Jordan Justen [Mon, 2 Oct 2017 07:08:55 +0000 (00:08 -0700)]
intel/nir: Zero local index const struct for valgrind & nir_serialize
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Jordan Justen [Mon, 2 Oct 2017 07:05:28 +0000 (00:05 -0700)]
nir: Zero local_size const struct for valgrind & nir_serialize
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Jordan Justen [Sat, 21 Oct 2017 01:54:17 +0000 (18:54 -0700)]
glsl: Add field initializers for glsl_struct_field default constructor
This helps valgrind when encode_type_to_blob is used.
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Jason Ekstrand [Thu, 14 Sep 2017 23:49:14 +0000 (16:49 -0700)]
compiler/types: Support [de]serializing void types
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Jason Ekstrand [Fri, 15 Sep 2017 23:22:00 +0000 (16:22 -0700)]
nir/intrinsics: Set the correct num_indices for load_output
Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Connor Abbott [Wed, 13 Sep 2017 00:16:22 +0000 (20:16 -0400)]
glsl: move shader_cache type handling to glsl_types
Not sure if this is the best place to put it, but we're going to need
this for NIR too.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Alex Smith [Wed, 25 Oct 2017 11:18:54 +0000 (12:18 +0100)]
vulkan: Update headers and registry to 1.0.64
Acked-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Alex Smith <asmith@feralinteractive.com>
Matthew Nicholls [Wed, 25 Oct 2017 13:20:43 +0000 (14:20 +0100)]
ac/nir: generate correct instruction for atomic min/max on unsigned images
v2: fix silly typo
Cc: "17.2 17.3" <mesa-stable@lists.freedesktop.org>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Roland Scheidegger [Wed, 25 Oct 2017 00:39:20 +0000 (02:39 +0200)]
gallium/util: remove some block alignment assertions
These assertions were revisited a couple of times in the past, and they
still weren't quite right.
The problem I was seeing (with some other state tracker) was a copy between
two 512x512 s3tc textures, but from mip level 0 to mip level 8. Therefore,
the destination has only size 2x2 (not a full block), so the box width/height
was only 2, causing the assertion to trigger for src alignment.
As far as I can tell, such a copy is completely legal, and because a correct
assertion would get ridiculously complicated just get rid of it for good.
Reviewed-by: Brian Paul <brianp@vmware.com>
Eric Engestrom [Tue, 24 Oct 2017 13:57:11 +0000 (14:57 +0100)]
meson: be explicit about the version required
This way, we know what we're allowed to use (no nested include lists
for instance) and users get immediate feedback when trying to use
unsupported versions, rather than a cryptic crash or things being
silently not built correctly.
Cc: Dylan Baker <dylan@pnwbakers.com>
Signed-off-by: Eric Engestrom <eric.engestrom@imgtec.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Erik Faye-Lund [Mon, 23 Oct 2017 18:54:03 +0000 (20:54 +0200)]
meson: add opt-out of libunwind
Libunwind has some issues on some platforms, so let's allow people
who have issues to opt-out. This is similar to what we do in automake,
and the implementation is modelled after our opt-out for valgrind.
Signed-off-by: Erik Faye-Lund <kusmabite@gmail.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Harish Krupo [Wed, 4 Oct 2017 11:56:28 +0000 (14:56 +0300)]
gles2: support for GL_EXT_occlusion_query_boolean
Following test checking entrypoints passes:
dEQP-EGL.functional.get_proc_address.extension.gl_ext_occlusion_query_boolean
Piglit test 'ext_occlusion_query_boolean-any-samples' passes with these changes.
No changes/regression observed in WebGL occlusion tests or Intel CI.
v2: add es2="2.0" for glapi entrypoints, clean up xml
dispatch_sanity changes (fix 'make check')
Signed-off-by: Harish Krupo <harish.krupo.kps@intel.com>
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Tapani Pälli [Wed, 4 Oct 2017 11:56:27 +0000 (14:56 +0300)]
mesa: enum checks for GL_EXT_occlusion_query_boolean
Some of the checks are valid for generic ES 3.2 as well.
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Samuel Pitoiset [Tue, 24 Oct 2017 15:23:43 +0000 (17:23 +0200)]
radv: print NIR before LLVM IR and disassembly
It's still printed after linking, but it makes more sense to
have SPIRV->NIR->LLVM IR->ASM.
Fixes:
f0a2bbd1a4 (radv: move nir print after linking is done)
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Bas Nieuwenhuizen [Wed, 25 Oct 2017 01:43:00 +0000 (03:43 +0200)]
radv: Fix truncation issue hexifying the cache uuid for the disk cache.
Going from binary to hex has a 2x blowup.
Fixes:
14216252923 'radv: create on-disk shader cache'
Reviewed-by: Dave Airlie <airlied@redhat.com>
Timothy Arceri [Wed, 18 Oct 2017 02:46:44 +0000 (13:46 +1100)]
radv: enable lower to scalar nir pass
This will allow dead components of varyings to be removed.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Timothy Arceri [Thu, 5 Oct 2017 00:44:39 +0000 (11:44 +1100)]
ac: add support for explicit component packing
This is needed for RADV to support explicit component packing.
This is also required to use the new NIR component splitting /
packing passes.
V2:
- add commponent packing support for interpolate_at* intrinsics
- improve store packing support when not all varyings are scalar
as spotted by Bas the store source was incorrectly offset.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Timothy Arceri [Mon, 23 Oct 2017 01:26:12 +0000 (12:26 +1100)]
i965: fix unused var warnings in release build
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Dave Airlie [Wed, 25 Oct 2017 01:23:08 +0000 (02:23 +0100)]
radv: use device name in cache creation like radeonsi.
Not sure how useful this is, but it makes it more consistent.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Cc: "17.3" <mesa-stable@lists.freedesktop.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Tue, 24 Oct 2017 22:05:11 +0000 (08:05 +1000)]
radv: use a define for the transition point between cp and compute shader
For certain buffer meta ops we can use the CP or a compute shader,
we should use a define to rather than hardcoding 4096, allows
for easier testing and more consistency.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Kenneth Graunke [Tue, 24 Oct 2017 17:34:10 +0000 (10:34 -0700)]
docs: Mark GL_KHR_no_error as done.
Drivers have supported KHR_no_error for a while. We'd been leaving it
marked as "in progress" because there's a zillion places that could get
slightly more optimized. But, Timothy and Samuel have already done
piles of work, and I think we have a solid implementation at this point.
Let's check it off the list.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Kenneth Graunke [Fri, 20 Oct 2017 22:53:50 +0000 (15:53 -0700)]
i965: Call gen6_upload_push_constants() even when the stage is disabled.
This properly sets stage_state->push_constant_dirty = true, so that we
emit 3DSTATE_CONSTANT_XS to disable the constant buffer for the shader
stage. It also sets stage_state->push_const_size = 0.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Kenneth Graunke [Mon, 23 Oct 2017 02:28:35 +0000 (19:28 -0700)]
i965: Drop a bunch of downcasting and upcasting of gl_program pointers.
We have a gl_program and we want a gl_program. There's no point in
converting to brw_program and back again. This probably made more
sense in the old days before Tim dropped a layer of subclassing.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Kenneth Graunke [Fri, 20 Oct 2017 22:50:28 +0000 (15:50 -0700)]
i965: Move _mesa_shader_write_subroutine_indices down a level.
Now we call it in one place instead of making every caller do it.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Dave Airlie [Tue, 24 Oct 2017 03:03:29 +0000 (04:03 +0100)]
radv: only emit dfsm packets if dfsm is allowed.
radeonsi only emits these when dfsm is enabled, so for now
just hinge them on a flag we never set.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Rob Clark [Sat, 14 Oct 2017 14:08:50 +0000 (10:08 -0400)]
meson: build freedreno
Mostly copy/pasta from Dylan Baker's conversion of nouveau and i965.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Rob Clark [Tue, 24 Oct 2017 15:07:52 +0000 (11:07 -0400)]
meson: extract out variable for nir_algebraic.py
Also needed in freedreno/ir3.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Rob Clark [Sat, 14 Oct 2017 17:00:28 +0000 (13:00 -0400)]
freedreno/ir3: use a flag instead of setting PYTHONPATH
Similar to
848da662224326ccfbe6647bc82f4f89ca22c762, pass an arg to
ir3_nir_trig.py to add to python path, rather than using $PYTHONPATH,
to prep for meson build support.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Kenneth Graunke [Mon, 23 Oct 2017 22:46:47 +0000 (15:46 -0700)]
i965: Don't disable CCS for RT dependencies when dispatching compute.
Compute shaders don't have access to the framebuffer, so there's no
point in worrying whether a texture is bound as a render target.
This saves a bunch of resolves in GFXBench4 Manhattan 3.1, but doesn't
seem to impact performance at all, at least on Apollolake.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Eric Anholt [Tue, 17 Oct 2017 22:41:25 +0000 (15:41 -0700)]
i965: Fix memmem compiler warnings.
gcc is throwing this warning in my meson build:
../src/intel/compiler/brw_eu_validate.c:50:11: warning
argument 1 null where non-null expected [-Wnonnull]
return memmem(haystack.str, haystack.len,
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
needle.str, needle.len) != NULL;
~~~~~~~~~~~~~~~~~~~~~~~
The first check for CONTAINS has a NULL error_msg.str and 0 len. The
glibc implementation will exit without looking at any haystack bytes if
haystack.len < needle.len, so this was safe, but silence the warning
anyway by guarding against implementation variablility.
Fixes:
122ef3799d56 ("i965: Only insert error message if not already present")
Reviewed-by: Matt Turner <mattst88@gmail.com>
Rob Clark [Thu, 24 Aug 2017 13:34:48 +0000 (09:34 -0400)]
freedreno: per-context fd_pipe
To enable per-context priorities, we need to have per-context pipe's.
Unfortunately we still need to keep the global screen pipe, mostly just
for screen->get_timestamp().
Signed-off-by: Rob Clark <robdclark@gmail.com>
Rob Clark [Thu, 24 Aug 2017 13:30:25 +0000 (09:30 -0400)]
freedreno: rename pipe -> vsc_pipe
To add context priority support we need to have an fd_pipe per context,
rather than per-screen. Which conflicts with existing ctx->pipe (which
is actually a visibility stream pipe (hw resource). So just rename it.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Rob Clark [Thu, 24 Aug 2017 13:16:53 +0000 (09:16 -0400)]
freedreno: pass context flags through to fd_context_init()
Prep work for later patch.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Brian Paul [Mon, 23 Oct 2017 21:25:24 +0000 (15:25 -0600)]
gallium/util: use util_snprintf() in u_socket_connect()
Instead of plain snprintf(). To fix the MSVC build.
snprintf() is used in various places in Mesa/gallium, but apparently,
not in code built with MSVC.
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
Benjamin Gordon [Fri, 20 Oct 2017 21:34:57 +0000 (15:34 -0600)]
configure: Allow android as an EGL platform
I'm working on radeonsi support in the Chrome OS Android container
(ARC++). Mesa in ARC++ uses autotools instead of Android.mk, but all
the necessary EGL bits are there, so the existing check is too strict.
Signed-off-by: Benjamin Gordon <bmgordon@chromium.org>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
Marek Olšák [Wed, 4 Oct 2017 03:07:50 +0000 (05:07 +0200)]
radeonsi: postponed KILL isn't postponed anymore, but maintains WQM
This restores performance for the drirc workaround, i.e.
KILL_IF does:
visible = src0 >= 0;
kill_flag &= visible; // accumulate kills
amdgcn_kill(wqm_vote(visible)); // kill fully dead quads only
And all helper pixels are killed at the end of the shader:
amdgcn_kill(kill_flag);
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Wed, 4 Oct 2017 13:37:30 +0000 (15:37 +0200)]
radeonsi: use postponed KILL only when derivatives are used
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Wed, 4 Oct 2017 03:06:57 +0000 (05:06 +0200)]
ac: use llvm.amdgcn.kill with LLVM 6.0
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Wed, 4 Oct 2017 02:51:39 +0000 (04:51 +0200)]
ac: replace ac_build_kill with ac_build_kill_if_false
This will be a new LLVM intrinsic and will also work nicely with
llvm.amdgcn.wqm.vote.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Timothy Arceri [Mon, 23 Oct 2017 22:08:55 +0000 (09:08 +1100)]
radv: move nir print after linking is done
We now have linking optimisations so we want to delay dumping the
nir until after these are complete.
Fixes:
06f05040eb73 (radv: Link shaders)
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Dave Airlie [Mon, 23 Oct 2017 00:34:54 +0000 (10:34 +1000)]
mesa/bufferobj: don't double negate the range
This fixes a regression I introduced refactoring this code,
I managed to invert range twice, I moved the inversion into
the common code, but forgot to stop doing it in the callee.
Fixes: GL45-CTS.multi_bind.dispatch_bind_buffers_base
Fixes:
35ac13ed3 (mesa/bufferobj: consolidate some codepaths between ubo/ssbo/atomics.)
Reported-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Timothy Arceri [Mon, 23 Oct 2017 00:10:47 +0000 (11:10 +1100)]
radv: clone meta shaders before linking
The IR is reused in different pipeline combinations so we need
to clone it to avoid link time optimistaions messing up the
original copy.
Fixes:
06f05040eb73 (radv: Link shaders)
Reviewed-by: Dave Airlie <airlied@redhat.com>
Brian Paul [Tue, 17 Oct 2017 19:59:28 +0000 (13:59 -0600)]
gallium/util: don't call close() on Windows in u_tests.c
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Brian Paul [Tue, 17 Oct 2017 19:59:03 +0000 (13:59 -0600)]
mesa: use util_strdup() macro in u_debug_symbol.c
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Brian Paul [Tue, 17 Oct 2017 19:58:47 +0000 (13:58 -0600)]
mesa: use util_strdup() macro in symbol_table.c
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Brian Paul [Tue, 17 Oct 2017 19:57:45 +0000 (13:57 -0600)]
util: add util_strdup() wrapper macro
To work around MSVC warning that strdup() is a deprecated POSIX function.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Brian Paul [Tue, 17 Oct 2017 19:11:03 +0000 (13:11 -0600)]
gallium/util: replace gethostbyname() with getaddrinfo()
Compiling with MSVC options /we4995 /we4996 (a subset of /sdl) generates
a warning that the gethostbyname() function is deprecated in favor of
getaddrinfo() or GetAddrInfoW(). Replace the call with getaddrinfo().
Untested. There are no callers to u_socket_connect() in Gallium.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Alex Smith [Mon, 23 Oct 2017 09:37:42 +0000 (10:37 +0100)]
radv: Update code pointer correctly if a variant is already created
This was the actual cause of GPU hangs fixed by
0fdd531457ec ("radv:
Fix pipeline cache locking issues"), since multiple threads would end
up trying to create the variants for a single entry.
Now that we're locking around the whole of this function, this isn't
really necessary (we either create all or none of the variants), but
fix this anyway in case things change later.
Signed-off-by: Alex Smith <asmith@feralinteractive.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
CC: 17.3 <mesa-stable@lists.freedesktop.org>
Kenneth Graunke [Thu, 19 Oct 2017 21:38:30 +0000 (14:38 -0700)]
i965: Revert absolute mode for constant buffer pointers.
The kernel doesn't initialize the value of the INSTPM or CS_DEBUG_MODE2
registers at context initialization time. Instead, they're inherited
from whatever happened to be running on the GPU prior to first run of a
new context. So, when we started setting these, other contexts in the
system started inheriting our values. Since this controls whether
3DSTATE_CONSTANT_* takes a pointer or an offset, getting the wrong
setting is fatal for almost any process which isn't expecting this.
Unfortunately, VA-API and Beignet don't initialize this (nor does older
Mesa), so they will die horribly if we start doing this. UXA and SNA
don't use any push constants, so they are unaffected.
Until we have some kind of solution to this problem, I'm going to revert
this patch and abandon using the feature for now. It will lead to fewer
pushed UBO ranges on Broadwell+, which may lead to lower performance,
though I don't have any data on the impact.
Cc: "17.3 17.2" <mesa-stable@lists.freedesktop.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102774
Dylan Baker [Fri, 20 Oct 2017 22:57:15 +0000 (15:57 -0700)]
meson: build imx driver
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Dylan Baker [Fri, 20 Oct 2017 22:45:22 +0000 (15:45 -0700)]
meson: build etnaviv driver + winsys
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Eric Anholt [Tue, 17 Oct 2017 20:46:50 +0000 (13:46 -0700)]
ac: Silence a compiler warning about results[0].
We know that num_components will be > 0, but it doesn't.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Eric Anholt [Tue, 17 Oct 2017 20:45:09 +0000 (13:45 -0700)]
ac: Fix a compiler warning for possibly undefined "name"
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Dylan Baker [Sat, 21 Oct 2017 00:34:51 +0000 (17:34 -0700)]
meson: fix egl build for meson version < 0.43
Meson 0.43 added the ability to pass nested lists to
include_directories, so the code that we have works for 0.43, but not
for 0.42. This patch changes the include_directories list to be flat so
it works with 0.42
fixes:
108d257a16859898f5c ("meson: build libEGL")
Tested-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Rhys Kidd <rhyskidd@gmail.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Nicolai Hähnle [Thu, 12 Oct 2017 09:21:26 +0000 (11:21 +0200)]
amd/common/gfx9: workaround DCC corruption more conservatively
Fixes KHR-GL45.texture_swizzle.smoke and others on Vega.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102809
Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Emil Velikov [Mon, 23 Oct 2017 13:30:37 +0000 (14:30 +0100)]
docs/release-calendar: update - 17.3.0-rc1 is out
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Ilia Mirkin [Sat, 21 Oct 2017 19:15:41 +0000 (15:15 -0400)]
glsl: fix derived cs variables
There are two issues with the current implementation. First, it relies
on the layout(local_size_*) happening in the same shader as the main
function, and secondly it doesn't work for variable group sizes.
In both cases, the simplest fix is to move the setup of these derived
values to a later time, similar to how the gl_VertexID workarounds are
done. There already exist system values defined for both of the derived
values, so we use them unconditionally, and lower them after linking is
performed.
While we're at it, we move to using gl_LocalGroupSizeARB instead of
gl_WorkGroupSize for variable group sizes.
Also the dead code elimination avoidance can be removed, since there
can be situations where gl_LocalGroupSizeARB is needed but has not been
inserted for the shader with main function. As a result, the lowering
code has to insert its own copies of the system values if needed.
Reported-by: Stephane Chevigny <stephane.chevigny@polymtl.ca>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103393
Cc: mesa-stable@lists.freedesktop.org
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Emil Velikov [Mon, 23 Oct 2017 12:07:06 +0000 (13:07 +0100)]
docs: add 17.4.0-devel release notes template
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Emil Velikov [Mon, 23 Oct 2017 12:00:43 +0000 (13:00 +0100)]
mesa: bump version to 17.4.0-devel
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Juan A. Suarez Romero [Thu, 19 Oct 2017 10:46:02 +0000 (12:46 +0200)]
radv: automake: include radv_extensions.py in the tarball
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Bas Nieuwenhuizen [Mon, 23 Oct 2017 01:27:28 +0000 (03:27 +0200)]
ac/nir: Only clamp shadow reference on radeonsi.
Vulkan CTS does not expect the value to be clamped (at least for D32),
and it makes a differences even though depth is in [0,1], due
to strict inequalities.
I couldn't find anything in the Vulkan spec about this, but the test
seemed to be copied from GL tests and the GL spec only specifies
clamping for fixed point formats. Hence I expect radeonsi to run into
this at some point as well, but given that they still have a usecase
with the Z16->Z32 promotion, I'll leave that for someone else to clean
up.
This at least fixes radv dEQP-VK.texture.shadow.* on VI.
Fixes:
0f9e32519bb 'ac/nir: clamp shadow texture comparison value on VI'
Reviewed-by: Dave Airlie <airlied@redhat.com>
Bas Nieuwenhuizen [Sun, 22 Oct 2017 16:43:14 +0000 (18:43 +0200)]
radv: Disallow indirect outputs for GS on GFX9 as well.
Since it also uses the output vector before writing to memory.
Fixes:
e38685cc62e 'Revert "radv: disable support for VEGA for now."'
Reviewed-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Bas Nieuwenhuizen [Sun, 22 Oct 2017 16:32:14 +0000 (18:32 +0200)]
ac/nir: Fix nir_texop_lod on GFX for 1D arrays.
Fixes:
1bcb953e166 'radv: handle GFX9 1D textures'
Reviewed-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Sun, 22 Oct 2017 20:23:29 +0000 (06:23 +1000)]
radv/ac/nir: only emit tess factors to storage if tes reads them
Otherwise we just need to write them to the tf ring.
this seems to improve the tessellation demo on Bonarie
~2190->~2230 fps
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Bas Nieuwenhuizen [Sat, 21 Oct 2017 22:56:09 +0000 (00:56 +0200)]
radv: Don't use vgpr indexing for outputs on GFX9.
Due to LLVM bugs. Fixes a bunch of dEQP-VK.glsl.indexing.*
tests.
Fixes:
e38685cc62e 'Revert "radv: disable support for VEGA for now."'
Reviewed-by: Dave Airlie <airlied@redhat.com>
Bas Nieuwenhuizen [Sat, 21 Oct 2017 20:00:37 +0000 (22:00 +0200)]
ac/nir: Account for compact array index in GS input load from LDS.
Mirrors the vram path.
Fixes:
d4ecc3c9299 'ac/nir: Add loading from LDS for merged GS.'
Reviewed-by: Dave Airlie <airlied@redhat.com>
Bas Nieuwenhuizen [Sat, 21 Oct 2017 19:14:13 +0000 (21:14 +0200)]
radv: Don't compile shaders when they are cached already.
When the gs_copy_shader is NULL (due to an incomplete cache), but
the main shaders are found, we still do the nir, but we shouldn't
compile the shaders again. For merged shaders we should also account
for the missing shaders.
Fixes:
ce03c119ce0 'radv: Add code to compile merged shaders.'
Reviewed-by: Dave Airlie <airlied@redhat.com>
Bas Nieuwenhuizen [Sat, 21 Oct 2017 18:16:57 +0000 (20:16 +0200)]
radv: Don't check for max GL GS invocations.
We specify 127 instead of 32 as the limit in vulkan.
Fixes:
6bc42855f92 'radv: enable GS on GFX9'
Reviewed-by: Dave Airlie <airlied@redhat.com>
Bas Nieuwenhuizen [Sat, 21 Oct 2017 16:23:47 +0000 (18:23 +0200)]
radv: Don't explicitly reference vertex shader for draw_id.
With merged shaders the vertex shader may not exist. This got in
because the offending patch was written before merged shaders were
upstream, but committed after.
Fixes:
75dfab24a2c 'radv: refactor indirect draws with radv_draw_info'
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Bas Nieuwenhuizen [Sat, 21 Oct 2017 16:20:24 +0000 (18:20 +0200)]
radv: Don't reset cmd_buffer->state.dirty.
Otherwise for non-indexed draws we set and immediately unset
RADV_CMD_DIRTY_INDEX_BUFFER. As all the set functions should
clear their own bit, this is unnecessary.
Fixes:
341529dbee5 'radv: use optimal packet order for draws'
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Bas Nieuwenhuizen [Sat, 21 Oct 2017 16:16:15 +0000 (18:16 +0200)]
radv: Correctly detect changed shaders for vertex descriptors.
As they were emitted after the new pipeline, the changed pipeline
detection was not working anymore.
Fixes:
341529dbee5 'radv: use optimal packet order for draws'
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Bas Nieuwenhuizen [Sat, 21 Oct 2017 01:07:43 +0000 (03:07 +0200)]
ac/nir: Set larged wrokgroup size for GS on GFX9.
They don't take a single wave anymore and we need the barriers.
Fixes:
6bc42855f92 'radv: enable GS on GFX9'
Reviewed-by: Dave Airlie <airlied@redhat.com>
Bas Nieuwenhuizen [Sat, 21 Oct 2017 01:04:35 +0000 (03:04 +0200)]
ac/nir: Take the max workgroup size of all provided shaders.
Fixes:
ffaf4d608a1 'radv: Enable tessellation shaders for GFX9.'
Reviewed-by: Dave Airlie <airlied@redhat.com>
Alex Smith [Thu, 19 Oct 2017 10:49:39 +0000 (11:49 +0100)]
radv: Fix pipeline cache locking issues
Need to lock around the whole process of retrieving cached shaders, and
around GetPipelineCacheData.
This fixes GPU hangs observed when creating multiple pipelines in
parallel, which appeared to be due to invalid shader code being pulled
from the cache.
Signed-off-by: Alex Smith <asmith@feralinteractive.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Lionel Landwerlin [Fri, 20 Oct 2017 17:49:30 +0000 (18:49 +0100)]
anv: don't assert on device init on Cannonlake
v2: Warn that support is still in alpha (Jason)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Lionel Landwerlin [Fri, 20 Oct 2017 17:28:48 +0000 (18:28 +0100)]
anv: disable stencil pma fix on Gen > 9
This workaround isn't listed on Gen10.
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Lionel Landwerlin [Fri, 20 Oct 2017 14:46:48 +0000 (15:46 +0100)]
blorp: enable R32G32B32X32 blorp ccs copies
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Eric Anholt [Sat, 21 Oct 2017 00:14:52 +0000 (17:14 -0700)]
meson: Fix vc5 deps on the XML-generated headers.
I typoed and was depending on v3d_xml.h (the gzipped xml)_, not on the
v3d_packet_v33_pack.h that the compiler and QPU packing actually use.
Eric Anholt [Sat, 21 Oct 2017 00:05:54 +0000 (17:05 -0700)]
broadcom/vc5: Propagate vc4 aliasing fix to vc5.
See
e5fea0d621af2b14cf6c5e364eeaf293db460f2a
Stefan Schake [Fri, 20 Oct 2017 14:16:44 +0000 (16:16 +0200)]
broadcom/vc4: Fix aliasing issue
This was causing Android clang version 3.8.256229 to miscompile,
presumably due to strict aliasing.
Fixes:
14dc281c1332 ("vc4: Enforce one-uniform-per-instruction after optimization.")
Dylan Baker [Wed, 18 Oct 2017 22:11:39 +0000 (15:11 -0700)]
meson: Add support for EGL glvnd
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Tested-by: Eric Engestrom <eric.engestrom@imgtec.com>
Reviewed-by: Lyude Paul <lyude@redhat.com>
Dylan Baker [Wed, 18 Oct 2017 19:20:43 +0000 (12:20 -0700)]
meson: build libEGL
This is based heavily on Daniel Stone's work for the same, rebased on
master and with a number of TODO's fixed.
This does not implement glvnd (which is coming in a later patch)
Meson builds egl slightly differently than autotools, namely it doesn't
build an intermediate shared library. It doesn't do this because meson
doesn't have problems with the name of the library being dynamically
generated, so the glvnd and non-glvnd code can follow the same path.
v2: - Don't reuse variable (Eric E.)
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Tested-by: Eric Engestrom <eric.engestrom@imgtec.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
Dylan Baker [Tue, 17 Oct 2017 22:02:58 +0000 (15:02 -0700)]
meson: move wayland_drm_protocol generation to wayland-drm
These files are needed by both vulkan wayland-wsi and by egl
wayland-wsi, since the XML file is in src/egl/wayland/wayland-drm and we
can include this directory in such a way that it will be loaded before
egl and vulkan this allows us to avoid multiple calls to the same
generator.
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Reviewed-and-Tested-by: Eric Engestrom <eric@engestrom.ch>
Dylan Baker [Wed, 18 Oct 2017 22:47:11 +0000 (15:47 -0700)]
meson: Don't allow glx to be built without platform_x11
Previously this failed to change with_glx to disabled from auto if
platform_x11 was unset or if no opengl apis were being built.
v2: - swap conditional positions
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Reviewed-and-Tested-by: Eric Engestrom <eric@engestrom.ch>
Dylan Baker [Fri, 20 Oct 2017 23:28:32 +0000 (16:28 -0700)]
meson: bump libdrm_amdgpu requirement to 2.4.85
fixes:
b60372570323243ec73 ("configure.ac: Bump libdrm_amdgpu version to 2.4.85.")
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Eric Anholt [Tue, 17 Oct 2017 17:41:55 +0000 (10:41 -0700)]
nir: Print the components referenced for split or packed shader in/outs.
Having 4 variables all called "gl_in_TexCoord0@n" isn't very informative,
much better to see:
decl_var shader_in INTERP_MODE_NONE float gl_in_TexCoord0 (VARYING_SLOT_VAR0.x, 1, 0)
decl_var shader_in INTERP_MODE_NONE float gl_in_TexCoord0@0 (VARYING_SLOT_VAR0.y, 1, 0)
decl_var shader_in INTERP_MODE_NONE float gl_in_TexCoord0@1 (VARYING_SLOT_VAR0.z, 1, 0)
decl_var shader_in INTERP_MODE_NONE float gl_in_TexCoord0@2 (VARYING_SLOT_VAR0.w, 1, 0)
v2: Handle arrays and structs better (by Timothy)
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Eric Anholt [Mon, 16 Oct 2017 22:32:30 +0000 (15:32 -0700)]
nir: Add a safety check that we don't remove dead I/O vars after lowering.
The pass only looks at var load/store intrinsics, not input load/store
intrinsics, so assert that we don't see the other type.
v2: Adjust comment indentation.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Andres Rodriguez [Fri, 20 Oct 2017 22:42:13 +0000 (18:42 -0400)]
radv: disable implicit sync for radv allocated bos v3
Implicit sync kicks in when a buffer is used by two different amdgpu
contexts simultaneously. Jobs that use explicit synchronization
mechanisms end up needlessly waiting to be scheduled for long periods
of time in order to achieve serialized execution.
This patch disables implicit synchronization for all radv allocations
except for wsi bos. The only systems that require implicit
synchronization are DRI2/3 and PRIME.
v2: mark wsi bos as RADV_MEM_IMPLICIT_SYNC
v3: Add drm version check (Bas)
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Andres Rodriguez [Fri, 20 Oct 2017 22:42:12 +0000 (18:42 -0400)]
radv: factor out radv_alloc_memory
This allows us to pass extra parameters to the memory allocation
operation that are not defined in the vulkan spec. This is useful for
internal usage.
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Andres Rodriguez [Fri, 20 Oct 2017 22:02:16 +0000 (18:02 -0400)]
radv: Expose VK_EXT_global_priority
Expose the extension string as supported
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Andres Rodriguez [Fri, 20 Oct 2017 22:02:15 +0000 (18:02 -0400)]
radv: don't skip PS/VS partial flush
This patch helps lower high priority compute latency. Found by
bisecting a perf regression on computeparticles with high priority
compute queues enabled.
Reverting this micro-optimization doesn't seem to have any negative
effect on performance on Dota2 or ssao.
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Andres Rodriguez [Fri, 20 Oct 2017 22:02:14 +0000 (18:02 -0400)]
radv: Implement VK_EXT_global_priority
This extension allows the caller to change a queue's system wide
priority. This is useful for applications with specific
latency constraints.
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Andres Rodriguez [Fri, 20 Oct 2017 22:02:13 +0000 (18:02 -0400)]
radeonsi: hardcode shader WAVE_LIMIT to the maximum value
This is part of a cooperative scheduling approach used by radv. All
drivers in the stack must opt-in to resource arbitration, otherwise GL
based apps will be able to ignore system priorities.
We always hardcode the field to its maximum value, instead of attempting
to calculate an approximate usage. In testing, there were no benefits to
using anything other than the maximum.
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Andres Rodriguez [Fri, 20 Oct 2017 22:02:12 +0000 (18:02 -0400)]
radv: hardcode shader WAVE_LIMIT to the maximum value
When WAVE_LIMIT is set, a submission will opt-in for SPI based resource
scheduling. Because this mechanism is cooperative, we must ensure that
all submissions have this field set, otherwise they will bypass resource
arbitration.
We always hardcode the field to its maximum value, instead of attempting
to calculate an approximate usage. In testing, there were no benefits to
using anything other than the maximum.
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Andres Rodriguez [Fri, 20 Oct 2017 22:02:11 +0000 (18:02 -0400)]
vulkan: update headers & registry to VK 1.0.63
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Bas Nieuwenhuizen [Fri, 20 Oct 2017 22:22:48 +0000 (00:22 +0200)]
configure.ac: Bump libdrm_amdgpu version to 2.4.85.
For VK_EXT_global_priority in radv.
Acked-by: Andres Rodriguez <andresx7@gmail.com>
Eric Anholt [Fri, 20 Oct 2017 21:00:11 +0000 (14:00 -0700)]
broadcom/vc5: Use SETMSF to handle discards.
A bit of spec text suggested that (like vc4) condition codes should be
used for discards, and the simulator was fine with it, but the 7268
disagrees and you have to use SETMSF instead or the color comes through.
Fixes glsl-fs-discard-01 and many of the interpolation-with-clipping
tests.
Eric Anholt [Fri, 20 Oct 2017 19:43:11 +0000 (12:43 -0700)]
broadcom/vc5: Set the snorm/unorm packing functions to be lowered.
We don't have native instructions for them, so set up the lowering. Once
we support the bfi instructions that get generated, they should start
actually working.
Eric Anholt [Fri, 20 Oct 2017 19:32:36 +0000 (12:32 -0700)]
broadcom/vc5: Fix pasteo that broke vertex texturing.
We weren't ever filling in the texture state record, so we'd dereference
NULL from the shader.
Eric Anholt [Thu, 19 Oct 2017 22:22:13 +0000 (15:22 -0700)]
broadcom/vc5: Move default attribute value setup to the CSO and fix them.
I was generating some stub values to bring the driver up, but fill them in
properly now. We now set 1.0 or 1u as appropriate, and thanks to being in
their own BO it fixes piglit failures on the 7268 (where our 4-byte
alignment was insufficient).
Fixes const-packHalf2x16.shader_test