nir/validate: Allow array derefs of vectors for nir_var_mem_global

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Karol Herbst <kherbst@redhat.com>

nir/lower_io: Add support for nir_var_mem_global

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Karol Herbst <kherbst@redhat.com>

nir/lower_io: Add a 32 and 64-bit global address formats

These are simple scalar addresses.

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

nir: Add load/store/atomic global intrinsics

These correspond roughly to reading/writing OpenCL global pointers.  The
idea is that they just take a bare address and load/store from it.  Of
course, exactly what this address means is driver-dependent.

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Karol Herbst <kherbst@redhat.com>

st/nine: Enable debug info if NDEBUG is not set

We want to have debug info as well if using
meson's debugoptimized when ndebug is off.

v2: use u_debug functions that do something
even if DEBUG is not set.

Signed-off-by: Axel Davy <davyaxel0@gmail.com>

st/nine: Immediately upload user provided textures

Fixes regression caused by
42d672fa6a766363e5703f119607f7c7975918aa
st/nine: Bind src not dst in nine_context_box_upload

Before that patch, for user provided textures,
when the texture was destroyed, the safety
check for pending uploads, which according to
the code "Following condition cannot happen currently",
was flushing the queue and thus triggering the upload.

After the patch, the texture destruction was delayed after
the upload. However the user frees the texture buffer,
as it thinks the texture released.

Instead of reverting the faulty patch,
this patch instead flushes the csmt queue right away
after queuing the upload for this type of textures.
This is more future-proof, as we may want to bind the
surface for other reasons in the future.

Signed-off-by: Axel Davy <davyaxel0@gmail.com>
Cc: 18.3 <mesa-stable@lists.freedesktop.org>

i965: Always compile fp64 funcs when needed

Compilation of user-specified shaders with software fp64 works by
compiling on demand an "fp64-funcs" shader implementing various fp64
operations and then linking it into the "user shader".

In

   commit 64b8c86d37ebb1e1d286c69d642d52b7bcf051d3
   Author: Timothy Arceri <tarceri@itsqueeze.com>
   Date:   Thu Jan 17 17:16:29 2019 +1100

       glsl: be much more aggressive when skipping shader compilation

we changed the behavior of the shader cache to skip compilation earlier
when we get a cache hit.

After the aforementioned commit, compiling a user program using fp64
would store into the cache an entry for the fp64-funcs shader.
Subsequent compilations of uncached user shaders using fp64 would fail
in compile_fp64_funcs() after finding a cache entry for the fp64-funcs,
but being unprepared to read from the cache.

It's unclear to me how to retrieve the cached NIR of the fp64-funcs (if
it even is cached), so just call _mesa_glsl_compile_shader() with
force_recompile=true in order to ensure we generate the fp64-funcs
successfully.

Tested-by: Mark Janes <mark.a.janes@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>

intel/compiler: Add a file-level description of brw_eu_validate.c

Acked-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Reviewed-by: Francisco Jerez <currojerez@riseup.net>

freedreno: add renderonly scanout

This allows creating a fd_screen with a renderonly object which will be
used to allocated scanout resources.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Eric Anholt <eric@anholt.net>
[slight tweak to fix uninitialized 'prsc' in debug print]
Signed-off-by: Rob Clark <robdclark@gmail.com>

freedreno/a2xx: fix unused variable warning

Signed-off-by: Rob Clark <robdclark@gmail.com>

tgsi: remove culldist semantic from docs

The semantic was removed in e6d93893662d.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>

ac/nir_to_llvm: fix clamp shadow reference for more hardware

Fixes the following piglit test on my VEGA and matches the behaviour in the
tgsi backend.

tests/spec/glsl-1.10/execution/samplers/glsl-fs-shadow2D-clamp-z.shader_test

Fixes: 625dcbbc4566 ("amd/common: pass address components individually to ac_build_image_intrinsic")
Reviewed-by: Marek Olšák <marek.olsak@amd.com>

gallium: Make sure we return is_unorm/is_snorm for compressed formats.

The util helpers were looking for a non-void channels in a non-mixed
format and returning its snorm/unorm state.  However, compressed formats
don't have non-void channels, so they always returned false.  V3D wants to
use util_format_is_[su]norm for its border color clamping workarounds, so
fix the functions to return the right answer for these.

This now means that we ignore .is_mixed.  I could retain the is_mixed
check, but it doesn't seem like a useful feature -- the only code I could
find that might care is freedreno's blit, which has some notes about how
things are wonky in this area anyway.

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

gallium: Fix comment about possible colorspaces.

Two typos, and missing one of the colorspaces.

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

gallium: Enable unit tests as actual meson unit tests.

These tests don't need swrast, so we can always enable them when
build_tests is set.  Most of them run to successful completion quickly
(.9s on my SKL).

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

mapi: print function declarations for shared glapi

Earlier commit aimed to remove unneeded function declarations. Namely
OpenGL entrypoints which are not applicable for OpenGLES*

Although it did not consider the shared glapi which needs all,
including hidden ones. Resulting in warning/errors like the following

../build/src/mapi/shared-glapi/glapi_mapi_tmp.h:26014:15:
error: no previous prototype for ‘shared_dispatch_stub_1414’ [-Werror=missing-prototypes]

This patch addressed that.

Cc: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reported-by: Eric Anholt <eric@anholt.net>
Fixes: 6148cce388f ("mapi: drop unneeded gl_dispatch_stub declarations")
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Acked-by: Eric Anholt <eric@anholt.net>

freedreno: limit tiling to PIPE_BIND_SAMPLER_VIEW

1ce5d757d04 dropped this limit.. which is probably the right thing to
do.  But it results in an extra tiled->linear blit for glReadPixels()
(ie. dEQP/piglit) which is hitting some intermittent corruption (looks
like cache) on a6xx, causing a lot of spurious fails.

Since we are getting close to 19.0 branchpoint, re-instate this limit
for now, until the blitter problems are resolved.

Fixes: 1ce5d757d04 freedreno: core buffer modifier support
Signed-off-by: Rob Clark <robdclark@gmail.com>

radv: fix computing number of user SGPRs for streamout buffers

Streamout buffers are emitted like push constants.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>

appveyor: Revert commits adding Cygwin support.

This reverts commits 00ad77b9f683e561b1ac45fbb89eb2bafe45c8c6 and
5334dafee265d78abdfcf30e2c693e0791bfecf5.

This avoids Appveyor build breakage due to Cygwin, but more importantly,
there are several problems with these patches, as highlighted to my
recent mesa-dev mail.  So better to revert for now, and pursue Cygwin
support after these have been address.

android: fix build issues with libmesa_anv_gen* libraries

We need this include path to find nir/nir_xfb_info.h.

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>

intel/batch-decoder: fix a vb end address calculation

According to the loop implementation (in 'ctx_print_buffer' function),
which advances dword by dword over vertex buffer(vb),
the vb size should be aligned by 4 bytes too.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=109449
Signed-off-by: Andrii Simiklit <andrii.simiklit@globallogic.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>

intel/batch-decoder: fix vertex buffer size calculation for gen<8

It should be incremented by one according to
how it is calculated by 'emit_vertex_buffer_state':
  "\#if GEN_GEN < 8
      .BufferAccessType = step_rate ? INSTANCEDATA : VERTEXDATA,
      .InstanceDataStepRate = step_rate,
   \#if GEN_GEN >= 5
      .EndAddress = ro_bo(bo, end_offset - 1),
   \#endif
   \#endif"

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=109449
Signed-off-by: Andrii Simiklit <andrii.simiklit@globallogic.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>

meson/vdpau: add missing soversion

This mirrors what autotools does in src/gallium/state_trackers/vdpau/Makefile.am
and src/gallium/targets/vdpau/Makefile.am:

  VDPAU_MAJOR = 1
  VDPAU_MINOR = 0
  libvdpau_gallium_la_LDFLAGS = -version-number $(VDPAU_MAJOR):$(VDPAU_MINOR)

Reported-by: Igor Gnatenko <i.gnatenko.brain@gmail.com>
Fixes: 68076b87474e7959c161 "meson: build gallium vdpau state tracker"
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>

anv: drop always-successful VkResult

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>

anv/allocator: Avoid race condition in anv_block_pool_map.

Accessing bo->map and then pool->center_bo_offset without a lock is
racy. One way of avoiding such race condition is to store the bo->map +
center_bo_offset into pool->map at the time the block pool is growing,
which happens within a lock.

v2: Only set pool->map if not using softpin (Jason).
v3: Move things around and only update center_bo_offset if not using
softpin too (Jason).

Cc: Jason Ekstrand <jason@jlekstrand.net>
Reported-by: Ian Romanick <idr@freedesktop.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=109442
Fixes: fc3f58832015cbb177179e7f3420d3611479b4a9
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>

meson: Add warnings and errors when using ICC

ICC tries to be helpful by not erroring when it sees something that it
doesn't understand, which is completely the opposite of helpful. Meson
0.49.0 does much better at handling this by really trying to make ICC
error, but there are some things in mesa that still get ignored until
0.49.1

v2: - Fix id check, which is 'intel' not 'icc'

Cc: 18.3 <mesa-stable@lists.freedesktop.org>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com> (v1)

meson: Fix compiler checks for SWR with ICC

This is a bit fragile, as the way this "fixes" the check is to move the
one that we know is correct before the one that is incorrectly reported
as working. In meson 0.49.1 (which isn't out yet) this is fixed that the
incorrect check is reported as a failure.

Fixes: e0b037d6979b266d4959c1e31746d4d19c941fdb
       ("meson: Build SWR driver")
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=109129
Acked-by: Eric Engestrom <eric.engestrom@intel.com>

meson: fix swr KNL build

There's a typo in one of the #defines that breaks compilation.

Fixes: e0b037d6979b266d4959c1e31746d4d19c941fdb
       ("meson: Build SWR driver")
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=109023
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>

gallivm: Return true from arch_rounding_available() if NEON is available

LLVM uses the single instruction "FRINTI" to implement llvm.nearbyint.
Fixes the rounding tests of lp_test_arit.

Bug: https://bugs.gentoo.org/665570
Reviewed-by: Roland Scheidegger <sroland@vmware.com>

gallium: Enable ASIMD/NEON on aarch64.

NEON (now called ASIMD) is available on all aarch64 CPUs. Our code was
missing an aarch64 path, leading to util_cpu_caps.has_neon always being
false on aarch64.

Reviewed-by: Eric Anholt <eric@anholt.net>

gallium: use put image shm2 path (v2)

This fixes the drisw paths to use the new shm2 interface, so that
we don't trigger the X server overflow checks when the x offset is non-zero.

This just hides the versioning in drisw, and either passes the src_x
or adds the offset fixup for the fallback path.

Cc: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Adam Jackson <ajax@redhat.com>

glx: add support for putimageshm2 path (v2)

v2: pass x,0 in as the offset coords at glx level not earlier

Cc: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Adam Jackson <ajax@redhat.com>

dri_interface: add put shm image2 (v2)

This adds a new interface to the swrast interface to fix an shm put image bug.

The current code adds the x,y src offsets into the offset parameters,
however if the x offset is > 0, and the put image copies up to the height
of the image, this can trigger an X server validation check to fail and
the renderering to get BadMatch.

This patch fixes it to pass the x offset coord in as a src x.

We cannot pass the Y coordinate due to the horrible code mangling the
image w/h vs stride in swrastXPutImage.

v2: drop srcx,y from api

Cc: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Adam Jackson <ajax@redhat.com>

mapi: remove machinery handling CSV files

We haven't have one in years, so just drop the code.

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

mapi: remove old, unused ES* generator code

As of earlier commit, everyone has switched to the new script for the ES
dispatch.

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

mapi/es2api: remove no longer present entrypoints

With the previous scripts API from the following was incorrectly
exported. Drop them from the list, since they're no longer around.

GL_EXT_blend_func_extended
GL_EXT_texture_integer

Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

mapi/es*api: remove GL_EXT_multi_draw_arrays entrypoints

Now we use the upstream XML file and a cleaner generator. Thus the
symbols are no longer exported and we can drop them from this list.

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

mapi/es*api: remove GL_OES_EGL_image entrypoints

As some point in the past we fixed the scripts so, these are no longer
exported. Drop them from the list.

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

Revert "mapi/new: sort by slot number"

This reverts commit a1f5d9412cf7cacb3534635f6c2409fafbe6574e.

We no longer needed to sort - it was meant only to ease compare against
the old generated files.

scons: wire the new generator for es1 and es2

Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

meson: wire the new generator for es1 and es2

v2: use ${foo})_py naming (Dylan)
v3: use symbolic name for genCommon.py

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com> (v2)

autotools: wire the new generator for es1 and es2

The output produced functionally identical, with the following changes:
 - A cosmetic: swapped ABI compatible types [ GLclampf -> GLfloat, etc ]
 - B cosmetic: renamed parameters [ zNear -> n, etc ]
 - C dropped extension entrypoints - invalid/incorrect

To make things easier to validate, normalise both old/new headers run
the sed patterns A, B and C to both sets.

A
      s/\<GLclampf\>/GLfloat/g; s/\<GLclampx\>/GLfixed/g;
      s/\<GLvoid\>/void/g;

B
      s/\ \* / */g; s/\<texture\>/target/g;
      s/\<plane\>/p/g; s/\<depth\>/d/g; s/\<modeAlpha\>/modeA/g;
      s/\<shader\>/program/g; s/\<obj\>/shaders/g; s/\<equation\>/eqn/g;
      s/\<param\>/data/g; s/\<params\>/data/g; s/\<buffers\>/buffer/g;
      s/\<src\>/mode/g; s/\<count\>/n/g; s/\<zNear\>/n/g; s/\<zFar\>/f/g;
      s/\<zfail\>/dpfail/g; s/\<zpass\>/dppass/g; s/\<buf\>/index/g;
      s/\<value\>/target/g; s/\<cap\>/target/g; s/\<maskNumber\>/index/g;
      s/\<srcRGB\>/sfactorRGB/g; s/\<dstRGB\>/dfactorRGB/g;
      s/\<srcAlpha\>/sfactorAlpha/g; s/\<dstAlpha\>/dfactorAlpha/g;
      s/\<primitiveMode\>/mode/g; s/\<primcount\>/instancecount/g;
      s/\<top\>/t/g; s/\<bottom\>/b/g; s/\<left\>/l/g; s/\<right\>/r/g;
      s/\<x\>/v0/g; s/\<y\>/v1/g; s/\<z\>/v2/g; s/\<w\>/v3/g;
      s/\<sfactor\>/mode/g; s/\<dfactor\>/dst/g; s/\<attribindex\>/bindingindex/g;
      s/\<internalFormat\>/internalformat/g; s/\<bufSize\>/bufsize/g;

C
glMultiDrawArraysEXT
glMultiDrawElementsEXT

glBindFragDataLocationEXT

glGetTexParameterIivEXT
glGetTexParameterIuivEXT
glTexParameterIivEXT
glTexParameterIuivEXT

v2:
 - gl_dispatch_stub declarations are addressed with previous patch
 - the public_entries table is no longer generated

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

mapi/new: remove duplicate GLvoid/void substitution

We already do it a few lines above - drop the duplicate.

Note that for consistency sake, we keep the substitution since the GL
API is a mixed bad - some use GLvoid while others a normal void.

We might want to merge this back in GLVND.

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

mapi/new: fixup the GLDEBUGPROCKHR typedef to the non KHR one

This way we can reuse the latter, which is already present in the
headers that we use. Thus we can drop the manual typedef we generate.

We might want to merge this back in GLVND.

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

mapi/new: don't print info we don't need for ES1/ES2

There is no need for the noop functions, the public_stubs and
public_entries table or table size defines. Remove those.

Pretty much all of this is applicable to GLVND, although it
requires preparatory work.

v2:
 - python style fixes (Dylan)
 - use "gldispatch" instead of not "glesv1" "glesv2"
 - remove the public_entries table/array (Erik)

v3:
 - use if == "gldispatch", instead of "in" (Kyle)

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com> (v2)

mapi/new: split out public_entries handling

The only instance that requires the public_entries table is the
dispatch library - split that into another function.

We have to be careful with when undefining the guard, so split it out.

We might want to merge this back in GLVND.
Minor GLVND cleanup will be needed first.

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

mapi/new: reinstate _NO_HIDDEN suffixes in the new generator

Strictly speaking we can rework the rest of the code so we do not need
those. That said, this will require a series on it's own so let's carry
this local quirk for now.

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

mapi/new: use the static_data offsets in the new generator

Otherwise the incorrect ones will be used, effectively breaking the ABI.

Note: some entries in static_data.py list a suffixed API, while (for ES*
at least) we expect the one w/o suffix.

v2:
 - rework path handling (Dylan)
 - use else if chain (Erik)

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

mapi/new: sort by slot number

Makes it easier to compare the newly generated header against the old
one. Will be reverted after the transition.

mapi/new: import mapi scripts from glvnd

Currently we have over 20 scripts that generate the libGL* dispatch and
various other functionality. More importantly we're using local XML
files instead of the Khronos provides one(s). Resulting in an
increasing complexity of writing, maintaining and bugfixing.

One fairly annoying bug is handling of statically exported symbols.
Today, if we enable a GL extension for GLES1/2, we add a special tag to
the xml. Thus the ES dispatch gets generated, but also since we have no
separate notion of GL/ES1/ES2 static functions it also gets exported
statically.

This commit adds step one towards clearing and simplifying our setup.
It imports the mapi generator from GLVND.

  012fe39 ("Remove a couple of duplicate typedefs.")

v2: use local genCommon.py

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

mapi: move genCommon.py to src/mapi/new

The helper will also be used by the new Khronos gl.xml aware generator.

v2: Move existing one, instead of duplicating it.
v3: Correct genCommon.py references in meson [Erik]
v4: Drop the file from the EGL EXTRA_DIST [Erik]

Suggested-by: Kyle Brenneman <kbrenneman@nvidia.com>
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>

genCommon.py: Fix typo in _LIBRARY_FEATURE_NAMES.

Port glvnd commit 37fc6caa4b8 ("Fix typo in _LIBRARY_FEATURE_NAMES.")
from Michal Srb.

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

mapi: add all _glapi_table entrypoints to static_data.py

Currently various parts of mesa use the glapi_table differently.

Some use _glapi_get_proc_offset() to get the offset, while others
directly reference the specific offset via _gloffset_Function.

Add all static entries, to ensure things don't break as we flip to the
upstream XML + new mapi generator.

Note: the offsets are also used for the alias remap table, thus we need
to ensure we honour the correct offsets range or it will break.

Currently this is done via MAX_OFFSETS constant, although a better
solution is in the works.

v2: add FramebufferTexture2DMultisampleEXT
v3: add MAX_OFFSETS guard

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com> (v1)
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>

mapi: sort static entrypoints numerically

A few of the entrypoints were incorrectly placed. Sort those to align
with the rest of the list.

Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

Revert "mesa/main: remove ARB suffix from glGetnTexImage"

This reverts commit f1998e15ffccf260552bf559abe1a733a8ce990e.

This changes the ABI, such that glGetnTexImageARB entry-point from the
GLAPI gets removed. Thus accessing many functions by offset (as we do)
will result in getting the wrong one.

Follow-up work will swap the by-offset handling, but for now revert
this patch.

Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

mapi: drop unneeded gl_dispatch_stub declarations

These declarations are not used anywhere - be that generated code or
otherwise.

[Emil: format the hunk from Erik into a patch]
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>

mesa: correctly use os.path.join in our python scripts

With Windows in mind, using forward slash isn't the right thing to do.
Even if it just works, we might want to fix it.

As here, use __file__ instead of argv[0] and sys.path.insert over
sys.path.append. With the path tweak being reportedly faster.

Suggested-by: Dylan Baker <dylan@pnwbakers.com>
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>

freedreno: automake: ship ir3_nir_trig.py in the tarball

Fixes: aa0fed10d35 ("freedreno: move ir3 to common location")
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>

egl/glvnd: sync egl.xml from Khronos

Fixes: 98984b7cdd79c15cc733 "egl: add glvnd entrypoints for EGL_MESA_query_driver"
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Emil Velikov <emil.velikov@collabora.com>

travis: bump libdrm to 2.4.97

Fixes: c02f761bdfc87d0a1bfd "winsys/amdgpu: use the new BO list API"
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>

egl: Implementation of egl dri2 drivers for MESA_query_driver

Signed-off-by: Veluri Mithun <velurimithun38@gmail.com>
Acked-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

egl: add glvnd entrypoints for EGL_MESA_query_driver

Fixes: fbdd7bde29863935106c "egl: Implement EGL API for MESA_query_driver"
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

egl: Implement EGL API for MESA_query_driver

Signed-off-by: Veluri Mithun <velurimithun38@gmail.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

egl: update headers from Khronos

Cheating a tiny bit as these headers aren't in the Khronos repo yet, but
I expect them to be within a couple days.

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

egl: finalize EGL_MESA_query_driver

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

intel/compiler: Reset default flag register in brw_find_live_channel()

emit_uniformize() emits SHADER_OPCODE_FIND_LIVE_CHANNEL with its
flag_subreg set, so that the IR knows which flag is accessed. However
the flag is only used on Gen7 in Align1 mode.

To avoid setting unnecessary bits in the instruction words, get the
information we need and reset the default flag register. This allows
round-tripping through the assembler/disassembler.

Reviewed-by: Francisco Jerez <currojerez@riseup.net>

gallium: Add forgotten docs for PIPE_CAP_GLSL_TESS_LEVELS_AS_INPUTS.

Thanks to Ilia for catching this.

Revert "Implement EGL API for MESA_query_driver"

This reverts commit ff621a505594a3217f24fb785c51f43580f1bf78.

with default warnings configuration, this commit generates:

   ../src/egl/main/eglapi.c:2654:1: error: no previous prototype for
            ‘eglGetDisplayDriverConfig’ [-Werror=missing-prototypes]

Revert "Implementation of egl dri2 drivers for MESA_query_driver"

This reverts commit 2720f78ef231a5a19ea69731755ee4242acbc580.

Implementation of egl dri2 drivers for MESA_query_driver

Signed-off-by: Veluri Mithun <velurimithun38@gmail.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

Implement EGL API for MESA_query_driver

Signed-off-by: Veluri Mithun <velurimithun38@gmail.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Acked-by: Eric Engestrom <eric.engestrom@intel.com>

Add extension doc for MESA_query_driver

Signed-off-by: Veluri Mithun <velurimithun38@gmail.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Acked-by: Eric Engestrom <eric.engestrom@intel.com>

nir: Length of boolean vtn_value now is 1

During conversion type-length was lost due to math.

v2 (Jason Ekstrand):
 - Use a size/offset of 4 bytes

Fixes: 44227453ec03 (nir: Switch to using 1-bit Booleans for almost everything)
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=109353
Signed-off-by: Sergii Romantsov <sergii.romantsov@globallogic.com>
Tested-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>

st/mesa: fix PRIMITIVES_GENERATED query after the "pipeline stat single" changes

When this functionality was added, the PRIMITIVES_GENERATED query was
accidentally omitted. This causes issues for drivers that support
transform feedback."

Fixes: d644698b443 ("gallium: Add the ability to query a single
pipeline statistics counter")

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>

st/mesa: purge framebuffers when unbinding a context

This fixes pipe_surface "leaks".

Cc: 18.3 <mesa-stable@lists.freedesktop.org>
Acked-by: Timothy Arceri <tarceri@itsqueeze.com>

docs: add note about sending merge-requests from forks

Sending MRs from the main Mesa repository increase clutter in the
repository, and decrease visibility of project-wide branches. So it's
better if MRs are sent from forks instead.

Let's add a note about this, in case its not obvious to everyone.

Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>

freedreno: set modifier when exporting buffer

Fixes an assert we start hitting with kms/gbm:

  #0  0x0000007fbf3d6e3c in raise () from /lib64/libc.so.6
  #1  0x0000007fbf3c4a68 in abort () from /lib64/libc.so.6
  #2  0x0000007fbf3d04e8 in __assert_fail_base () from /lib64/libc.so.6
  #3  0x0000007fbf3d0550 in __assert_fail () from /lib64/libc.so.6
  #4  0x0000007fbf5a73c4 in gbm_dri_bo_create (gbm=0x5820f0, width=2160, height=1440, format=875713112, usage=0, modifiers=0x695e00, count=1) at ../src/gbm/backends/dri/gbm_dri.c:1150
  #5  0x0000007fbf5a49c4 in gbm_bo_create_with_modifiers (gbm=0x5820f0, width=2160, height=1440, format=875713112, modifiers=0x695e00, count=1) at ../src/gbm/main/gbm.c:491
  #6  0x0000007fbbac3d64 in get_back_bo (dri2_surf=0x6f4cc0) at ../src/egl/drivers/dri2/platform_drm.c:258
  #7  0x0000007fbbac4318 in dri2_drm_image_get_buffers (driDrawable=0x704490, format=4098, stamp=0x6fc730, loaderPrivate=0x6f4cc0, buffer_mask=1, buffers=0x7fffffe210) at ../src/egl/drivers/dri2/platform_drm.c:409
  #8  0x0000007fbf5a5318 in image_get_buffers (driDrawable=0x704490, format=4098, stamp=0x6fc730, loaderPrivate=0x70e150, buffer_mask=1, buffers=0x7fffffe210) at ../src/gbm/backends/dri/gbm_dri.c:135
  #9  0x0000007fbe4308c4 in dri_image_drawable_get_buffers (drawable=0x6fc730, images=0x7fffffe210, statts=0x6f2660, statts_count=1) at ../src/gallium/state_trackers/dri/dri2.c:339
  #10 0x0000007fbe430c44 in dri2_allocate_textures (ctx=0x614b30, drawable=0x6fc730, statts=0x6f2660, statts_count=1) at ../src/gallium/state_trackers/dri/dri2.c:466
  #11 0x0000007fbe435580 in dri_st_framebuffer_validate (stctx=0x714160, stfbi=0x6fc730, statts=0x6f2660, count=1, out=0x7fffffe3b8) at ../src/gallium/state_trackers/dri/dri_drawable.c:85
  #12 0x0000007fbe7b2c84 in st_framebuffer_validate (stfb=0x6f2190, st=0x714160) at ../src/mesa/state_tracker/st_manager.c:222
  #13 0x0000007fbe7b4884 in st_api_make_current (stapi=0x7fbf0430d8 <st_gl_api>, stctxi=0x714160, stdrawi=0x6fc730, streadi=0x6fc730) at ../src/mesa/state_tracker/st_manager.c:1074
  #14 0x0000007fbe434f44 in dri_make_current (cPriv=0x703c20, driDrawPriv=0x704490, driReadPriv=0x704490) at ../src/gallium/state_trackers/dri/dri_context.c:301
  #15 0x0000007fbe42c910 in driBindContext (pcp=0x703c20, pdp=0x704490, prp=0x704490) at ../src/mesa/drivers/dri/common/dri_util.c:579
  #16 0x0000007fbbabab40 in dri2_make_current (drv=0x69d170, disp=0x69c6e0, dsurf=0x6f4cc0, rsurf=0x6f4cc0, ctx=0x70cb40) at ../src/egl/drivers/dri2/egl_dri2.c:1456
  #17 0x0000007fbbaa8ef4 in eglMakeCurrent (dpy=0x69c6e0, draw=0x6f4cc0, read=0x6f4cc0, ctx=0x70cb40) at ../src/egl/main/eglapi.c:862
  #18 0x0000007fbf5736ac in InternalMakeCurrentVendor (dpy=dpy@entry=0x614fb0, draw=draw@entry=0x6f4cc0, read=read@entry=0x6f4cc0, context=context@entry=0x70cb40, apiState=apiState@entry=0x6fc940, vendor=0x6975f0) at libegl.c:861
  #19 0x0000007fbf573764 in InternalMakeCurrentDispatch (dpy=0x614fb0, draw=0x6f4cc0, read=0x6f4cc0, context=0x70cb40, vendor=0x6975f0) at libegl.c:630
  #20 0x0000000000403640 in init_egl (egl=0x5805a8 <gl>, gbm=0x580528 <gbm>, samples=0) at ../common.c:263
  #21 0x0000000000403c1c in init_cube_smooth (gbm=0x580528 <gbm>, samples=0) at ../cube-smooth.c:225
  #22 0x0000000000408618 in main (argc=1, argv=0x7fffffe8d8) at ../kmscube.c:145

Fixes: 1ce5d757d04 freedreno: core buffer modifier support
Signed-off-by: Rob Clark <robdclark@gmail.com>

radv: always pass the GFX9 fence data to si_cs_emit_cache_flush()

Remove two useless checks.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>

radv: compute the GFX9 fence VA at allocation time

Instead of doing every time we emit cache flushes.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>

radv: only allocate the GFX9 fence and EOP BOs for the gfx queue

It's invalid to emit a ZPASS_DONE event on the compute queue, and
the fence BO is unused on the compute queue (ie. we don't flush
CB or DB caches).
This saves some space in the upload BO.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>

radv: remove old_fence parameter from si_cs_emit_write_event_eop()

This parameter is actually useless as the immediate value
can always be zero.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>

radv: improve gathering of load_push_constants with dynamic bindings

For example, if a pipeline has two stages VS and FS. And if only
the fragment stage needs dynamic bindings, we shouldn't allocate
an extra user SGPR for the vertex stage. Of course, if the vertex
stage loads constants, it needs an user SGPR.

This should reduce the number of SET_SH_REG packets that are emitted.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>

gallium: Add PIPE_CAP_GLSL_TESS_LEVELS_AS_INPUTS

In the Intel backend, it makes the most sense to treat gl_TessLevelInner
and gl_TessLevelOuter as ordinary shader inputs.  For Radeon, it makes
more sense to treat them as system values which get special handling.

We already have a compiler option for this, but the Iris driver will
need a capability bit so we can set it appropriately.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

nv50,nvc0: mark textures dirty on fb update

We may have to flush the cache if there are any textures presently bound
that refer to the outgoing framebuffer. This is only checked at
validation time.

Fixes a number of dEQP-GLES3.functional.fbo.color.repeated_clear.sample.*
tests, which would bind a texture, then clear it while the binding was
in effect, and then render to a different texture. This seems legal
under the "no feedback loops" rule.

Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Karol Herbst <kherbst@redhat.com>

ac/nir_to_llvm: fix interpolateAt* for structs

This fixes the arb_gpu_shader5 interpolateAt* tests that contain
structs.

Acked-by: Marek Olšák <marek.olsak@amd.com>

ac/nir_to_llvm: add bindless support for uniform handles

Reviewed-by: Marek Olšák <marek.olsak@amd.com>

radeonsi/nir: add missing piece for bindless image support

This fixes some piglit tests and is was TGSI does.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>

freedreno: core buffer modifier support

Split out of a patch from Fritz Koenig to decouple from a6xx UBWC
enablement, and added fd_resource_create_with_modifiers().

loader: fix the no-modifiers case

Normally modifiers take precendence over use flags, as they are more
explicit.  But if the driver supports modifiers, but the xserver does
not, then we should fallback to the old mechanism of allocating a buffer
using 'use' flags.

Fixes: 069fdd5f9facbd72fb6a289696c7b74e3237e70f
Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@chromium.org>

freedreno: add query for dmabuf modifiers

freedreno: drm_fourcc.h header include

Add Qualcomm modifier for UBWC

svga: add new gallium formats to the format conversion table

Fixes a static assertion which broke the build.

Fixes: 3ee240890 "gallium: add SINT formats to have exact counterparts to SNORM formats"
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Neha Bhende<bhenden@vmware.com>

radeonsi: rename rfence -> sfence

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

radeonsi: rename rbo, rbuffer to buf or buffer

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

radeonsi: rename rsrc -> ssrc, rdst -> sdst

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

radeonsi: rename rquery -> squery

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

radeonsi: rename r600_resource -> si_resource

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

vulkan: make generated enum to strings helpers available from c++

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>

radeonsi: remove r600 from comments

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

winsys/amdgpu: rename rfence, rsrc, rdst -> afence, asrc, adst

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