radv: emit pa_sc_mode_cntl_0 with multisample state.

We don't have the meta kludge with 0 viewports anymore,
so we can always enable them.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>

i965: Avoid problems from referencing orphaned BOs after growing.

Growing the batch/state buffer is a lot more dangerous than I thought.

A number of places emit multiple state buffer sections, and then write
data to the returned pointer, or save a pointer to brw->batch.state.bo
and then use it in relocations.  If each call can grow, this can result
in stale map references or stale BO pointers.  Furthermore, fences refer
to the old batch BO, and that reference needs to continue working.

To avoid these woes, we avoid ever swapping the brw->batch.*.bo pointer,
instead exchanging the brw_bo structures in place.  That way, stale BO
references are fine - the GEM handle changes, but the brw_bo pointer
doesn't.  We also defer the memcpy until a quiescent point, so callers
can write to the returned pointer - which may be in either BO - and
we'll sort it out and combine the two properly in the end.

v2/v3:
- Handle stale pointers in the shadow copy case, where realloc may or
  may not move our shadow copy to a new address.
- Track the partial map explicitly, to avoid problems with buffer reuse
  where multiple map modes exist (caught by Chris Wilson).

v4:
- Don't use realloc in the CPU shadow case, it isn't safe.

Fixes: 2dfc119f22f257082ab0 "i965: Grow the batch/state buffers if we need space and can't flush."
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com> [v3]
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>

i965: Rename 'aux' to 'prog_data' in program cache.

'aux' is a very generic name, suggesting it can be a bunch of things.
However, it's always the brw_*_prog_data structure.  So, call it that.

Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>

swr: allow a single swr architecture to be builtin

Part 2 of 2 (part 1 is autoconf changes, part 2 is C++ changes)

When only a single SWR architecture is being used, this allows that
architecture to be builtin rather than as a separate libswrARCH.so that
gets loaded via dlopen.  Since there are now several different code
paths for each detected CPU architecture, the log output is also
adjusted to convey where the backend is getting loaded from.

This allows SWR to be used for static mesa builds which are still
important for large HPC environments where shared libraries can impose
unacceptable application startup times as hundreds of thousands of copies
of the libs are loaded from a shared parallel filesystem.

Based on an initial implementation by Tim Rowley.

v2: Refactor repetitive preprocessor checks to reduce code duplication
v3: Formatting changes per Bruce C. Also delay screen creation until end
    to avoid leaks when failure conditions are hit.

Signed-off-by: Chuck Atkins <chuck.atkins@kitware.com>
Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
CC: Tim Rowley <timothy.o.rowley@intel.com>
Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>

swr: (autoconf) allow a single swr architecture to be builtin

Part 1 of 2 (part 1 is autoconf changes, part 2 is C++ changes)

When only a single SWR architecture is being used, this allows that
architecture to be builtin rather than as a separate libswrARCH.so that
gets loaded via dlopen.  Since there are now several different code
paths for each detected CPU architecture, the log output is also
adjusted to convey where the backend is getting loaded from.

This allows SWR to be used for static mesa builds which are still
important for large HPC environments where shared libraries can impose
unacceptable application startup times as hundreds of thousands of copies
of the libs are loaded from a shared parallel filesystem.

Based on an initial implementation by Tim Rowley.

v2: Fix comment placement pointed out by Bruce C.

Signed-off-by: Chuck Atkins <chuck.atkins@kitware.com>
Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>
CC: Tim Rowley <timothy.o.rowley@intel.com>
Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>

swr: fix clang 5 null cast warning

Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Bruce Cherniak <bruce.cherniak@intel.com>

mesa/program: Fix -Wunused-param warning

v2: Don't annotate, but remove the unused ctx parameter

Signed-off-by: Gert Wollny <gw.fossdev@gmail.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>

mesa/program/prog_execute.c: Silence -Wunused-param

v2: Don't annotate, but remove the unused ctx parameter

Signed-off-by: Gert Wollny <gw.fossdev@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

mesa: Make numSamples an unsigned int

As a followup to the previous patch propagate the change of numSamples
from int to unsigned to gl_config::samples and consequently fix some
-Wsign-compare warnings.

Signed-off-by: Gert Wollny <gw.fossdev@gmail.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>

gallium: Make (num_)samples an unsigned int

According to the ARB_multisample num_samples is a non-negative integer.
Consequently define it as such, fail in glx/choose_visual if a negative
number is given.

v2: split patch into gallium and mesa part

Signed-off-by: Gert Wollny <gw.fossdev@gmail.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

docs: correct a typo in releasing instructions

Cc: Emil Velikov <emil.velikov@collabora.com>
Cc: Juan A. Suarez Romero <jasuarez@igalia.com>
Signed-off-by: Andres Gomez <agomez@igalia.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Reviewed-by: Juan A. Suarez <jasuarez@igalia.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

docs: move untar line in basic testing instructions for coherence

For scons, windows/mingw dealing with LLVM_CONFIG is done before
untarring. This is also more convenient for copy and paste.

Cc: Emil Velikov <emil.velikov@collabora.com>
Cc: Juan A. Suarez Romero <jasuarez@igalia.com>
Signed-off-by: Andres Gomez <agomez@igalia.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Reviewed-by: Juan A. Suarez <jasuarez@igalia.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

docs: add a notice whenever a release is the final in a series

Cc: Emil Velikov <emil.velikov@collabora.com>
Cc: Juan A. Suarez Romero <jasuarez@igalia.com>
Signed-off-by: Andres Gomez <agomez@igalia.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Reviewed-by: Juan A. Suarez <jasuarez@igalia.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

docs: add final release note for 17.2.8

Cc: Emil Velikov <emil.velikov@collabora.com>
Cc: Juan A. Suarez Romero <jasuarez@igalia.com>
Signed-off-by: Andres Gomez <agomez@igalia.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Reviewed-by: Juan A. Suarez <jasuarez@igalia.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

docs: add final release note for 17.1.10

Cc: Emil Velikov <emil.velikov@collabora.com>
Cc: Juan A. Suarez Romero <jasuarez@igalia.com>
Signed-off-by: Andres Gomez <agomez@igalia.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Reviewed-by: Juan A. Suarez <jasuarez@igalia.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

st/vdpau: release held lock in error path

Signed-off-by: Grazvydas Ignotas <notasas@gmail.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Cc: mesa-stable@lists.freedesktop.org

docs: update calendar, add news and link release notes to 17.3.3

Signed-off-by: Juan A. Suarez Romero <jasuarez@igalia.com>

docs: add sha256 checksums for 17.3.3

Signed-off-by: Juan A. Suarez Romero <jasuarez@igalia.com>
(cherry picked from commit bc1503b13fcf8190262757ea7f86613e14e25981)

docs: add release notes for 17.3.3

Signed-off-by: Juan A. Suarez Romero <jasuarez@igalia.com>
(cherry picked from commit 80f5f279b3f9fc752ba35b1cb2878a936f8ace90)

anv: avoid segmentation fault due to vk_error()

vk_error() is a macro that calls __vk_errorf() with instance == NULL.

Then, __vk_errorf() passes a pointer to instance->debug_report_callbacks
to vk_debug_error(), which segfaults as this pointer is invalid but not
NULL.

Fixes: e5b1bd6ab8 "vulkan: move anv VK_EXT_debug_report implementation to common code."
Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>

ac/nir: Fix vector extraction if source vector has >4 elements.

v2: Add forgotten argument and start offset.

Fixes: 91074bb11bda "radv/ac: Implement Float64 SSBO stores."
Tested-by: Timothy Arceri <tarceri@itsqueeze.com>
Acked-by: Timothy Arceri <tarceri@itsqueeze.com>

ac/nir: Use correct 32-bit component writemask for 64-bit SSBO stores.

Fixes: 91074bb11bda "radv/ac: Implement Float64 SSBO stores."
Tested-by: Timothy Arceri <tarceri@itsqueeze.com>
Acked-by: Timothy Arceri <tarceri@itsqueeze.com>

ac/nir: Fix TCS output LDS offsets.

When a channel was not set we also did not increase the LDS address,
while that obviously should happen.

The output loading code was inadvertently fixed which resulted in a
mismatch causing the SaschaWillems tessellation demo to result
in corrupt rendering.

Fixes: 7898eb9a60 "ac: rework load_tcs_{inputs,outputs}"
Reviewed-by: Dave Airlie <airlied@redhat.com>

radv: Use correct bindings for inputRate in key generation.

The bindings also have an index field.

Fixes: 49d035122e "radv: Add single pipeline cache key."
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104677
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>

radv: Implement VK_ANDROID_native_buffer.

Passes
  dEQP-VK.api.smoke.*
  dEQP-VK.wsi.android.*

with android-cts-7.1_r12 .

Unlike the initial anv implementation this does
use syncobjs instead of waiting on the CPU.

This is missing meson build coverage for now.

One possible todo is that linux 4.15 now has a
sycall that allows us to export amdgpu fence to
a sync_file, which allows us not to force all
fences and semaphores to use syncobjs. However,
I had trouble with my kernel crashing regularly
with NULL pointers, and I'm not sure how beneficial
it is in the first place given that intel uses
syncobjs for all fences if available.

Reviewed-by: Dave Airlie <airlied@redhat.com>

radv: Add create image flag to not use DCC/CMASK.

If we import an image, we might not have space in the
buffer for CMASK, even though it is compatible.

Reviewed-by: Dave Airlie <airlied@redhat.com>

radv: Generate VK_ANDROID_native_buffer.

Reviewed-by: Dave Airlie <airlied@redhat.com>

radv: Replace an assert with unreachable.

Otherwise we get uninitialized variable warnings for es_vgpr_comp_cnt.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>

radv: Remove DCC check on CS resolve dst image.

Gives a warning when the assert is disabled, and not even
necessarily true.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>

gallivm: support avx512 (16x32) in interleave2_half

lp_build_interleave2_half was not doing the right thing for avx512-style
16-wide loads.

This path is hit in the swr driver with a 16-wide vertex shader. It is
called from lp_build_transpose_aos, when doing texel fetches and the
fetched data needs to be transposed to one component per output register.

Special-case the post-load swizzle operations for avx512 16x32 (16-wide
32-bit values) so that we move the xyzw components correctly to the outputs.

Reviewed-by: Roland Scheidegger <sroland@vmware.com>

vbo: fix VBO optimization regression

The optimization in change 8e4efdc895ea ("vbo: optimize some display
list drawing") missed the loopback case.  This is used when the
glBegin/End primitive doesn't have a uniform set of vertex attributes.
The new Piglit gl-1.0-dlist-materials test hits this.

So check the aligned_vertex_buffer_offset(list) value and adjust the
buffer offset accordingly.

We also need to remove the 'start == 0' assertion in the loopback
code since it no longer applies.

Reviewed-by: Roland Scheidegger <sroland@vmware.com>

meson: ensure that xmlpool_options.h is generated for targets that need it

Currently a couple of gallium targets race with xmlpool_options.h being
generated, don't do that.

Signed-off-by: Dylan Baker <dylan.c.baker@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>

ac: fix visit_ssa_undef() for doubles

V2: use LLVMIntTypeInContext()

Fixes: f4e499ec7914 "radv: add initial non-conformant radv vulkan driver"
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>

ac/nir: account for view index in the user sgpr allocation.

The view index user sgpr wasn't being accounted for properly,
this refactors out the code to decide if it's required and then
uses that info to account for it.

Fixes: 180c1b924e (ac/nir: Add shader support for multiviews.)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600: enable ARB_enhanced_layouts

Only one piglit test fails,
sso-vs-gs-fs-array-interleave

There are 3 tests using ssbo without checking sizes failing also
but those are test bugs.

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

intel: Future-proof ring names for aubinator_error_decode

The kernel is moving to a $class$instance naming scheme in preparation
for accommodating more rings in the future in a consistent manner. It is
already using the naming scheme internally, and now we are looking at
updating some soft-ABI such as the error state to use the new naming
scheme. This of course means we need to teach aubinator_error_decode how
to map both sets of ring names onto its register maps.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Michel Thierry <michel.thierry@intel.com>

i965: Bind null render targets for shadow sampling + color.

Portal 2 appears to bind RGBA8888_UNORM textures to a sampler2DShadow,
and calls shadow2D() on it.  This causes undefined behavior in OpenGL.

Unfortunately, our sampler appears to hang in this scenario, which is
not acceptable.  Just give them a null surface instead, which returns
all zeroes.

Fixes GPU hangs in Portal 2 on Kabylake.

Huge thanks to Jason Ekstrand for noticing this crazy behavior while
sifting through crash dumps.

Cc: mesa-stable@lists.freedesktop.org
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104487
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>

anv/query: implement multiview interactions

From the Vulkan spec with KHX extensions:

  "If queries are used while executing a render pass instance that has
   multiview enabled, the query uses N consecutive query indices
   in the query pool (starting at query) where N is the number of bits
   set in the view mask in the subpass the query is used in.

   How the numerical results of the query are distributed among the
   queries is implementation-dependent. For example, some implementations
   may write each view's results to a distinct query, while other
   implementations may write the total result to the first query and write
   zero to the other queries. However, the sum of the results in all the
   queries must accurately reflect the total result of the query summed
   over all views. Applications can sum the results from all the queries to
   compute the total result."

In our case we only really emit a single query (in the first query index)
that stores the aggregated result for all views, but we still need to manage
availability for all the other query indices involved, even if we don't
actually use them.

This is relevant when clients call vkGetQueryPoolResults and pass all N
queries to retrieve the results. In that scenario, without this patch,
we will never see queries other than the first being available since we
never emit them.

v2: we need the same treatment for timestamp queries.

v3 (Jason):
 - Better an if instead of an early return.
 - We can't write to this memory in the CPU, we should use
   MI_STORE_DATA_IMM and emit_query_availability (Jason).

v4 (Jason):
 - No need to take the value to write as parameter, just hard code it to 0.

Fixes test failures in some work-in-progress CTS multiview+query tests.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>

vc5: add missing files to the tarball

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>

broadcom: add missing headers to the tarball

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>

i965/screen: Allow drirc to set 'allow_rgb10_configs' again.

Since setup of ALLOW_RGB10_CONFIGS was moved to i965's own
brw_config_options.xml, this was hard-coded to false and
could not be overriden by drirc. Add some parsing into
i965's private screen->optionCache to enable drirc again.

Fixes: b391fb26df9f1b ("dri_util: remove ALLOW_RGB10_CONFIGS option (v2)")
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Cc: Marek Olšák <marek.olsak@amd.com>
Cc: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>

anv: return VK_ERROR_OUT_OF_DEVICE_MEMORY when surface size is out of HW limits

Signed-off-by: Samuel Iglesias Gonsálvez <siglesias@igalia.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>

ac: tidy up array indexing logic

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>

mesa/st: translate SO info in glsl_to_nir() case

This was handled for VS, but not for GS.

Fixes for gallium drivers using nir:
spec@arb_gpu_shader5@arb_gpu_shader5-xfb-streams-without-invocations
spec@arb_gpu_shader5@arb_gpu_shader5-xfb-streams*
spec@arb_transform_feedback3@arb_transform_feedback3-ext_interleaved_two_bufs_gs*
spec@ext_transform_feedback@geometry-shaders-basic
spec@ext_transform_feedback@* use_gs
spec@glsl-1.50@execution@geometry@primitive-id*
spec@glsl-1.50@execution@geometry@tri-strip-ordering-with-prim-restart gl_triangle_strip *
spec@glsl-1.50@transform-feedback-builtins
spec@glsl-1.50@transform-feedback-type-and-size

v2: don't call st_translate_program_stream_output) for TCS

v3: drop scanning patch outputs as TCS can't output xfb

Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Tested-by: Karol Herbst <kherbst@redhat.com>

r600/sb: add lds related peepholes.

if no destination:
a) convert _RET instructions to non _RET variants if no dst
b) set src0 to undefined if it's a READ, this should get DCE then.

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: use different stacks for tracking lds and queue usage.

The normal ssa renumbering isn't sufficient for LDS queue access,
this uses two stacks, one for the lds queue, and one for the
lds r/w ordering.

The LDS oq values are incremented in their use in a linear
fashion.
The LDS rw values are incremented in their definitions and used
in the next lds operation to ensure reordering doesn't occur.

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: schedule LDS ops in appropriate places.

So LDS ops have to be SLOT_X,
and LDS OQ reads have read port restrictions so we try
and force those into only having one per slot and avoiding
bank swizzles.

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: hit the scheduler with a big hammer to avoid lds splits.

This tries to avoid an lds queue read getting scheduled separately
from an lds ret read, the non-sb code uses the same style of hammer,
this isn't foolproof.

We can do better, but it's a bit tricky, as you have to scan ahead
and either schedule more lds oq moves and more lds reads and that
could lead to you running out of space anyways.

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: adding lds oq tracking to the scheduler

This adds support for tracking the lds oq read/writes
so can avoid scheduling other things in between.

This patch just adds the tracking and assert to show
problems.

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: add gcm support to avoid clause between lds read/queue read

You have to schedule LDS_READ_RET _, x and MOV reg, LDS_OQ_A_POP
in the same basic block/clause. This makes sure once we've issues
and MOV we don't add another block until we balance it with an
LDS read.

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: handle lds special dest registers.

This adds lds to the geom emit handling

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: handle LDS operations in folding.

Don't try and fold LDS using expressions.

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: add finalising for lds output queue special values.

We need to convert these to the hw special registers.

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: add initial support for parsing lds operations.

This handles parsing the LDS ops and queue accessess.

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: disable if conversion for hs

This fixes bad interactions with the LDS special values.

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: lds ops have no dst register.

Although these are op3s they don't have a dst reg.

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: introduce special register values for lds support.

For LDS read/write ordering we use the LDS_RW value, reads
will wait on previous writes.
For LDS read/read from LDS queue ordering we use the LDS_OQ
values, we define two for now, though initially we'll just
support OQA.

Also add the check for the lds oq values

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: update last_cf if alu is the last clause

It's rare to have a final alu clause on normal shaders (exports)
but tess shaders write to LDS as their output, so we see some
alu clauses, and the CF_END get put in the wrong place.

This makes sure to update last_cf correctly.

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: start adding GDS support

This adds support for GDS ops to sb backend.

This seems to work for atomics and tess factor writes.

Acked-By: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: add tess/compute initial state registers.

This stops them being optimised out.

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/sb: fix a bug emitting ar load from a constant.

Some tess shaders were doing MOVA_INT _, c0.x on cayman, and then
hitting an assert in sb_bc_finalize.cpp:translate_kcache.

This makes sure the toplevel kcache tracker gets updated,
and the clause gets fixed up.

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Cc: <mesa-stable@lists.freedesktop.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600/shader: only emit add instruction if param has a value.

Just saves a pointless a = a + 0;

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600: emit 0 gds_op for tf write.

This field is ignored for tf writes so should be 0.

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

r600: add support for ARB_shader_clock.

Reviewed-by: Gert Wollny <gw.fossedev@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

radv/ws: get rid of useless return value

This also used boolean, so nice to kill that.

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>

radv: Initialize DCC on transition from preinitialized.

Looks like the decompress does not handle invalid encodings well,
which happens with random memory. Of course apps should not use it
with random memory, but they are allowed to ....

Fixes: 44fcf58744 "radv: Disable DCC for GENERAL layout and compute transfer dest."
Reviewed-by: Dave Airlie <airlied@redhat.com>

ac: fix buffer overflow bug in 64bit SSBO loads

Fixes: 441ee1e65b04 "radv/ac: Implement Float64 SSBO loads"
Reviewed-by: Marek Olšák <marek.olsak@amd.com>

ac: fix nir_intrinsic_get_buffer_size for radeonsi

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>

i965: Pass brw_growing_bo to grow_buffer().

Cleaner.

Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>

i965: Make a helper for recreating growing buffers.

Now that we have two of these, we're duplicating a bunch of this logic.
The next commit will add more logic, which would make the duplication
seem worse.

This ends up setting EXEC_OBJECT_CAPTURE on the batch, which isn't
necessary (it's already captured), but it should be harmless.

Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>

i965: Replace cpu_map pointers with a "use_shadow_copy" boolean.

Having a boolean for "we're using malloc'd shadow copies for all
buffers" is cleaner than having a cpu_map pointer for each.  It was
okay when we had one buffer, but this is more obvious.

Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>

intel/fs: Optimize and simplify the copy propagation dataflow logic.

Previously the dataflow propagation algorithm would calculate the ACP
live-in and -out sets in a two-pass fixed-point algorithm.  The first
pass would update the live-out sets of all basic blocks of the program
based on their live-in sets, while the second pass would update the
live-in sets based on the live-out sets.  This is incredibly
inefficient in the typical case where the CFG of the program is
approximately acyclic, because it can take up to 2*n passes for an ACP
entry introduced at the top of the program to reach the bottom (where
n is the number of basic blocks in the program), until which point the
algorithm won't be able to reach a fixed point.

The same effect can be achieved in a single pass by computing the
live-in and -out sets in lock-step, because that makes sure that
processing of any basic block will pick up the updated live-out sets
of the lexically preceding blocks.  This gives the dataflow
propagation algorithm effectively O(n) run-time instead of O(n^2) in
the acyclic case.

The time spent in dataflow propagation is reduced by 30x in the
GLES31.functional.ssbo.layout.random.all_shared_buffer.5 dEQP
test-case on my CHV system (the improvement is likely to be of the
same order of magnitude on other platforms).  This more than reverses
an apparent run-time regression in this test-case from my previous
copy-propagation undefined-value handling patch, which was ultimately
caused by the additional work introduced in that commit to account for
undefined values being multiplied by a huge quadratic factor.

According to Chad this test was failing on CHV due to a 30s time-out
imposed by the Android CTS (this was the case regardless of my
undefined-value handling patch, even though my patch substantially
exacerbated the issue).  On my CHV system this patch reduces the
overall run-time of the test by approximately 12x, getting us to
around 13s, well below the time-out.

v2: Initialize live-out set to the universal set to avoid rather
    pessimistic dataflow estimation in shaders with cycles (Addresses
    performance regression reported by Eero in GpuTest Piano).
    Performance numbers given above still apply.  No shader-db changes
    with respect to master.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104271
Reported-by: Chad Versace <chadversary@chromium.org>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>

gallium: remove PIPE_CAP_USER_CONSTANT_BUFFERS

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>

st/mesa: assume that user constant buffers are always supported

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>

nine: assume that user constant buffers are always supported

Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>

gallium: remove PIPE_CAP_TEXTURE_SHADOW_MAP

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>

st/mesa: expose ARB_sync unconditionally

All drivers support it.

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>

gallium: remove PIPE_CAP_TWO_SIDED_STENCIL

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>

glsl: remove unneeded extern "C" {} bracketing around Mesa includes

The two headers already have the right extern "C" annotations.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>

mesa: move gl_external_samplers() to program.[ch]

The function is only called from a couple places.  It doesn't make
sense to have it in mtypes.h

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>

st/mesa: include util/bitscan.h in st_glsl_to_tgsi_temprename.cpp

And use "" instead of <> for including Mesa headers, as we do elsewhere.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>

glsl: include util/bitscan.h in serialize.cpp

Instead of relying on indirect inclusion of the header.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>

util: include string.h in u_dynarray.h

To get memset() prototype.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>

mesa: remove unneeded #includes of main/compiler.h

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>

st/mesa: remove unneeded #includes of main/compiler.h

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>

st/mesa: include main/compiler.h in st_cb_queryobj.c

To get CPU_TO_LE32() macro.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>

mesa: include util/macros.h in format_fallback.c

To get definition of unreachable() macro.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>

mesa: include compiler.h in disk_cache.c

Instead of indirect inclusion to get CPU_TO_LE32() macro.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>

mesa/program: change validate_inputs() local var 'inputs' to GLbitfield64

Both state->prog->info.inputs_read and state->InputsBound are GLbitfield64
so it seems that the OR of those values should be of the same type.
I'm not sure this fixes any actual issues though.

Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>

vbo: reindent vbo_attrib.h to use 3 spaces

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

vbo: whitespace, formatting fixes in vbo_exec_api.c

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

vbo: add assertions, comments in vbo_exec_api.c

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

vbo: whitespace, formatting fixes in vbo_exec_draw.c

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

vbo: use inputs_read var to simplify code

v2: add some const qualifiers, per Ian.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

vbo: whitespace, formatting fixes in vbo_split_copy.c

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

vbo: use a new local 'array' variable in bind_vertex_list() loop

Make the code a bit more concise.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

vbo: remove unneeded #includes in vbo_context.c

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

vbo: whitespace, formatting fixes in vbo_context.c

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

vbo: change vbo_context attribute map arrays to GLubyte

The values will never be larger than VBO_ATTRIB_MAX (currently 44).

v2: add STATIC_ASSERT to be sure VBO_ATTRIB_MAX can fit in ubyte,
per Emil.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

vbo: lift common code out of switch cases

Both switch cases began with the same code.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>