Unfortunately, this also means that we need to use a slightly different
algorithm for assign_constant_locations. The old algorithm worked based on
the assumption that each read of a uniform value read exactly one float.
If it encountered a MOV_INDIRECT, it would immediately bail and push the
whole thing. Since we can now read ranges using MOV_INDIRECT, we need to
be able to push a series of floats without breaking them up. To do this,
we use an algorithm similar to the on in split_virtual_grfs.
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
This commit moves us to an instruction based model rather than a
register-based model for indirects. This is more accurate anyway as we
have to emit instructions to resolve the reladdr. It's also a lot simpler
because it gets rid of the recursive reladdr problem by design.
One side-effect of this is that we need a whole new algorithm in
move_uniform_array_access_to_pull_constants. This new algorithm is much
more straightforward than the old one and is fairly similar to what we're
already doing in the FS backend.
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Now that we have MOV_INDIRECT opcodes, we have all of the size information
we need directly in the opcode. With a little restructuring of the
algorithm used in assign_constant_locations we don't need param_size
anymore. The big thing to watch out for now, however, is that you can have
two ranges overlap where neither contains the other. In order to deal with
this, we make the first pass just flag what needs pulling and handle
assigning pull constant locations until later.
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Instead of using reladdr, this commit changes the FS backend to emit a
MOV_INDIRECT whenever we need an indirect uniform load. We also have to
rework some of the other bits of the backend to handle this new form of
uniform load. The obvious change is that demote_pull_constants now acts
more like a lowering pass when it hits a MOV_INDIRECT.
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
While we're at it, we also add support for the possibility that the
indirect is, in fact, a constant. This shouldn't happen in the common case
(if it does, that means NIR failed to constant-fold something), but it's
possible so we should handle it.
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
The subnr field is in bytes so we don't need to multiply by type_sz.
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
It should work fine without it and the visitor can set it if it wants.
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
This will fix the spurious error message: "Failed to query GPU properties."
that was unintentionally added in cc01b63d73.
This patch changes the function to return an int so that the caller is able to
do stuff based on the return value.
The equivalent of this patch was in the original series that fixed up the
warning, but I dropped it at the last moment. It is required to make the desired
behavior of not warning when trying to query GPU properties from the kernel
unless there is something the user can do about it.
v2: Use strerror (Jason)
Make EINVAL check similar in all places (Ian)
NOTE: Broadwell appears to actually have some issue where the kernel returns
ENODEV when it shouldn't be. I will investigate this separately.
Reported-by: Chris Forbes <chrisf@ijw.co.nz>
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
I neglected to free the sampler view which was created earlier in the
function. So for each glCallLists() command that used the bitmap atlas
to draw text, we'd leak a sampler view object.
Also, check for st_create_texture_sampler_view() failure and record
GL_OUT_OF_MEMORY.
Cc: "11.1 11.2" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Charmaine Lee <charmainel@vmware.com>
I removed that return 0 by mistake. Ooops.
Fixes: 6e23fd4 ("nvc0: allow to use compute support on GM200")
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
This works like a charm but please not that NVF0_COMPUTE have to be set
because compute support is still not enabled by default on GK110+. This
will require more testing to make sure it won't break the 3D state.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Emil Velikov:
- Attribute the src/{glsl,compiler}/nir move
- Flesh out to separate SConscript
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
Not supported by MSVC, and completely unnecessary -- inline functions
work just as well.
NIR_SRC_INIT/NIR_DEST_INIT could and probably should be replaced by the
inline functions.
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
It doesn't seem needed, and is not available on MSVC.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
Not supported by MSVC and consistent through NIR.
[Emil Velikov: rebase]
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
Rather than having two almost identical Makefiles, with various VPATH
hacks just fold them, using COMMON_* variables and actually getting
things buildable/shipable.
v2: whitespace fixes, remove Makefile.sources-arch
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Tim Rowley <timothy.o.rowley@intel.com>
This fixes GS piglit failures after adding SI_PARAM_SHADER_BUFFERS,
which bumped NUM_USER_SGPRS and uncovered this bug on SI.
If this was fixed in LLVM, these workarounds wouldn't be needed.
LLVM would have to look at the calling convention to know how many SGPR
inputs are declared, and add VCC and the scratch wave offset (which is
enabled even if we spill SGPRs but not VGPRs, oh well).
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Except:
- u_cache_test -- too long
- translate_test -- unreliable (it's probably testing corner cases that
translate module doesn't care about.)
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
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>
The old GLSL IR based lowering doesn't quite work right in all cases,
and fails several dEQP-GLES31 and Vulkan CTS tests. Jason's new
approach in NIR passes all the tests. There's not likely to be a ton
of advantage to lowering early in GLSL IR anyway, so...switch.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
The algorithm used is different from both the naive suggestion from the
GLSL spec and the one used in GLSL IR today. Unfortunately, the GLSL IR
implementation that we have today doesn't handle denormals (for those that
care) or the case where the float source is +-inf.
Reviewed-by: Matt Turner <mattst88@gmail.com>
It's not really doing enough anymore to justify a helper function.
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Reveiewed-by: Kristian Høgsberg <krh@bitplanet.net>
There are several passes where we need to specify some set of variable
modes that the pass needs top operate on. This lets us easily do that.
Acked-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
- Incorporate flatshade flag into the shader generation
- Use provoking vertex (vc) in shader when flat shading.
Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
This reverts commit 62fa868728.
dEQP-GLES3.functional.occlusion_query.* was unhappy about that change.
Still not really sure *what* the other slots in the sample results
buffer are.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
This one is slightly annoying, since trying to write RBRC from draw
would clobber values set in the tiling/gmem code. We could do command-
stream patching for RBRC, as is done on a3xx. Although since it seems
to be a rarely used feature, it is easier just to do RMW to set/clear
the bit.
Fixes dEQP-GLES3.functional.rasterizer_discard.basic.write_depth_triangles
and related tests.
a3xx still needs the same feature, although there it probably makes more
sense to take advantage of the existing cmdstream patching which is
required for RBRC for other reasons.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Seems like a4xx needs offset added to array index for all arrays,
whereas a3xx only for cubemap arrays. Fixes a whole swath of dEQP fails
(roughly *sampler2darray*).
Signed-off-by: Rob Clark <robclark@freedesktop.org>
We need to increment offset by # of vertices, not by # of prims. Fixes
a bunch of dEQP fails involving prims other than points. For example,
dEQP-GLES3.functional.transform_feedback.position.lines_separate
Signed-off-by: Rob Clark <robclark@freedesktop.org>
