Kenneth Graunke [Mon, 11 Aug 2014 05:49:55 +0000 (22:49 -0700)]
i965/vp: Use the sampler for pull constant loads on Gen7/7.5.
This improves performance in Trine 2 at 1280x720 (windowed) on "Very
High" settings by 30% (in the interactive menu) to 45% (in the forest
by the giant frog) on Haswell GT3e.
It also now generates the same assembly on Gen7 as it does on Gen8,
which always used the sampler for both types.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Mon, 11 Aug 2014 03:41:42 +0000 (20:41 -0700)]
i965/vec4: Drop gen <= 7 assertion in pull constant load handling.
I don't see any reason for this to exist.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Sun, 10 Aug 2014 14:10:55 +0000 (07:10 -0700)]
i965/eu: Set src0 file to IMM on Gen8+ flow control instructions.
According to the documentation, we need to set the source 0 register
type to IMM for flow control instructinos that have both JIP and UIP.
Out of paranoia, just make all flow control instructions use IMM;
there's no benefit to using ARF anyway, and it could trouble that's
difficult to diagnose.
See commit
9584959123b0453cf5313722357e3abb9f736aa7, which did the
analogous change in the gen8_generator code.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Sun, 10 Aug 2014 14:06:36 +0000 (07:06 -0700)]
i965/eu: Refactor brw_WHILE to share a bit more code on Gen6+.
We're going to add a Gen8+ case shortly, which would need to duplicate
this code again. Instead, share it.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Sat, 28 Jun 2014 23:08:39 +0000 (16:08 -0700)]
i965/eu: Emulate F32TO16 and F16TO32 on Broadwell.
When we combine the Gen4-7 and Gen8+ generators, we'll need to handle
half float packing/unpacking functions somehow. The Gen8+ generator
code today just emulates the behavior of the Gen7 F32TO16/F16TO32
instructions, including the align16 mode bugs.
Rather than messing with fs_generator/vec4_generator, I decided to just
emulate the instructions at the brw_eu_emit.c layer.
v2: Change gen >= 7 asserts to gen == 7 (suggested by Chris Forbes).
Fix regressions on Haswell in VS tests due to type assertions.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Mon, 11 Aug 2014 22:53:54 +0000 (15:53 -0700)]
i965/vec4: Port Gen8 SET_VERTEX_COUNT handling to vec4_generator.
Broadwell requires the number of vertices written by the geometry shader
to be specified in a separate register, as part of the terminating
message's payload.
This also means GS_OPCODE_THREAD_END needs to increment mlen.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Mon, 11 Aug 2014 15:13:05 +0000 (08:13 -0700)]
i965/vec4: Switch to MOV, not OR, for GS_OPCODE_THREAD_END on Gen8.
Either should work.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Mon, 11 Aug 2014 03:06:44 +0000 (20:06 -0700)]
i965/vec4: Use MOV, not OR, to set URB write channel mask bits.
g0.5 has nothing of value to contribute to m0.5. In both the VS and GS
payload, g0.5 contains the scratch space pointer - which is definitely
not of any use. The GS payload also contains FFTID, but the URB write
message header doesn't want FFTID.
The only reason I used OR was because Eric originally requested it.
On Broadwell, I used MOV, and that's worked out fine.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Kenneth Graunke [Sun, 10 Aug 2014 23:15:51 +0000 (16:15 -0700)]
i965/fs: Don't set flag_subreg_nr = 1 on predicated FB write setup.
On Haswell, we implement "discard" via predicated SEND messages, using
f0.1 instead of f0.0. To accomplish this, we set inst->flag_subreg to 1
on the FS_OPCODE_FB_WRITE.
Most instructions using fs_inst::flag_subreg expand to a single assembly
instruction. However, FS_OPCODE_FB_WRITE can generate several MOVs for
setting up header information. We don't want to set flag_subreg on
those, so override the default state back to 0.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Kenneth Graunke [Mon, 11 Aug 2014 15:41:36 +0000 (08:41 -0700)]
i965/vec4: Respect ir->force_writemask_all in Gen8 code generation.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Cc: "10.2" <mesa-stable@lists.freedesktop.org>
Kenneth Graunke [Mon, 11 Aug 2014 15:15:57 +0000 (08:15 -0700)]
i965/vec4: Set NoMask for GS_OPCODE_SET_VERTEX_COUNT on Gen8+.
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Cc: "10.2" <mesa-stable@lists.freedesktop.org>
Jason Ekstrand [Tue, 12 Aug 2014 18:12:47 +0000 (11:12 -0700)]
gallium/r300: Fix a link error in the tests
The link error occurs because the static libraries are linked in the wrong
order. This fixes it.
Signed-off-by: Jason Ekstrand <jason.ekstrand@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=82483
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Matt Turner [Mon, 11 Aug 2014 18:12:43 +0000 (11:12 -0700)]
i965: Return NONE from brw_swap_cmod on unknown input.
Comparing ~0u with a packed enum (i.e., 1 byte) always evaluates to
false. Shouldn't gcc warn about this?
Reported-by: Connor Abbott <cwabbott0@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Neil Roberts [Wed, 23 Jul 2014 10:25:31 +0000 (11:25 +0100)]
docs: Update release notes and GL3.txt for GL_ARB_texture_compression_bptc
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Neil Roberts [Thu, 31 Jul 2014 13:07:50 +0000 (14:07 +0100)]
mesa/meta: Support decompressing floating-point formats
Previously the Meta implementation of glGetTexImage would fall back to
_mesa_get_teximage if the texturing is not using an unsigned normalised
format. However in order to support the half-float formats of BPTC textures we
can make it render to a floating-point renderbuffer instead. This patch makes
decompression_state have two FBOs, one for the GL_RGBA format and one for
GL_RGBA32F. If a floating-point texture is encountered it will try setting up
a floating-point FBO. It will now also check the status of the FBO and fall
back to _mesa_get_teximage if the FBO is not complete.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Neil Roberts [Thu, 17 Jul 2014 13:45:01 +0000 (14:45 +0100)]
swrast: Enable GL_ARB_texture_compression_bptc
Enables BPTC texture compression on the software rasterizer.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Neil Roberts [Thu, 17 Jul 2014 13:38:20 +0000 (14:38 +0100)]
i965: Enable the GL_ARB_texture_compression_bptc extension
Enables the BPTC extension on Gen>=7 and adds the necessary format mappings to
get the right surface type value.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Neil Roberts [Fri, 25 Jul 2014 16:38:22 +0000 (17:38 +0100)]
mesa/main: Modify generate_mipmap_compressed to cope with float textures
Once we add BPTC texture support we will need to generate mipmaps for
compressed floating point textures too. Most of the code seems to already be
there but it just needs a few extra lines to get it to use GL_FLOAT instead of
GL_UNSIGNED_BYTE as the type for the temporary buffers.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Neil Roberts [Thu, 17 Jul 2014 13:33:10 +0000 (14:33 +0100)]
mesa: Add texstore functions for BPTC-compressed textures
This adds compressors for all four of the BPTC compressed-texture formats. The
compressor is written from scratch and takes a very simple approach. It always
uses a single mode of the BPTC format (4 for unorm and 3 for half-floats) and
picks the two endpoints by dividing the texels into those which have more or
less than the average luminance of the block and then calculating an average
color of the texels within each division.
It's probably not really sensible to try to use BPTC compression at runtime
because for example with the Nvidia offline compression tool it can take in
the order of an hour to compress a full-screen image. With that in mind I
don't think it's worth having a proper compressor in Mesa and this approach
gives reasonable results for a usage that is basically a corner case.
v2: Always use the custom compressor, even for the unorm formats. Fix the
quantization step for the half-float format compressor. Fixed a typo which
was breaking the right-hand edge of half-float textures with a width that
isn't a multiple of four.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Neil Roberts [Thu, 17 Jul 2014 13:30:29 +0000 (14:30 +0100)]
mesa: Add texel fetch functions for BPTC-compressed textures
Adds functions to fetch from any of the four BPTC-compressed formats.
v2: Set the alpha component to 1.0 when fetching from the half-float formats
instead of leaving it uninitialised. Don't linearize the alpha component
when fetching from sRGB.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Neil Roberts [Thu, 17 Jul 2014 13:18:27 +0000 (14:18 +0100)]
mesa: Add the format enums for BPTC-compressed images
This adds the following four Mesa image format enums which correspond to the
four BPTC compressed texture formats:
MESA_FORMAT_BPTC_RGBA_UNORM
MESA_FORMAT_BPTC_SRGB_ALPHA_UNORM
MESA_FORMAT_BPTC_RGB_SIGNED_FLOAT
MESA_FORMAT_BPTC_RGB_UNSIGNED_FLOAT
It also updates the format information functions to handle these and the
corresponding GL enums.
v2: Also modify _mesa_get_format_color_encoding, _mesa_get_srgb_format_linear
and _mesa_get_uncompressed_format
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Neil Roberts [Wed, 6 Aug 2014 15:52:14 +0000 (16:52 +0100)]
mesa/format_info: Add support for the BPTC layout
Adds the ‘bptc’ layout to get_channel_bits. The channel bits for BPTC depend
on the mode but as it only has to be an approximation this sets it to 8 for
the two UNORM formats and 16 for the two half-float formats. These represent
the minimum number of bits of variation that can be generated by the
interpolation of the two formats.
This doesn't quite match what we do for S3TC which only returns 4 even though
it can similarly generate 8 bits from the interpolation. However it does match
what we return for ETC2. For reference, NVidia seems to return 8 bits for the
UNORM formats and 32 bits for the half-float formats.
v2: Change the number of bits to 8/8/8/8 for the UNORM formats and 16/16/16
for the half-float formats.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Neil Roberts [Wed, 6 Aug 2014 15:52:14 +0000 (16:52 +0100)]
mesa/format_info: Add support for compressed floating-point formats
If the name of a compressed texture format has ‘FLOAT’ in it it will now set
the data type of the format to GL_FLOAT. This will be needed for the BPTC
half-float formats.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Neil Roberts [Fri, 25 Jul 2014 10:31:07 +0000 (11:31 +0100)]
mesa: Fix the base format for GL_COMPRESSED_RGB_BPTC_*_FLOAT_ARB
The signed and unsigned half-float BPTC-compressed formats were being reported
as having a base format of GL_RGBA but they don't store an alpha channel so it
should be GL_RGB.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Neil Roberts [Thu, 17 Jul 2014 13:21:26 +0000 (14:21 +0100)]
mesa: Add the GL_ARB_texture_compression_bptc extension
This adds a boolean in the gl_extensions struct for
GL_ARB_texture_compression_bptc as well as an entry in extension_table.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Andreas Boll [Mon, 4 Aug 2014 10:48:50 +0000 (12:48 +0200)]
winsys/radeon: fix nop packet padding for hawaii
The initial firmware for hawaii does not support type3 nop packet.
Detect the new hawaii firmware with query RADEON_INFO_ACCEL_WORKING2.
If the returned value is 3, then the new firmware is used.
This patch uses type2 for the old firmware and type3 for the new firmware.
It fixes the cases when the old firmware is used and the user wants to
manually enable acceleration.
The two possible scenarios are:
- the kernel has no support for the new firmware.
- the kernel has support for the new firmware but only the old firmware
is available.
Additionaly this patch disables GPU acceleration on hawaii if the kernel
returns a value < 2. In this case the kernel hasn't the required fixes
for proper acceleration.
v2:
- Fix indentation
- Use private struct radeon_drm_winsys instead of public struct radeon_info
- Rename r600_accel_working2 to accel_working2
v3:
- Use type2 nop packet for returned value < 3
v4:
- Fail to initialize winsys for returned value < 2
Cc: mesa-stable@lists.freedesktop.org
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Jérôme Glisse <jglisse@redhat.com>
Cc: Marek Olšák <marek.olsak@amd.com>
Cc: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Andreas Boll <andreas.boll.dev@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Brian Paul [Tue, 12 Aug 2014 14:01:41 +0000 (08:01 -0600)]
mesa: regenerate gl_mangle.h
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Brian Paul [Tue, 12 Aug 2014 13:31:46 +0000 (07:31 -0600)]
mesa: update wglext.h to version
20140810
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Brian Paul [Tue, 12 Aug 2014 13:31:25 +0000 (07:31 -0600)]
mesa: update glxext.h to version
20140810
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Brian Paul [Tue, 12 Aug 2014 13:30:52 +0000 (07:30 -0600)]
mesa: update glext.h to version
20140810
This brings in the new OpenGL 4.5 features.
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Charmaine Lee [Tue, 12 Aug 2014 13:37:12 +0000 (07:37 -0600)]
svga: Add a limit to the maximum surface size
This patch adds a limit to the maximum surface size which is
based on the maximum size of a single mob. If this value is not
available, the maximum surface size is by default set to 128 MB.
Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
José Fonseca [Tue, 12 Aug 2014 13:24:38 +0000 (14:24 +0100)]
mesa/st: Move declaration to top of block.
To fix MSVC build failure.
Trivial.
Ilia Mirkin [Wed, 6 Aug 2014 04:43:29 +0000 (00:43 -0400)]
mesa/st: add support for dynamic sampler offsets
Replace the plain sampler index with a register reference to a sampler.
We also need to keep track of the sampler array size when there is a
relative reference so that we can mark the whole array used.
To facilitate implementation, we add a separate ADDR register that
exclusively handles the sampler relative address. Other approaches would
be more invasive.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Acked-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Christian König [Mon, 11 Aug 2014 14:40:43 +0000 (16:40 +0200)]
radeon/uvd: fix gpu_address for video surfaces
We need to get the new gpu_address as well when
reallocating the cs buffer.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=82428
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Tested-by: Kai Wasserbäch <kai@dev.carbon-project.org>
Chris Forbes [Sun, 3 Aug 2014 07:55:55 +0000 (19:55 +1200)]
mesa: Add a new function for getting the nonconst sampler array index
If the array index is not a constant expression, the existing support
will assume a zero offset (giving us the sampler index of the base of
the array).
For dynamically uniform indexing of sampler arrays, we need both that
and the indexing expression.
Signed-off-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Chris Forbes [Sun, 3 Aug 2014 05:57:05 +0000 (17:57 +1200)]
glsl: Allow dynamically uniform sampler array indexing with 4.0/gs5
V2: Expand comment to explain what dynamically uniform expressions are
about.
Signed-off-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Ilia Mirkin [Thu, 7 Aug 2014 03:45:05 +0000 (23:45 -0400)]
nvc0/ir: describe the tex arguments for fermi/kepler
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Ilia Mirkin [Wed, 9 Jul 2014 04:41:11 +0000 (00:41 -0400)]
nvc0/ir: add kepler+ support for indirect texture references
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Ilia Mirkin [Wed, 6 Aug 2014 05:22:49 +0000 (01:22 -0400)]
nvc0/ir: add base tex offset for fermi indirect tex case
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Kenneth Graunke [Mon, 11 Aug 2014 22:05:54 +0000 (15:05 -0700)]
i965: Revert part of
f5cc3fdcf1680b116612fac7c39f1bd79f5e555e.
Fixes non-termination in various Piglit tests.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Eric Anholt [Sat, 9 Aug 2014 18:01:53 +0000 (11:01 -0700)]
vc4: Flip which primitives are considered front-facing.
This mostly fixes glxgears rendering.
Eric Anholt [Sat, 9 Aug 2014 18:00:51 +0000 (11:00 -0700)]
vc4: Don't forget to set the depth clear value in the packet.
This gets glxgears partially rendering again.
Eric Anholt [Tue, 5 Aug 2014 21:24:29 +0000 (14:24 -0700)]
vc4: Add support for gl_FragCoord.
This isn't passing all tests (glsl-fs-fragcoord-zw-ortho, for example),
but it does get a bunch more tests passing.
v2: Rebase on helpers change.
Eric Anholt [Tue, 5 Aug 2014 21:23:40 +0000 (14:23 -0700)]
vc4: Refactor shader input setup again.
This makes some space for handling special inputs like fragcoords.
Eric Anholt [Tue, 5 Aug 2014 18:00:51 +0000 (11:00 -0700)]
vc4: Clean up the tile alloc buffer size.
This prevents some simulator assertion failures, but it does mean (since
I've dropped the "* 16" padding) that on real hardware you need a kernel
that does overflow memory management (currently, "drm/vc4: Add support for
binner overflow memory allocation." in my kernel tree).
Eric Anholt [Tue, 5 Aug 2014 18:00:08 +0000 (11:00 -0700)]
vc4: Clarify some values implicitly chosen for binning config.
These #defines are 0, but it should help make math above make more sense.
Eric Anholt [Tue, 5 Aug 2014 17:54:56 +0000 (10:54 -0700)]
vc4: Improve simulator memory allocation.
This should reduce a bunch of spurious failures in sim.
Eric Anholt [Tue, 5 Aug 2014 01:30:33 +0000 (18:30 -0700)]
vc4: Handle stride==0 in VBO validation
Eric Anholt [Mon, 4 Aug 2014 23:38:07 +0000 (16:38 -0700)]
vc4: Stash some debug code for looking at what BOs are at what hindex.
When you're debugging validation, it's nice to know what the BOs are for.
Eric Anholt [Mon, 4 Aug 2014 20:01:29 +0000 (13:01 -0700)]
vc4: Use GEM under simulation even for non-winsys BOs.
In addition to reducing sim-specific code, it also avoids our local handle
allocation conflicting with the host GEM's handle numbering, which was
causing vc4_gem_hindex() to not distinguish between winsys BOs and the
same-numbered non-winsys bo.
Eric Anholt [Mon, 4 Aug 2014 20:00:56 +0000 (13:00 -0700)]
vc4: Don't forget to unmap the GEM BO when freeing.
Otherwise it'll stick around forever.
Eric Anholt [Sun, 3 Aug 2014 04:28:34 +0000 (21:28 -0700)]
vc4: Add validation of raster-format textures.
... and reject everything else, for now.
v2: Rebase on v2 of the rendering config validation change.
Eric Anholt [Sun, 3 Aug 2014 04:23:20 +0000 (21:23 -0700)]
vc4: Drop VC4_PACKET_PRIMITIVE_LIST_FORMAT.
It's not relevant to our command streams any more.
v2: Fix indentation and a typo in the comment.
Eric Anholt [Sun, 3 Aug 2014 04:06:50 +0000 (21:06 -0700)]
vc4: Add validation that vertex indices don't overflow VBO bounds.
Eric Anholt [Sun, 3 Aug 2014 03:44:39 +0000 (20:44 -0700)]
vc4: Fix the shader record size for extended strides.
It turns out they aren't packed when attributes are missing, according to
both docs and simulation.
Eric Anholt [Sun, 3 Aug 2014 03:44:39 +0000 (20:44 -0700)]
vc4: Fix the shader record size for extended strides.
It turns out they aren't packed when attributes are missing, according to
both docs and simulation.
v2: Drop unused variable.
Eric Anholt [Sun, 3 Aug 2014 03:30:18 +0000 (20:30 -0700)]
vc4: Add a bunch of validation of render mode configuration.
v2: Fix a build break after some previous rebase.
Eric Anholt [Sun, 3 Aug 2014 03:19:38 +0000 (20:19 -0700)]
vc4: Store the (currently always linear) tiling format in the resource.
Eric Anholt [Sat, 2 Aug 2014 00:11:38 +0000 (17:11 -0700)]
vc4: Add a bunch of validation of the binning mode config.
Eric Anholt [Sat, 2 Aug 2014 03:23:31 +0000 (20:23 -0700)]
vc4: Validate that the same BO doesn't get reused for different purposes.
We don't care if things like vertex data get smashed by render target
data, but we do need to make sure that shader code doesn't get rendered
to.
v2: Fix overflowing read of gl_relocs[] that incorrect flagged of some
VBOs as shader code.
Eric Anholt [Sat, 2 Aug 2014 00:31:40 +0000 (17:31 -0700)]
vc4: Use the packet #defines in the kernel validation code.
Eric Anholt [Sat, 2 Aug 2014 00:17:03 +0000 (17:17 -0700)]
vc4: Rename GEM_HANDLES to be in a namespace.
It's not a real VC4 hardware packet, but I've put in a comment to explain
it.
Eric Anholt [Sat, 2 Aug 2014 00:05:21 +0000 (17:05 -0700)]
vc4: Clean up TMU write validation.
The comment conflicted with the support in the code, so I moved the TMU
write validation to where the comment was, and dropped some dead arguments
from the functions while changing their signatures.
Eric Anholt [Sat, 2 Aug 2014 00:01:44 +0000 (17:01 -0700)]
vc4: Update a comment about shader validation
Eric Anholt [Fri, 1 Aug 2014 23:02:37 +0000 (16:02 -0700)]
vc4: Add proper translation from Zc to Zs for vertex output.
This fixes the remaining failure in depthfunc.
Eric Anholt [Fri, 1 Aug 2014 20:32:49 +0000 (13:32 -0700)]
vc4: Add support for depth clears and tests within a tile.
This doesn't load/store the Z contents across submits yet. It also
disables early Z, since it's going to require tracking of Z functions
across multiple state updates to track the early Z direction and whether
it can be used.
v2: Move the key setup to before the search for the key.
Eric Anholt [Fri, 1 Aug 2014 22:33:06 +0000 (15:33 -0700)]
vc4: Avoid flushing when mapping buffers that aren't in the batch.
This should prevent a bunch of unnecessary flushes for things like
updating immediate vertex data.
Eric Anholt [Thu, 31 Jul 2014 05:17:56 +0000 (22:17 -0700)]
vc4: Drop the flush at the end of the draw
Now we actally get multiple draw calls per submit.
Eric Anholt [Fri, 1 Aug 2014 18:24:29 +0000 (11:24 -0700)]
vc4: Align following shader recs to 16 bytes.
Otherwise, the low address bits will end up being interpreted as attribute
counts.
Eric Anholt [Thu, 31 Jul 2014 20:14:00 +0000 (13:14 -0700)]
vc4: Fix a potential src buffer overflow in shader rec validation.
Eric Anholt [Thu, 31 Jul 2014 19:19:29 +0000 (12:19 -0700)]
vc4: Keep a reference to BOs queued for rendering.
Otherwise, once we're not flushing at the end of every draw, we'll free
things like gallium resources, and free the backing GEM object, before
we've flushed the rendering using it to the kernel.
Eric Anholt [Thu, 31 Jul 2014 19:46:13 +0000 (12:46 -0700)]
vc4: Compute the proper end address of the relocated command lists.
render_cl_size/bin_cl_size includes relocations, while the hardware buffer
doesn't. If you don't emit a HALT packet, the command parser continues
until the end register's value. We can't allow executing unvalidated
buffer contents (and it's actually harmful in the render lists Mesa is
emitting, since VC4_PACKET_STORE_MS_TILE_BUFFER_AND_EOF doesn't trigger a
halt).
Eric Anholt [Thu, 31 Jul 2014 19:45:41 +0000 (12:45 -0700)]
vc4: Walk tiles horizontally, then vertically.
I was confused looking at my addresses in dumps because I was seeing the
tile branch offsets jumping all over.
Eric Anholt [Wed, 23 Jul 2014 18:21:04 +0000 (11:21 -0700)]
vc4: Track clears veresus uncleared draws, and the clear color.
This is a step toward queueing more than one draw per frame.
Fixes piglit attribute0 test, since we get a working clear color now.
Eric Anholt [Thu, 31 Jul 2014 18:23:22 +0000 (11:23 -0700)]
vc4: Move the rest of RCL setup to flush time.
We only want to set up render target config and clear colors once per
frame.
Eric Anholt [Thu, 31 Jul 2014 18:22:17 +0000 (11:22 -0700)]
vc4: Move render command list calls to vc4_flush()
Eric Anholt [Thu, 31 Jul 2014 18:19:41 +0000 (11:19 -0700)]
vc4: Move bin command list ending commands to vc4_flush()
Eric Anholt [Wed, 23 Jul 2014 03:16:10 +0000 (20:16 -0700)]
vc4: Rename fields in the kernel interface.
I decided I didn't like "len" compared to "size", and I keep typing
shader_rec instead of shader_record[s] elsewhere, so make it consistent.
Eric Anholt [Wed, 23 Jul 2014 03:10:01 +0000 (20:10 -0700)]
vc4: Fix things to validate more than one shader state in a submit.
Eric Anholt [Mon, 21 Jul 2014 18:27:35 +0000 (11:27 -0700)]
vc4: Rewrite the kernel ABI to support texture uniform relocation.
This required building a shader parser that would walk the program to find
where the texturing-related uniforms are in the uniforms stream.
Note that as of this commit, a new kernel is required for rendering on
actual VC4 hardware (currently that commit is named "drm/vc4: Introduce
shader validation and better command stream validation.", but is likely to
be squashed as part of an eventual merge of the kernel driver).
Eric Anholt [Mon, 21 Jul 2014 18:26:24 +0000 (11:26 -0700)]
vc4: Add docs for the drm interface
Eric Anholt [Fri, 18 Jul 2014 21:18:23 +0000 (14:18 -0700)]
vc4: Add load/store to the validator
Eric Anholt [Fri, 18 Jul 2014 20:06:01 +0000 (13:06 -0700)]
vc4: Switch simulator to using kernel validator
This ensures that when I'm using the simulator, I get a closer match to
what behavior on real hardware will be. It lets me rapidly iterate on the
kernel validation code (which otherwise has a several-minute turnaround
time), and helps catch buffer overflow bugs in the userspace driver
faster.
Eric Anholt [Fri, 18 Jul 2014 20:28:34 +0000 (13:28 -0700)]
vc4: Drop pointless shader state struct
Eric Anholt [Thu, 17 Jul 2014 04:39:05 +0000 (21:39 -0700)]
vc4: Add support for texture rectangles
v2: Rebase on helpers change.
Eric Anholt [Tue, 15 Jul 2014 19:29:32 +0000 (12:29 -0700)]
vc4: Add support for texturing (under simulation)
Only rgba8888 works, and only a single texture unit, and it's only under
simulation because I haven't built the kernel interface yet.
v2: Rebase on helpers.
v3: Fold in the don't-break-the-arm-build fix.
Eric Anholt [Mon, 11 Aug 2014 21:40:06 +0000 (14:40 -0700)]
vc4: Drop PIPE_SHADER_CAP_MAX_ADDRS
Fixes the build since
c10332bbb8889d733bdaa729ef23cbd90176b55d
Marek Olšák [Wed, 6 Aug 2014 21:58:10 +0000 (23:58 +0200)]
gallium: remove PIPE_SHADER_CAP_MAX_ADDRS
This limit is fixed in Mesa core and cannot be changed.
It only affects ARB_vertex_program and ARB_fragment_program.
The minimum value for ARB_vertex_program is 1 according to the spec.
The maximum value for ARB_vertex_program is limited to 1 by Mesa core.
The value should be zero for ARB_fragment_program, because it doesn't
support ARL.
Finally, drivers shouldn't mess with these values arbitrarily.
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Marek Olšák [Sun, 3 Aug 2014 03:36:23 +0000 (05:36 +0200)]
st/mesa: compute supported GL versions at DRIscreen creation
This computes all GL versions before any context is created.
It's a requirement for GLX_MESA_query_renderer.
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Marek Olšák [Sun, 3 Aug 2014 11:58:20 +0000 (13:58 +0200)]
gallium: pass st_config_options to query_versions
So move it from dri_context to dri_screen.
This will be needed for version computations.
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Marek Olšák [Sun, 3 Aug 2014 03:35:10 +0000 (05:35 +0200)]
mesa: return version 0 if the computed core profile version is too low
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Marek Olšák [Sun, 3 Aug 2014 03:27:10 +0000 (05:27 +0200)]
mesa: add _mesa_get_version, a ctx-independent variant of _mesa_compute_version
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Marek Olšák [Sun, 3 Aug 2014 03:17:08 +0000 (05:17 +0200)]
mesa: add a context-independent variant of _mesa_override_gl_version
v2: changed GLboolean -> bool
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Marek Olšák [Sun, 3 Aug 2014 02:51:31 +0000 (04:51 +0200)]
mesa: make _mesa_init_constants context-independent and public
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Marek Olšák [Sun, 3 Aug 2014 02:42:50 +0000 (04:42 +0200)]
mesa: make _mesa_init_extensions context-independent
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Marek Olšák [Sun, 3 Aug 2014 02:36:19 +0000 (04:36 +0200)]
st/mesa: make st_init_limits context-independent
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Marek Olšák [Sun, 3 Aug 2014 02:31:56 +0000 (04:31 +0200)]
mesa: move ShaderCompilerOptions into gl_constants
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Marek Olšák [Sun, 3 Aug 2014 02:20:31 +0000 (04:20 +0200)]
st/mesa: make st_init_extensions context-independent
Setting Const.MaxSamples needed a rework, so that it doesn't call
st_choose_format, which depends on st_context.
Other than that, there is no change in functionality.
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Marek Olšák [Sun, 3 Aug 2014 01:40:49 +0000 (03:40 +0200)]
mesa: make _mesa_override_glsl_version context-independent
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Marek Olšák [Sun, 3 Aug 2014 00:54:52 +0000 (02:54 +0200)]
gallium/stapi: move setting GL versions to the state tracker
All flags are set for st/mesa, so the state tracker doesn't have to check
them.
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>