure how the old code worked anyway - we bound the
attribute location for "texcoords", while all the shaders capitalized
or spelled it differently.
v2: Convert another instance in brw_meta_fast_clear.c.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
When a user defines a point size array and enables it, the point
size value set via glPointSize should be ignored. To achieve this,
we can simply toggle ctx->VertexProgram.PointSizeEnabled.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=42187
Signed-off-by: Plamena Manolova <plamena.manolova@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This is done instead of copy propagating the VPM reads into the
instructions using them, because VPM reads have to stay in order.
shader-db results:
total instructions in shared programs: 78509 -> 78114 (-0.50%)
instructions in affected programs: 5203 -> 4808 (-7.59%)
total estimated cycles in shared programs: 234670 -> 234318 (-0.15%)
estimated cycles in affected programs: 5345 -> 4993 (-6.59%)
Signed-off-by: Varad Gautam <varadgautam@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Tested-by: Rhys Kidd <rhyskidd@gmail.com>
This file will contain optimization passes for both vpm reads
and writes.
Signed-off-by: Varad Gautam <varadgautam@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Some rasterization code relies (for sse) on the first and third planes
(but not the second for now) being 128bit aligned, and we didn't get that
on 32bit - I mistakenly thought the 64bit number in the struct would get
the thing aligned to 64bit even on 32bit archs.
Stephane Marchesin really figured this out.
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
CC: <mesa-stable@lists.freedesktop.org>
The logic was comparing actual ints, not true/false values.
This meant that it was emitting always multiple line segments instead of just
one even if the stipple test had the same result, which looks inefficient, and
the segments also overlapped thus breaking line aa as well.
(In practice, with the no-op default line stipple pattern, for a 10-pixel
long line from 0-9 it was emitting 10 segments, with the individual segments
ranging from 0-1, 0-2, 0-3 and so on.)
This fixes https://bugs.freedesktop.org/show_bug.cgi?id=94193
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
CC: <mesa-stable@lists.freedesktop.org>
All these img filter loops iterate through NUM_CHANNELS, not QUAD_SIZE.
In practice both are of course the same unchangeable value (4), but it
makes the code look a bit confusing. Moreover, some of the functions were
actually given an array of 4 values according to the declaration, yet the
code was addressing values 0/4/8/12 out of it, so fix this by just saying
it's a pointer to floats like the other functions.
While here, also add comment about not quite correct filtering.
There's no actual code difference.
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
The filt_args->offset wasn't assigned but was always used later leading
to a crash (as far as I can tell, texel offsets don't actually make much
sense with anisotropic filtering, but because there's no explicit setting
if offsets are enabled there the array is always accessed).
This fixes https://bugs.freedesktop.org/show_bug.cgi?id=94481
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
CC: <mesa-stable@lists.freedesktop.org>
This allows drivers to make smarter decisions e.g. about whether the image
has to be decompressed.
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
This is required to preserve the image variable's coherent/restrict/volatile
qualifiers in TGSI.
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Frontends should have this information readily available, and it simplifies
image LOAD/STORE/ATOM* handling especially with indirect image access.
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Iago Toral Quiroga
Connor Abbott
Jason Ekstrand
Samuel Iglesias Gonsálvez
Jordan Justen
Kenneth Graunke
Eric Anholt
Plamena Manolova
Varad Gautam
Roland Scheidegger
Nicolai Hähnle