Alex Deucher [Wed, 10 Jun 2015 15:39:30 +0000 (11:39 -0400)]
radeonsi: properly set the raster_config for KV
This enables the second RB on asics that support it which
should boost performance.
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: mesa-stable@lists.freedesktop.org
Marek Olšák [Thu, 30 Apr 2015 15:02:38 +0000 (17:02 +0200)]
radeonsi: add amdgpu support for querying the GPU reset state
Reviewed-by: Christian König <christian.koenig@amd.com>
Marek Olšák [Thu, 16 Apr 2015 18:44:54 +0000 (20:44 +0200)]
radeonsi: add VI hardware support
Marek Olšák [Sat, 11 Jul 2015 11:22:22 +0000 (13:22 +0200)]
radeonsi: add definitions for VI status registers
Useful for debugging hangs with the read-register interface.
I checked that this adds the same register fields as the kernel driver.
Acked-by: Michel Dänzer <michel.daenzer@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Marek Olšák [Thu, 16 Apr 2015 18:12:24 +0000 (20:12 +0200)]
radeonsi: add VI register definitions
Marek Olšák [Thu, 16 Apr 2015 18:35:27 +0000 (20:35 +0200)]
radeonsi: fix DRM version checks for amdgpu DRM 3.0.0
Marek Olšák [Thu, 16 Apr 2015 17:41:33 +0000 (19:41 +0200)]
winsys/amdgpu: add addrlib - texture addressing and alignment calculator
This is an internal project that Catalyst uses and now open source will do
too.
v2: squashed these commits in:
- winsys/amdgpu: fix warnings in addrlib
- winsys/amdgpu: set PIPE_CONFIG and NUM_BANKS in tiling_flags
Marek Olšák [Thu, 16 Apr 2015 20:43:23 +0000 (22:43 +0200)]
winsys/amdgpu: add a new winsys for the new kernel driver
v2: - lots of changes according to Emil Velikov's comments
- implemented radeon_winsys::read_registers
v3: - a lot of new work, many of them adapt to libdrm interface changes
Squashed patches:
winsys/amdgpu: implement radeon_winsys context support
winsys/amdgpu: add reference counting for contexts
winsys/amdgpu: add userptr support
winsys/amdgpu: allocate IBs like normal buffers
winsys/amdgpu: add IBs to the buffer list, adapt to interface changes
winsys/amdgpu: don't use KMS handles as reloc hash keys
winsys/amdgpu: sync buffer accesses to different rings
winsys/amdgpu: use dependencies instead of waiting for last fence v2
gallium/radeon: unify buffer_wait and buffer_is_busy in the winsys interface (amdgpu part)
winsys/amdgpu: track fences per ring and be thread-safe
winsys/amdgpu: simplify waiting on a variable in amdgpu_fence_wait
gallium/radeon: allow the winsys to choose the IB size (amdgpu part)
winsys/amdgpu: switch to new amdgpu_cs_query_fence_status interface
winsys/amdgpu: handle fence and dependencies merge
winsys/amdgpu follow libdrm change to move user fence into UMD
winsys/amdgpu: use amdgpu_bo_va_op for va map/unmap v2
winsys/amdgpu: use the new tiling flags
winsys/amdgpu: switch to new GTT_USWC definition
winsys/amdgpu: expose amdgpu_cs_query_reset_state to drivers
winsys/amdgpu: fix valgrind warnings
winsys/amdgpu: don't use VRAM with APUs that don't have much of it
winsys/amdgpu: require LLVM 3.6.1 for VI because of bug fixes there
winsys/amdgpu: remove amdgpu_winsys::num_cpus
winsys/amdgpu: align BO size to page size
winsys/amdgpu: reduce BO cache timeout
winsys/amdgpu: remove useless flushing and waiting in amdgpu_bo_set_tiling
winsys/amdgpu: use amdgpu_device_handle as a unique device ID instead of fd
winsys/amdgpu: use safer access to amdgpu_fence_wait::signalled
winsys/amdgpu: allow maximum IB size of 4 MB
winsys/amdgpu: add ip_instance into amdgpu_fence
gallium/radeon: add RING_COMPUTE instead of RADEON_FLUSH_COMPUTE
winsys/amdgpu: set the ring type at CS initilization
winsys/amdgpu: query the GART page size from the kernel
winsys/amdgpu: correctly wait for shared buffers to become idle
winsys/amdgpu: set the amdgpu_cs_fence structure only once at fence creation
winsys/amdgpu: add a specific error message for cs_submit -> -ENOMEM
winsys/amdgpu: check num_active_ioctls before calling amdgpu_bo_wait_for_idle
winsys/amdgpu: clear user fence BO after allocating it
winsys/amdgpu: fix user fences
winsys/amdgpu: make amdgpu_winsys_create public
winsys/amdgpu: remove thread offloading
winsys/amdgpu: flatten the amdgpu_cs_context structure and simplify more
v4: require libdrm 2.4.63
Christian König [Wed, 29 Apr 2015 13:35:02 +0000 (15:35 +0200)]
st/vdpau: add HEVC support v2
v2: fix return code
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Leo Liu [Thu, 25 Jun 2015 17:19:56 +0000 (13:19 -0400)]
st/omx/enc: stack frame tasks for the gathering
Put tasks to the FIFO queue for results
Signed-off-by: Leo Liu <leo.liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Leo Liu [Fri, 29 May 2015 18:50:44 +0000 (14:50 -0400)]
st/omx/enc: flush after eos handling v2
v2 (chk): reorder the flush
Signed-off-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Tue, 28 Apr 2015 13:31:37 +0000 (15:31 +0200)]
vl: add HEVC profiles and defines
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Leo Liu [Thu, 25 Jun 2015 16:09:11 +0000 (12:09 -0400)]
vl: add cap for stacking frames
Signed-off-by: Leo Liu <leo.liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Marek Olšák [Wed, 12 Aug 2015 23:51:37 +0000 (01:51 +0200)]
swrast: fix EXT_depth_bounds_test
zMin and zMax can't use _DepthMaxF, because the test is done in Z32_UNORM.
Probably a useless patch given how popular swrast is nowadays, but it helped
create and validate the piglit test.
v2: add an explicit cast to GLuint
Reviewed-by: Brian Paul <brianp@vmware.com>
Marek Olšák [Mon, 10 Aug 2015 00:23:21 +0000 (02:23 +0200)]
radeonsi: add support for EXT_depth_bounds_test
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Mon, 10 Aug 2015 00:18:43 +0000 (02:18 +0200)]
st/mesa: add EXT_depth_bounds_test
Reviewed-by: Brian Paul <brianp@vmware.com>
Marek Olšák [Mon, 10 Aug 2015 00:11:48 +0000 (02:11 +0200)]
gallium: add an interface for EXT_depth_bounds_test
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Marek Olšák [Mon, 10 Aug 2015 17:53:22 +0000 (19:53 +0200)]
st/mesa: small cleanup in st_extensions.c
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Edward O'Callaghan <eocallaghan at alterapraxis.com>
Marek Olšák [Mon, 10 Aug 2015 17:37:01 +0000 (19:37 +0200)]
gallium: add support for GLES texture float extensions (v3)
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=74329
v2: add a CAP for half floats
drivers should not expose the CAPs if they don't support the formats
v3: update relnotes
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Zoltan Gilian [Tue, 7 Jul 2015 21:38:27 +0000 (23:38 +0200)]
r600,compute: setup compute sampler states and views
v2: Add compute mode flag to sampler state setup (Marek).
Drop branches which avoid reference counting (Marek).
Simplify unset branch condition (Marek).
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Michel Dänzer [Fri, 14 Aug 2015 06:16:12 +0000 (15:16 +0900)]
st/clover: Fix build against LLVM 3.8 SVN r244928
raw_svector_ostream::flush() is now unnecessary and forbidden:
CXX llvm/libclllvm_la-invocation.lo
../../../../../src/gallium/state_trackers/clover/llvm/invocation.cpp: In function 'clover::module {anonymous}::build_module_llvm(llvm::Module*, unsigned int (&)[7])':
../../../../../src/gallium/state_trackers/clover/llvm/invocation.cpp:574:29: error: use of deleted function 'void llvm::raw_svector_ostream::flush()'
bitcode_ostream.flush();
^
In file included from /home/daenzer/src/llvm-git/llvm/include/clang/Basic/VirtualFileSystem.h:22:0,
from /home/daenzer/src/llvm-git/llvm/include/clang/Basic/FileManager.h:20,
from /home/daenzer/src/llvm-git/llvm/include/clang/Basic/SourceManager.h:38,
from /home/daenzer/src/llvm-git/llvm/include/clang/Frontend/CompilerInstance.h:16,
from ../../../../../src/gallium/state_trackers/clover/llvm/invocation.cpp:25:
/home/daenzer/src/llvm-git/llvm/include/llvm/Support/raw_ostream.h:512:8: note: declared here
void flush() = delete;
^
Makefile:862: recipe for target 'llvm/libclllvm_la-invocation.lo' failed
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
Tapani Pälli [Thu, 13 Aug 2015 06:30:35 +0000 (09:30 +0300)]
mesa: set correct error for non-renderable multisample textures
v2: same common error on gles31 and desktop OpenGL
(spotted by Erik Faye-Lund)
Signed-off-by: Marta Lofstedt <marta.lofstedt@linux.intel.com>
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Timothy Arceri <t_arceri@yahoo.com.au>
Ben Widawsky [Fri, 7 Aug 2015 20:58:37 +0000 (13:58 -0700)]
i965/skl: Remove early platform support
We do not want bug reports from this early stepping of SKL. Few if any were ever
shipped outside of Intel to early enabling partners, and none will be sold.
There is a functional change here. If you're using new mesa on an old
kernel/libdrm, the revid will be -1, and we'll use new SKL values instead of
early ones (a hopefully irrelevant improvement IMO).
v2: Remove hunk which warned before dying. Instead, default to normal SKL
support (Ken)
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Neil Roberts <neil@linux.intel.com>
Frank Binns [Wed, 12 Aug 2015 15:36:00 +0000 (16:36 +0100)]
egl: improve attribute checking for eglCreateContext
The EGL 1.4 spec states for eglCreateContext:
"attribute EGL_CONTEXT_CLIENT_VERSION is only valid when the current
rendering API is EGL_OPENGL_ES_API"
Additionally, if the EGL_KHR_create_context EGL extension is supported
(this is mandatory in EGL 1.5) then the EGL_CONTEXT_MAJOR_VERSION_KHR,
which is an alias for EGL_CONTEXT_CLIENT_VERSION, and
EGL_CONTEXT_MINOR_VERSION_KHR attributes are also accepted by
eglCreateContext with the extension spec stating:
"The values for attributes EGL_CONTEXT_MAJOR_VERSION_KHR and
EGL_CONTEXT_MINOR_VERSION_KHR specify the requested client API
version. They are only meaningful for OpenGL and OpenGL ES
contexts, and specifying them for other types of contexts will
generate an error."
Add the necessary checks against the extension and rendering APIs when
validating these attributes as part of eglCreateContext.
Signed-off-by: Frank Binns <frank.binns@imgtec.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
[Emil Velikov: Add newline before the spec quote (Matt)]
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Frank Binns [Wed, 12 Aug 2015 15:35:59 +0000 (16:35 +0100)]
egl: don't allow eglGetConfigs to set num_configs param to a negative value
When a buffer is provided to eglGetConfigs it's supposed to set the value
of the num_config parameter to the total number of configs that have been
copied into this buffer. For some reason the EGL spec doesn't consider it
to be an error to pass this function a buffer while specifying its size to
be less than 0. Given this, one would expect this combination to result in
the num_config parameter being set to 0 but this wasn't the case. This was
due to the buffer size being copied straight into num_configs without being
clamped to 0.
This was causing the following dEQP EGL test to fail:
dEQP-EGL.functional.query_config.get_configs.get_configs_bounds
Signed-off-by: Frank Binns <frank.binns@imgtec.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Frank Binns [Tue, 4 Aug 2015 13:32:45 +0000 (14:32 +0100)]
egl/x11: don't abort when creating a DRI2 drawable fails
When calling either eglCreateWindowSurface or eglCreatePixmapSurface it
was possible for an application to be aborted as a result of it failing
to create a DRI2 drawable on the server. This could happen due to an
application passing in an invalid native drawable handle, for example.
v2: Handle the case where an error has been set on the connection
Cc: <mesa-stable@lists.freedesktop.org>
Signed-off-by: Frank Binns <frank.binns@imgtec.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Frank Binns [Tue, 4 Aug 2015 13:32:44 +0000 (14:32 +0100)]
egl/x11: set EGL_BAD_NATIVE_(PIXMAP|WINDOW) for invalid pixmaps/windows
Both eglCreatePixmapSurface and eglCreateWindowSurface were incorrectly
setting the EGL error to be EGL_BAD_ALLOC when an invalid native drawable
handle was being passed in. The EGL spec states the following for
eglCreatePixmapSurface:
"If pixmap is not a valid native pixmap handle, then an EGL_BAD_-
NATIVE_PIXMAP error should be generated."
(eglCreateWindowSurface has similar text)
Correctly set the EGL error value based on xcb_get_geometry_reply returning
an error structure containing something other than BadAlloc.
v2: Check for BadAlloc error and update commit message to reflect this
Signed-off-by: Frank Binns <frank.binns@imgtec.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Frank Binns [Tue, 4 Aug 2015 13:32:43 +0000 (14:32 +0100)]
egl/x11: fix use of EGL_BAD_NATIVE_WINDOW
Commit
4ed23fd590 introduced some calls to _eglError inappropriately
passing it EGL_BAD_NATIVE_WINDOW. This was actually harmless in two of the
cases as _eglError gets called later on with a more appropriate error code
but (just to be safe) switch these to _eglLog calls instead.
The final case is a little trickier as it actually needs to set an error
of which the following are available (according to the EGL spec):
EGL_BAD_MATCH, EGL_BAD_CONFIG, EGL_BAD_NATIVE_(PIXMAP|WINDOW) and
EGL_BAD_ALLOC.
Of these, EGL_BAD_ALLOC seems to be the most appropriate given that
failure can occur either as a result of xcb_get_setup failing due to an
earlier error on the connection (where the most commonly occurring error
code is XCB_CONN_CLOSED_MEM_INSUFFICIENT) or as a result of the
xcb_screen_iterator_t 'rem' field being 0.
In addition to this, commit
af2aea40d2 unconditionally set the error to
EGL_BAD_NATIVE_WINDOW when creating a window or pixmap surface with a NULL
native handle. Change this to correctly set the error based on surface
type.
v2: Updated patch description (Emil Velikov)
Return EGL_BAD_NATIVE_PIXMAP when eglCreatePixmapSurface is called
with a NULL native pixmap handle
Signed-off-by: Frank Binns <frank.binns@imgtec.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Timothy Arceri [Thu, 13 Aug 2015 08:42:54 +0000 (18:42 +1000)]
mesa: remove extern from texture function
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Timothy Arceri [Wed, 12 Aug 2015 07:01:52 +0000 (17:01 +1000)]
glsl: make linker error message more informative
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Topi Pohjolainen [Thu, 25 Jun 2015 11:00:41 +0000 (14:00 +0300)]
i965: Stop aux data compare preventing program binary re-use
Items in the program cache consist of three things: key, the data
representing the instructions and auxiliary data representing
uniform storage. The data consisting of instructions is stored into
a drm buffer object while the key and the auxiliary data reside in
malloced section. Now the cache uploading is equipped with a check
that iterates over existing items and seeks to find a another item
using identical instruction data than the one being just uploaded.
If such is found there is no need to add another section into the
drm buffer object holding identical copy of the existing one. The
item just being uploaded should instead simply point to the same
offset in the underlying drm buffer object.
Unfortunately the check for the matching instruction data is
coupled with a check for matching auxiliary data also. This
effectively prevents the cache from ever containing two items
that could share a section in the drm buffer object.
The constraint for the instruction data and auxiliary data to
match is, fortunately, unnecessary strong. When items are stored
into the cache they will anyway contain their own copy of the
auxiliary data (even if they matched - which they in real world
never will). The only thing the items would be sharing is the
instruction data and hence we should only check for that to match
and nothing else.
No piglit regression in jenkins.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Topi Pohjolainen [Thu, 25 Jun 2015 11:35:26 +0000 (14:35 +0300)]
i965: Only write program to cache when it doesn't exist yet
Current logic re-writes the same data when existing data is found.
Not that this actually matters at the moment in practice, the
contraint for finding matching data is too severe to ever allow
data to be shared between two items in the cache.
Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Topi Pohjolainen [Thu, 25 Jun 2015 11:31:03 +0000 (14:31 +0300)]
i965: Rename brw_upload_item_data to brw_alloc_item_data
and simplify the interface to take directly the size and to return
the offset. The routine does nothing more than allocate, it doesn't
upload anything.
Reviewed-by: Anuj Phogat <anuj.phogat@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Tapani Pälli [Wed, 12 Aug 2015 08:13:40 +0000 (11:13 +0300)]
mesa: update MaxShaderStorageBlockSize to 2^27
Extension spec originally required 2^24 but 2^27 is the minimum value
required by OpenGL 4.5 and OpenGL ES 3.1 specifications.
Fixes:
ES31-CTS.shader_storage_buffer_object.basic-max
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Timothy Arceri <t_arceri@yahoo.com.au>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Tapani Pälli [Mon, 3 Aug 2015 07:46:33 +0000 (10:46 +0300)]
mesa: fix name returned for XFB varyings
_mesa_get_program_resource_name has logic to append '[0]' in name
if variable is an array, this should be skipped for XFB varyings
that have array index already appended.
v2: fix comment, change also GL_NAME_LENGTH query to match
the behaviour
Fixes:
ES31-CTS.program_interface_query.transform-feedback-types
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Martin Peres <martin.peres@linux.intel.com>
Edward O'Callaghan [Wed, 29 Jul 2015 11:49:45 +0000 (21:49 +1000)]
mesa: Fix printf format specifier warn of the ptrdiff_t
See §7.19.6.1, paragraph 7 of the ISO C specification.
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Marek Olšák [Tue, 11 Aug 2015 19:37:59 +0000 (21:37 +0200)]
r600g: allow setting geometry shader sampler states
We were ignoring them. This is both hilarious and sad.
Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Edward O'Callaghan <eocallaghan at alterapraxis.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Marek Olšák [Tue, 11 Aug 2015 20:36:51 +0000 (22:36 +0200)]
r600g: fix polygon offset scale
The value was copied from r300g, which uses 1/12 subpixels, but this hw
uses 1/16 subpixels.
Should fix piglit: gl-1.4-polygon-offset (formerly a glean test)
(untested, ported from radeonsi)
Reviewed-by: Edward O'Callaghan <eocallaghan at alterapraxis.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Cc: mesa-stable@lists.freedesktop.org
Marek Olšák [Tue, 11 Aug 2015 20:36:51 +0000 (22:36 +0200)]
radeonsi: fix polygon offset scale
The value was copied from r300g, which uses 1/12 subpixels, but this hw
uses 1/16 subpixels.
Fixes piglit: gl-1.4-polygon-offset (formerly a glean test)
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Cc: mesa-stable@lists.freedesktop.org
Marek Olšák [Mon, 10 Aug 2015 00:28:01 +0000 (02:28 +0200)]
radeonsi: enable VS_OUT_MISC_SIDE_BUS_ENA
This is recommended for better performance.
Diag tests always enable this.
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Sun, 9 Aug 2015 23:50:11 +0000 (01:50 +0200)]
radeonsi: add support for gl_PrimitiveID in the fragment shader
It must be obtained from the VS.
The GS scenario A must be enabled for PrimID to be generated for the VS.
+ 4 piglits
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Marek Olšák [Sun, 9 Aug 2015 22:52:21 +0000 (00:52 +0200)]
radeonsi: move VGT_GS_MODE to the VS state
The VS will want to select GS scenario A here (VS with PrimitiveID).
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Rob Clark [Wed, 12 Aug 2015 15:39:24 +0000 (11:39 -0400)]
freedreno/a4xx: format updates
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Tue, 11 Aug 2015 20:47:16 +0000 (16:47 -0400)]
freedreno/a3xx+a4xx: add texture buffer object support
Basic texture buffer support. Should be straightforward to add first/
last_element support. And with a bit of work in ir3 emulate larger
texture buffer sizes. But this seems to be enough for stk gl31 render
paths.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Tue, 11 Aug 2015 20:33:14 +0000 (16:33 -0400)]
ttn: add buffer texture type
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Tue, 11 Aug 2015 20:11:04 +0000 (16:11 -0400)]
freedreno/ir3: 'keeps' need neighbors found too
This shows up with a glamor shader, which does a TXF and uses the result
for conditional kill. Before we wouldn't group the fanin (collect)
neighbors which need to be allocated adjacently at RA, resulting in
badness.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Tue, 11 Aug 2015 20:09:48 +0000 (16:09 -0400)]
freedreno/ir3/print: print left/right neighbors too
When debugging compiler, this is useful to see.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Tue, 11 Aug 2015 15:47:46 +0000 (11:47 -0400)]
freedreno/ir3: use nir pass to lower const to scalar
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Tue, 11 Aug 2015 12:48:34 +0000 (08:48 -0400)]
freedreno/a4xx: point-size and spritelist fixes
a4xx needs similar treatment as
995f55a6
Also fixup a few point-size and vpsrepl issues and drop fix_blit_fp()
hack previously needed for mem2gmem.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Tue, 11 Aug 2015 00:41:45 +0000 (20:41 -0400)]
freedreno: cap cleanups
Move a few things around to group stuff that is common to a3xx/a4xx
together. Also, introduce is_ir3() for things that are more specific to
the compiler / shader-ISA than to the gpu generation.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Rob Clark [Mon, 10 Aug 2015 10:58:37 +0000 (06:58 -0400)]
mesa: add NV_read_{depth,stencil,depth_stencil} extensions
These extensions allow reading depth/stencil for GLES contexts, which is
useful for tools like apitrace.
Signed-off-by: Rob Clark <robclark@freedesktop.org>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Jason Ekstrand [Thu, 6 Aug 2015 21:26:47 +0000 (14:26 -0700)]
i965/shader: Don't use OptimizeForAOS for NIR vec4 vertex shaders
Shader-db results for vec4 programs using NIR on HSW:
total instructions in shared programs: 1838157 -> 1828469 (-0.53%)
instructions in affected programs: 275978 -> 266290 (-3.51%)
helped: 2827
HURT: 244
GAINED: 0
LOST: 0
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eduardo Lima Mitev <elima@igalia.com>
Nanley Chery [Fri, 7 Aug 2015 23:37:47 +0000 (16:37 -0700)]
mesa/teximage: report the correct function which triggered the error
This function would always report that a dimension or size error occurred
in glTexImage even when it was called from glCompressedTexImage. Replace
the static string with the dynamically determined caller name.
Reviewed-by: Tapani Palli <tapani.palli@intel.com>
Signed-off-by: Nanley Chery <nanley.g.chery@intel.com>
Oded Gabbay [Wed, 12 Aug 2015 15:22:53 +0000 (18:22 +0300)]
mesa/formats: don't byteswap when building array formats
Because we build here an array format, we don't need to swap the
bytes for big endian.
If it isn't an array format, the bytes will be swapped in
_mesa_format_convert.
v2: remove temp variable
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Cc: "10.5 10.6" <mesa-stable@lists.freedesktop.org>
Jason Ekstrand [Mon, 10 Aug 2015 08:32:23 +0000 (01:32 -0700)]
mesa/formats: Don't flip channels of null array formats
Before, if we encountered an array format of 0 on a BE system, we would
flip all the channels even though it's an invalid format. This would
result in a mostly invalid format with a swizzle of yyyy or wwww. Instead,
we should just return 0 if the array format stashed in the format info is
invalid.
Cc: "10.6 10.5" <mesa-stable@lists.freedesktop.org>
Jason Ekstrand [Mon, 10 Aug 2015 06:45:44 +0000 (23:45 -0700)]
mesa/formats: Fix swizzle flipping for big-endian targets
The swizzle defines where in the format you should look for any given
channel. When we flip the format around for BE targets, we need to change
the destinations of the swizzles, not the sources. For example, say the
format is an RGBX format with a swizzle of xyz1 on LE. Then it should be
wzy1 on BE; however, the code as it was before, would have made it 1zyx on
BE which is clearly wrong.
Reviewed-by: Iago Toral <itoral@igalia.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Cc: "10.6 10.5" <mesa-stable@lists.freedesktop.org>
Jason Ekstrand [Sat, 8 Aug 2015 16:00:21 +0000 (09:00 -0700)]
mesa/formats: Only do byteswapping for packed formats
Reviewed-by: Iago Toral <itoral@igalia.com>
Cc: "10.6 10.5" <mesa-stable@lists.freedesktop.org>
Matt Turner [Tue, 11 Aug 2015 22:21:03 +0000 (15:21 -0700)]
configure.ac: Always define __STDC_LIMIT_MACROS.
... which ensures that we get defines like LONG_MAX in C++.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91591
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
Matt Turner [Tue, 11 Aug 2015 01:50:48 +0000 (18:50 -0700)]
i965: Optimize brw_inst_set_bits() and brw_compact_inst_set_bits().
Cuts about 2k of .text.
text data bss dec hex filename
5017141 197160 27672 5241973 4ffc75 i965_dri.so before
5014981 197160 27672 5239813 4ff405 i965_dri.so after
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Matt Turner [Mon, 10 Aug 2015 23:57:58 +0000 (16:57 -0700)]
i965: Optimize brw_inst_bits() and brw_compact_inst_bits().
Cuts about 1k of .text.
text data bss dec hex filename
5018165 197160 27672 5242997 500075 i965_dri.so before
5017141 197160 27672 5241973 4ffc75 i965_dri.so after
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Emil Velikov [Tue, 11 Aug 2015 18:00:03 +0000 (19:00 +0100)]
docs: add news item and link release notes for 10.6.4
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Emil Velikov [Tue, 11 Aug 2015 17:54:18 +0000 (18:54 +0100)]
docs: add sha256 checksums for 10.6.4
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
(cherry picked from commit
99793e2541510fe208d29e69fedf97a6fff006f8)
Emil Velikov [Tue, 11 Aug 2015 15:39:10 +0000 (16:39 +0100)]
docs: add release notes for 10.6.4
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
(cherry picked from commit
6b2fcee64edadbd4db2293f5f4fc1a70e80c7251)
Marek Olšák [Sat, 8 Aug 2015 12:03:54 +0000 (14:03 +0200)]
gallium/radeon: fix r600g build if LLVM is disabled
MESA_LLVM_VERSION_PATCH is undefined.
Reviewed-by: Edward O'Callaghan <eocallaghan at alterapraxis.com>
Tested-by: Benjamin Bellec <b.bellec@gmail.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Grazvydas Ignotas [Sun, 9 Aug 2015 21:42:35 +0000 (00:42 +0300)]
r600g: use a bitfield to track dirty atoms
r600 currently has 73 atoms and looping through their dirty flags has
become costly because checking each flag requires a pointer
dereference before the read. To avoid having to do that add additional
bitfield which can be checked really quickly thanks to tzcnt instruction.
id field was added to struct r600_atom but that doesn't affect memory
usage for both 32 and 64 bit CPUs because it was stuffed into padding.
The performance improvement is ~2% for benchmarks that can have FPS in
the thousands but is hardly measurable in "real" programs.
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Grazvydas Ignotas [Sun, 9 Aug 2015 21:42:34 +0000 (00:42 +0300)]
r600g: don't mark unused atom dirty
On evergreen config_state is not used, so don't mark it dirty.
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Grazvydas Ignotas [Sun, 9 Aug 2015 21:42:33 +0000 (00:42 +0300)]
r600g: use a helper to add an initialized atom
Instead of writing to rctx->atoms directly use a helper to take
advantage of assert checks.
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Grazvydas Ignotas [Sun, 9 Aug 2015 21:42:32 +0000 (00:42 +0300)]
gallium/radeon: use helper functions to mark atoms dirty
This is analogous to r300_mark_atom_dirty() used by r300, and will
be used by later patches. For common radeon code, appropriate helper
is called through a function pointer.
No functional changes.
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Francisco Jerez [Fri, 1 May 2015 14:00:02 +0000 (17:00 +0300)]
docs: Mark ARB_shader_image_load_store as done on i965.
Francisco Jerez [Thu, 7 May 2015 15:56:01 +0000 (18:56 +0300)]
i965: Expose ARB_shader_image_load_store.
Reviewed-by: Paul Berry <stereotype441@gmail.com>
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Francisco Jerez [Mon, 27 Jul 2015 11:33:06 +0000 (14:33 +0300)]
i965/fs: Clamp image array indices to the array bounds on IVB.
This fixes the spec@arb_shader_image_load_store@invalid index bounds
piglit tests on IVB, which were causing a GPU hang and then a crash
due to the invalid binding table index result of the array index
calculation. Other generations seem to behave sensibly when an
invalid surface is provided so it doesn't look like we need to care.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Francisco Jerez [Mon, 27 Jul 2015 13:26:52 +0000 (16:26 +0300)]
i965/fs: Translate image load, store and atomic NIR intrinsics.
v2: Move array coordinate workaround into the surface builder.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Francisco Jerez [Sun, 28 Jun 2015 18:16:31 +0000 (21:16 +0300)]
i965/fs: Handle image uniforms in NIR programs.
v2: Move the image_params array back to brw_stage_prog_data.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Francisco Jerez [Tue, 5 May 2015 18:07:15 +0000 (21:07 +0300)]
i965: Implement logic to set up and upload an image uniform.
v2: Move the image_params array back to brw_stage_prog_data.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Francisco Jerez [Tue, 5 May 2015 18:05:45 +0000 (21:05 +0300)]
i965: Teach type_size() about the size of an image uniform.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Francisco Jerez [Thu, 30 Jul 2015 12:46:40 +0000 (15:46 +0300)]
i965/fs: Implement image load, store and atomic.
v2: Drop VEC4 suport.
v3: Rebase.
v4: Move array coordinate workaround into the surface builder.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Francisco Jerez [Thu, 30 Jul 2015 12:51:58 +0000 (15:51 +0300)]
i965/fs: Import image format conversion primitives.
Define bitfield packing, unpacking and type conversion operations in
terms of which the image format conversion code will be implemented.
These don't directly know about image formats: The packing and
unpacking functions take a 4-tuple of bit shifts and a 4-tuple of bit
widths as arguments, determining the bitfield position of each
component. Most of the remaining functions perform integer, fixed
point normalized, and floating point type conversions, mapping between
a target type with per-component bit widths given by a parameter and a
matching native representation of the same type.
v2: Drop VEC4 suport.
v3: Rebase.
v4: Fix clamping of negative floats in the unsigned case of
emit_convert_to_scaled().
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Francisco Jerez [Wed, 22 Apr 2015 13:45:28 +0000 (16:45 +0300)]
i965/fs: Import image format metadata queries.
Define some utility functions to query the bitfield layout of a given
image format and whether it satisfies a number of more or less
hardware-specific properties.
v2: Drop VEC4 suport.
v3: Add SKL support.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Francisco Jerez [Thu, 23 Jul 2015 16:32:08 +0000 (19:32 +0300)]
i965/fs: Import code to transform image coordinates into surface coordinates.
Accounting for the padding required for 1D arrays in certain cases.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Francisco Jerez [Wed, 22 Apr 2015 13:44:18 +0000 (16:44 +0300)]
i965/fs: Import image memory offset calculation code.
Define a function to calculate the memory address of the image
location given by a vector of coordinates. This is required in cases
where we need to fall back to untyped surface access, which take a raw
memory offset and know nothing about surface coordinates, type
conversion or memory tiling and swizzling. They are still useful
because typed surface reads don't support any 64 or 128-bit formats on
IVB, and they don't support any 128-bit formats on HSW and BDW.
The tiling algorithm is implemented based on a number of parameters
which are passed in as uniforms and determine whether the surface
layout is X-tiled, Y-tiled or untiled. This allows binding surfaces
of different tiling layouts to the pipeline without recompiling the
program.
v2: Drop VEC4 suport.
v3: Rebase.
v4: Add plenty of comments (Jason).
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Francisco Jerez [Wed, 22 Apr 2015 13:43:51 +0000 (16:43 +0300)]
i965/fs: Import image access validity checks.
These utility functions check whether an image access is valid.
According to the spec an invalid image access should have no effect on
the image and yield well-defined results. Typically the hardware
implements correct bounds and surface checking by itself, but in some
cases (typed atomics on IVB and untyped messages elsewhere) we need to
implement it in software to work around lacking hardware support.
v2: Drop VEC4 suport.
v3: Rebase.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Francisco Jerez [Sat, 23 Nov 2013 00:00:33 +0000 (16:00 -0800)]
i965: Define implementation constants for ARB_shader_image_load_store.
Reviewed-by: Paul Berry <stereotype441@gmail.com>
v2: Drop VS support pre-Gen8, drop GS support.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Francisco Jerez [Mon, 9 Feb 2015 19:04:53 +0000 (21:04 +0200)]
i965/gen7-8: Set up early depth/stencil control appropriately for image load/store.
v2: Store early fragment test mode in brw_wm_prog_data instead of
getting it from core mesa data structures (Ken).
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Francisco Jerez [Mon, 13 Jul 2015 11:21:07 +0000 (14:21 +0300)]
i965/gen7-8: Poke the 3DSTATE UAV access enable bits.
v2: Set the PS UAV-only bit on HSW (Ken).
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Francisco Jerez [Tue, 3 Feb 2015 15:14:10 +0000 (17:14 +0200)]
i965/gen7: Enable fragment shader dispatch if the program has image uniforms.
Shaders with image uniforms may have side effects. Make sure that
fragment shader threads are dispatched if the shader has any image
uniforms.
v2: Use brw_stage_prog_data::nr_image_params to find out if the shader
has image uniforms instead of checking core mesa data structures
(Ken).
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Francisco Jerez [Mon, 20 Jul 2015 14:13:17 +0000 (17:13 +0300)]
i965: Hook up image state upload.
v2: Add CS support. Move the image_params array back to
brw_stage_prog_data.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Francisco Jerez [Mon, 13 Jul 2015 14:19:29 +0000 (17:19 +0300)]
i965: Reserve enough parameter entries for all image uniforms used in the program.
v2: Add CS support.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Francisco Jerez [Wed, 21 Jan 2015 15:34:49 +0000 (17:34 +0200)]
i965: Define and initialize image parameter structure.
This will be used to pass image meta-data to the shader when we cannot
use typed surface reads and writes. All entries except surface_idx
and size are otherwise unused and will get eliminated by the uniform
packing pass. size will be used for bounds checking with some image
formats and will be useful for ARB_shader_image_size too. surface_idx
is always used.
v2: Add CS support. Move the image_params array back to
brw_stage_prog_data.
v3: Improve documentation.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Francisco Jerez [Sat, 2 May 2015 13:58:24 +0000 (16:58 +0300)]
i965: Implement surface state set-up for shader images.
v2: Add SKL support.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Francisco Jerez [Tue, 12 May 2015 12:56:54 +0000 (15:56 +0300)]
i965: Fix brw_memory_barrier() for SKL.
This works as-is on SKL, only the assertion needs to be relaxed.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Francisco Jerez [Tue, 12 May 2015 13:10:07 +0000 (16:10 +0300)]
i965: Add SKL support to brw_miptree_get_horizontal_slice_pitch().
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Timothy Arceri [Sun, 9 Aug 2015 04:44:30 +0000 (14:44 +1000)]
glsl: Add missing spec quote about atomic counter in structs
Reviewed-by: Thomas Helland <thomashelland90@gmail.com>
Alex Deucher [Mon, 10 Aug 2015 19:35:21 +0000 (15:35 -0400)]
radeonsi: add new OLAND pci id
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: mesa-stable@lists.freedesktop.org
Ilia Mirkin [Mon, 10 Aug 2015 21:41:36 +0000 (17:41 -0400)]
nouveau: no need to do tnl wakeup, state updates are always hooked up
A TNL state update now requires a DrawBuffer to be set, which it isn't
early on in context creation. Since we init swtnl from context init,
this caused crashes.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91570
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Cc: "10.6" <mesa-stable@lists.freedesktop.org>
Jason Ekstrand [Mon, 10 Aug 2015 18:52:50 +0000 (11:52 -0700)]
i965/fs: Make resolve_source_modifiers consistent with the vec4 version
Reviewed-by: Matt Turner <mattst88@gmail.com>
Jason Ekstrand [Mon, 10 Aug 2015 18:48:14 +0000 (11:48 -0700)]
i965/vec4_visitor: Make some function arguments const references
Reviewed-by: Matt Turner <mattst88@gmail.com>
Jason Ekstrand [Fri, 31 Jul 2015 15:36:35 +0000 (08:36 -0700)]
i965/fs: Don't do redundant RA setup on IVB+
Acked-by: Matt Turner <mattst88@gmail.com>
Jason Ekstrand [Fri, 31 Jul 2015 15:35:57 +0000 (08:35 -0700)]
i965/fs: Use dispatch_width instead of reg_width in alloc_reg_sets
reg_width is kind of an outdated concept.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Jason Ekstrand [Fri, 31 Jul 2015 03:53:04 +0000 (20:53 -0700)]
ra: Delete the conflict lists in ra_set_finalize
They are never used after the set is finalized so there's no reason to keep
them around.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Jason Ekstrand [Fri, 31 Jul 2015 03:49:22 +0000 (20:49 -0700)]
ra: Refactor ra_set_finalize
All this commit does is change an early return to an if with an else
clause.
Reviewed-by: Matt Turner <mattst88@gmail.com>