r600g: add multi ring support with dma as first second ring v4
We keep track of ring emission order in a stack, whenever we need to
flush we empty the stack in a fifo order. There is few helpers function
for bo mapping and other ring activities that will make sure that
the ring stack is properly flush and submitted.
v2: fix st flush path, and other flush path to properly flush all
rings if necessary
v3: - improve name of ring helpers
- make sure that each time a cs is gona be written it endup at
top of the stack to avoid any issue such as :
STACK[0] = dma (withbo A,B)
STACK[1] = gfx (withbo C,D)
Now if code try to emit a dma command relative to bo C or D
it will start writting cmd stream into the cs and once it
reach the point where it adds relocation it will flush.
At that point the cs will have cmd that don't have proper
relocation into the relocation buffer and kernel will just
refuse to run.
v4: - Drop the stack idea as it turn out there is no way to use it
or benefit from it. Any time the driver start command on other
ring, it always need to flush the previous ring. So make code
simpler by not using a stack.
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Add ring support, you can create a cs for each ring. DMA ring is
bit special regarding relocation as you must emit as much relocation
as there is use of the buffer.
v2: - Improved comment on relocation changes
- Use a single thread to queue cs submittion this simplify driver
code while not impacting performances. Rational for this is that
you have to wait for all previous submission to have completed
so there was never a case while we could have 2 different thread
submitting a command stream at the same time. This code just
consolidate submission into one single thread per winsys.
v3: - Do not use semaphore for empty queue signaling, instead use
cond var. This is because it's tricky to maintain an even number
of call to semaphore wait and semaphore signal (the number of
cs in the stack would for instance make that number vary).
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
gallivm,draw,llvmpipe: mass rename of unit->texture_unit/sampler_unit
Make it obvious what "unit" this is (no change in functionality).
draw still uses "unit" in places where it changes the shader by adding
texture sampling itself - it seems like this can't work with shaders
using dx10-style sample opcodes (can't mix gl-style and dx10-style
sample instructions in a shader).
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
Split the sampler interface to use separate sampler and texture (sampler_view)
state. This is needed to support dx10-style sampling instructions.
This is not quite complete since both draw/llvmpipe don't really track
textures/samplers independently yet, as well as the gallivm code not quite
using the right sampler or texture index respectively (but it should work
for the sampling codes used by opengl).
We are however losing some optimizations in the process, apply_max_lod will
no longer work, and we potentially could end up with more (unnecessary)
recompiles (if switching textures with/without mipmaps only so it shouldn't
be too bad).
v2: don't use different callback structs for sampler/sampler view functions
(which just complicates things), fix up sampling code to actually use the
right texture or sampler index, and similar for llvmpipe/draw actually
distinguish between samplers and sampler views.
v3: fix more of PIPE_MAX_SAMPLER / PIPE_MAX_SHADER_SAMPLER_VIEWS mismatches
(both in draw and llvmpipe), based on feedback from José get rid of unneeded
static sampler derived state.(which also fixes the only 2 piglit regressions
due to a forgotten assignment), fix comments based on Brian's feedback.
v4: remove some accidental unrelated whitespace changes
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
npix_x/y/z is wrong with NPOT textures, since it's always aligned to POT
if the level is non-zero, so we can't use that.
This fixes piglit/spec/EXT_texture_shared_exponent/fbo-generatemipmap-formats.
svga: use pipe_sampler_view_release() in svga_cleanup_tss_binding()
Fixes a crash when the Redway3D Turbine demo exits. We've made this
change in other places in the past. The root issue is texture objects
are being shared by multiple contexts and sampler views get shared too.
Sampler views have a context pointer and if that context gets deleted
we may try to reference that context when finally deleting the sampler
view.
pipe_sampler_view_release() avoids this problem because it takes
an explicit context.
Reviewed-by: Zack Rusin <zackr@vmware.com>
Check the return value of calls to u_upload_alloc() and
u_upload_data() and return early if needed.
Since we don't have a way to propagate errors all the way up to
Mesa through pipe_context::draw_vbo(), call debug_warn_once() so
the user might have some clue about OOM errors.
Note: This is a candidate for the 9.0 branch.
st/mesa: do proper error checking for u_upload_alloc() calls
We weren't properly checking the return value of these calls (and
calls to u_upload_data()) to detect OOM errors.
Note: This is a candidate for the 9.0 branch.
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
util: add some defensive coding in u_upload_alloc()
Some callers of this function were checking the 'ptr' result to see if
the function failed. But the correct way is to check the regular
return value for PIPE_ERROR_x. Now we initialize all the returned
values at the top of the function in case we do hit an error (like OOM).
Callers are more likely to detect OOM conditions now. But there
are some callers which don't do any error checking...
Note: This is a candidate for the 9.0 branch.
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
glsl: Add support for lowering 4x8 pack/unpack operations
Lower them to arithmetic and bit manipulation expressions.
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
Reviewed-by: Paul Berry <stereotype441@gmail.com>
That is, evaluate constant expressions for the following functions:
packSnorm4x8, unpackSnorm4x8
packUnorm4x8, unpackUnorm4x8
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
Reviewed-by: Paul Berry <stereotype441@gmail.com>
glsl: Extend ir_expression_operation for ARB_shading_language_packing
For each function {pack,unpack}{Snorm,Unorm}4x8, add a corresponding
opcode to enum ir_expression_operation. Validate the new opcodes in
ir_validate.cpp.
Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
Reviewed-by: Paul Berry <stereotype441@gmail.com>
r600g: Don't build llvm_wrapper.cpp when we aren't using LLVM
We were using the NEED_RADEON_GALLIUM conditional to decide whether or not
to build llvm_wrapper.cpp, which is required for using the LLVM backend.
llvm_wrapper.cpp needs to be linked against the LLVM IPO libary
and this library is only added to LLVM_LIBS if either opencl or the
r600-llvm-compiler is enabled.
The NEED_RADEON_GALLIUM conditional is set to true when enabling the
radeonsi driver, so if the radeonsi and r600 drivers are enabled without
also enabling opencl or r600-llvm-compiler, llvm_wrapper.cpp will be
built, but the IPO library won't be added to LLVM_LIBS. This was
causing unresolved symbol errors when buiding with this configuration.
https://bugs.freedesktop.org/show_bug.cgi?id=59831
Tested-by: Alex Deucher <alexander.deucher@amd.com>
Dereffing all the values in the two callers was just pointless, and
the function isn't inlined so there was actual code impact.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
i965: reuse _mesa_sizeof_type for index buffer types.
The core Mesa code has just one more case than this (GL_BITMAP), so I
don't see any cause to special-case it.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
i965: Drop debug check for knowing the size of a type.
This was added in b93684f5f311f89c965960ab42bfea71a397b180, but there's
no need for it -- get_size has to succeed, and it has an assert for us
in debug builds.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
i965: Stop worrying about alignment of vertex data.
For our current types, the required alignment is actually just 1 byte.
When we get doubles, we have to worry (those have to be aligned to the
natural size), but we don't have doubles yet and they'll just be a
special case.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
v2: A previous patch contained a spurious hunk that removed an
assignment to ir_variable::uniform_block. That hunk was moved to this
patch. Suggested by Carl Worth.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jerome Glisse