Lionel Landwerlin
f0e13827aa
intel/dump_gpu: set default device_override
...
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com >
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/2201 >
2020-08-07 11:27:54 +00:00
Lionel Landwerlin
c4bfa6dfbb
intel/dump_gpu: further track mapping of BOs
...
We can go further in tracking what BOs are written to by the driver by
tracking when a buffer in unmapped. A BO could be mmap, written, unmap
and never be written to again. In such case we can just write the BO's
content on the first exec buf after unmap and never write it again.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com >
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/2201 >
2020-08-07 11:27:54 +00:00
Lionel Landwerlin
63c90acc9e
intel/dump_gpu: only write BOs mapped by the driver
...
Track what BOs are mapped by the driver and only write those. We can
safely assume that when not mapped there is no data to save.
v2: Don't forget to return the ret (Rafael)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com >
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/2201 >
2020-08-07 11:27:54 +00:00
Mike Blumenkrantz
54c68498d4
zink: set primitive restart cap
...
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5912 >
2020-08-07 11:17:04 +00:00
Mike Blumenkrantz
5959178381
zink: use util_draw_vbo_without_prim_restart for unsupported prim modes
...
this fixes up primitive restart functionality for the primitive types that
vulkan doesn't support using primitive restart with
fixes #2873
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5912 >
2020-08-07 11:17:04 +00:00
Mike Blumenkrantz
f4583b4086
zink: move 8bit index handling out of u_primconvert path
...
putting this through util_translate_prim_restart_ib reduces our reliance on
u_primconvert, which ideally we don't want to be using
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5912 >
2020-08-07 11:17:04 +00:00
Mike Blumenkrantz
5edaf081cd
zink: basic primitive restart support for strip/fan topologies
...
this conditionally handles rewriting the index buffer to use vk-compatible
restart indexes and then enables it in the pipeline for supported draw
modes
fixes #3174
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5912 >
2020-08-07 11:17:04 +00:00
Bas Nieuwenhuizen
8531830038
radv: Add ETC2 support on RAVEN2.
...
I did a whole bunch of validation and it essentially works.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6215 >
2020-08-07 10:54:58 +00:00
Samuel Pitoiset
bea8930468
radv: allow to force-enable LLVM internally for a specific shader stage
...
For ACO debugging purposes, developers only.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6208 >
2020-08-07 07:45:06 +00:00
Christian Gmeiner
e63a7882a0
etnaviv: call nir_lower_bool_to_bitsize
...
Starting with commit 6f394343b1 ("nir/algebraic: i2f(f2i()) -> trunc()")
dEQP-GLES2.functional.shaders.operator.binary_operator.div.lowp_int_vertex
ends with an unsuppored flt instruction. Use nir_lower_bool_to_bitsize to
convert this flt to a flt32 which is supported. This fixes the introduced
regression.
Cc: 20.2 <mesa-stable@lists.freedesktop.org >
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6182 >
2020-08-07 03:25:36 +00:00
Alyssa Rosenzweig
c5b2319628
panfrost: Drop compiler cmdstream deps
...
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
1d0b3ef051
panfrost: Treat texture dimension as first-class
...
Instead of hiding behind a cmdstream enum. The raw values have a very
simple interpretation as dimension.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
5a3088e85f
panfrost: Inline max rt into compilers
...
Another dep.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
54f96544ec
pan/bi: Drop use of MALI_POSITIVE
...
Contributing to a dep on the main cmdstream. It's a fancy word for
subtract one...
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
c94013c819
panfrost: Add panloader/ to .gitignore
...
For out-of-tree trace infrastructure.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
52e3a67bb1
panfrost: Reduce bit dependency to disassembly only
...
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
14bb72c68b
pan/bi: Separate disasm/compiler targets
...
Likewise.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
b792d613e6
pan/mdg: Separate disassembler and compiler targets
...
The compiler depends on the disassembler, which is okay. But the
disassembler now no longer depends on the compiler circularly, avoiding
the need to being in GLSL/NIR.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
d62a6e7c5f
panfrost: Move pandecode into lib/
...
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
1c62b5528a
panfrost: Rename encoder/ to lib/
...
We'll want both encoding and decoding here, as a generic hardware
interface library based on GenXML.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
956b09653c
panfrost: Don't export exception_status
...
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
4ddc6e543c
panfrost: Remove panfrost-misc.h
...
Now unused.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
64f3c9da87
panfrost: Inline panfrost-misc.h into panfrost-job.h
...
We only need a small subset of the defines here.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
ebf7e87f2c
pan/decode: Remove shader replacement artefact
...
Did this ever work?
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
e4df95bf9f
panfrost: Move format stringify to decode.c
...
Let's simplify pandecode before moving it around. We don't need to share
this routine right now, and we'll be generating it soon.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Alyssa Rosenzweig
b80e753940
panfrost: Remove blend prettyprinters
...
Unused and mostly wrong too.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6196 >
2020-08-06 23:54:24 +00:00
Lionel Landwerlin
b43bcb05a1
anv: fix incorrect realloc failure handling
...
We don't want to leak in case of failure.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com >
Reported-by: Ivan Briano <ivan.briano@intel.com >
Fixes: 246261f0ad ("anv: prepare the driver for delayed submissions")
Fixes: 34f32a6d66 ("anv: implement VK_KHR_timeline_semaphore")
Reviewed-by: Ivan Briano <ivan.briano@intel.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6200 >
2020-08-06 23:07:56 +00:00
Rhys Perry
21b47cbd99
aco: fix C++11/C++14 compilation
...
static_assert without a message is only available since C++17.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com >
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl >
Fixes: d1f992f3c2 ('aco: rework barriers and replace can_reorder')
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/3374
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6216 >
2020-08-06 23:51:14 +01:00
Danylo Piliaiev
782ba8d3ae
st/mesa: Treat vertex outputs absent in outputMapping as zero in mesa_to_tgsi
...
After updating vertex outputs being written based on optimized NIR, they may
go out of sync with outputs in mesa IR. Which is translated to TGSI and used
together with NIR if draw doesn't have llvm.
It's much easier to treat such outputs as zero because there is no pass to
entirely get rid of them.
Similar to eeab9c93db but now for outputs.
Fixes: d684fb37bf
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/3365
Signed-off-by: Danylo Piliaiev <danylo.piliaiev@globallogic.com >
Reviewed-by: Marek Olšák <marek.olsak@amd.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6187 >
2020-08-06 16:53:50 +00:00
Dylan Baker
a92cfa66b4
Bump development version and clear new_features
...
Signed-off-by: Dylan Baker <dylan.c.baker@intel.com >
2020-08-06 09:34:42 -07:00
Jason Ekstrand
7f06d194fd
anv: Advertise shaderIntegerFunctions2
...
We advertised the extension string but never the feature bit. Doh!
Fixes: c57338b924 "anv: Enable SPV_INTEL_shader_integer_functions2..."
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com >
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6033 >
2020-08-06 16:14:16 +00:00
Icenowy Zheng
9e397956b0
panfrost: signal syncobj if nothing is going to be flushed
...
When nothing is going to be flushed, the kernel will get no job that
signals the syncobj.
Signal it by ourselves, otherwise it will never get signaled.
Closes : #3371
Signed-off-by: Icenowy Zheng <icenowy@aosc.io >
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com >
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6190 >
2020-08-06 13:16:25 +00:00
Bas Nieuwenhuizen
c6aadbae71
radv: Don't use both DCC and CMASK for single sample images.
...
Fixes: c67ef7695a "radv: Use ac_surface to allocate aux surfaces."
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6194 >
2020-08-06 10:50:22 +00:00
Jose Fonseca
e2c614a415
appveyor: Use Python3.
...
This implied upgrading to the Visual Studio 2019 image, not for VS
itself, but for the newer Python 3.8.5 version it contains, to avoid
UnicodeDecodeError inside modulefinder module when attempting to decode
our UTF-8 encoded Python scripts with cp1252 encoding.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6184 >
2020-08-06 09:46:48 +00:00
Jose Fonseca
0f9fb7ffaa
appveyor: Upgrade pip.
...
To avoid all those warnings.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6184 >
2020-08-06 09:46:48 +00:00
Vinson Lee
2e665458a9
util: Fix SCons build.
...
Fixes: 848e7b947d ("util: Move stack debug functions to src/util")
Signed-off-by: Vinson Lee <vlee@freedesktop.org >
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6199 >
2020-08-05 22:55:06 -07:00
Eric Anholt
56ab105182
freedreno: Add more asserts for DST_OFF/NUM_UNIT in indirect const uploads.
...
These are just empirical alignment numbers from looking at dEQP traces of
the blob driver (a330, a418, a540, a618, a630), with one exception noted
in the comments.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5990 >
2020-08-05 23:06:55 +00:00
Eric Anholt
3e970de360
freedreno: Increase the NUM_UNIT on compute's consts in indirect dispatch.
...
Avoids tripping the assert in the next commit -- the blob never uses
num_unit % 4 != 0 for indirect const uploads.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5990 >
2020-08-05 23:06:55 +00:00
Eric Anholt
f07e25bc6d
freedreno/ir3: Clean up instrlen setup.
...
We were calculating it with the gpu_id check in two places, do it once and
use ir3_compiler for the gpu_id dependency.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5990 >
2020-08-05 23:06:55 +00:00
Eric Anholt
8f637d66cc
freedreno: Split ir3_const's user buffer and indirect upload APIs.
...
They're almost entirely split by whether you're uploading user buffer or
from a BO. While I'm rewriting the API, drop the emit_const ->
fdN_emit_const wrapper in favor of a #define before the header and a
little helper for the asserts.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5990 >
2020-08-05 23:06:55 +00:00
Eric Anholt
154299934d
freedreno: Rename emit_const_bo() to emit_const_ptrs().
...
I keep thinking it's the "upload from inside a BO" path when it's not.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5990 >
2020-08-05 23:06:55 +00:00
Eric Anholt
51acfe2230
freedreno/ir3: Simpify the immediates from an array of vec4 to array of dwords.
...
We usually had to split the idx/swiz out of the dword index anyway. Note
that incidentally, immediates_size now increments in vec4s instad of
4*vec4s.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5990 >
2020-08-05 23:06:55 +00:00
Eric Anholt
e873c4da08
freedreno/ir3: Merge the redundant immediate_idx/immediates_count fields
...
I got tripped up again with the index vs count vs size fields and I'd
rather we didn't store the redundant info. Settle on immediates_count as
"how many dwords of immediates we have"
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5990 >
2020-08-05 23:06:55 +00:00
Rob Clark
5e922fbc16
glsl_to_nir: fix bitfield_extract with 16-bit operands
...
These are defined to explicitly take 32b values.
Reviewed-by: Marek Olšák <marek.olsak@amd.com >
Reviewed-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6073 >
2020-08-05 22:04:47 +00:00
Marek Olšák
92f5e94a93
glsl: improve precision determination for calls
...
Don't leave the precision as NONE for non-lowerable calls. Set it to HIGH
if a function really returns highp.
Reviewed-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6073 >
2020-08-05 22:04:47 +00:00
Marek Olšák
282a1e6288
glsl: don't lower to mediump for desktop OpenGL
...
Desktop OpenGL ignores all precision qualifiers.
Also, the lowering pass doesn't work if precision qualifiers are not set,
which is only possible with desktop OpenGL, causing random behavior.
Reviewed-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6073 >
2020-08-05 22:04:47 +00:00
Marek Olšák
01e0085637
glsl: don't create conversion opcodes for array types
...
Instead, convert all elements one by one.
This fixes piglit shaders@glsl-bug-110796.
Reviewed-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6073 >
2020-08-05 22:04:47 +00:00
Marek Olšák
5020403c70
glsl: don't lower atomic functions to mediump
...
Reviewed-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6073 >
2020-08-05 22:04:47 +00:00
Rob Clark
93076f60d3
glsl: don't inline intrinsics for mediump
...
They have an empty fxn body, trying to handle them results in the
intrinsic call being expanded into a no-op.
Reviewed-by: Marek Olšák <marek.olsak@amd.com >
Reviewed-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6073 >
2020-08-05 22:04:47 +00:00
Marek Olšák
48a6255186
glsl: fix constant expression evaluation for 16-bit types
...
Reviewed-by: Rob Clark <robdclark@chromium.org >
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6073 >
2020-08-05 22:04:47 +00:00