| Age | Commit message (Collapse) | Author |
|---|
|
Trying to make understanding code easier with small refactoring and renaming.
|
|
Fixes progs/demos/cubemap
|
|
|
|
|
|
|
|
|
|
|
|
Split into component pieces, stitch together at runtime using function
pointers. Make it possible to utilize the existing fastpaths as image-level
filters for generic mip-filtering routines.
Remove special case for rectangle filtering, as it can now be handled by
the 2d path.
As most of the mesa demo texturing was already covered by fast paths, its
harder to find examples of speedups, but tunnel gets a boost as mip-nearest
filtering is now able to access the img_2d_linear_wrap_POT functions
for sampling within a mipmap level.
|
|
|
|
|
|
|
|
When adding a new bitmap to the cache we have to check if the Z value is
changing and flush first if it is.
This is a modified version of a patch from Justin Dou <justin.dou@intel.com>
|
|
We do flush for cmd buffer in case there isn't enough space left for whole
rendering operation. This protects dma regions from getting released in middle
of state emit.
Signed-off-by: Pauli Nieminen <suokkos@gmail.com>
|
|
|
|
This makes it easier to predict size of next rendering operation so we
can do early flush.
|
|
This fixes some state atom check functions from returing wrong emit size.
There is emit code cleanup so that emit function selection is done in init
time instead of runtime.
Signed-off-by: Pauli Nieminen <suokkos@gmail.com>
|
|
- emit prediction
- fixes to emit_sizes for data
- clean up of excesive use of radeonEnsureCmdBufferSpace
Signed-off-by: Pauli Nieminen <suokkos@gmail.com>
|
|
Signed-off-by: Pauli Nieminen <suokkos@gmail.com>
|
|
Signed-off-by: Pauli Nieminen <suokkos@gmail.com>
|
|
Calling EnsureCmdBufSpace is not required because rendering pipeline has to quarentee free space.
Signed-off-by: Pauli Nieminen <suokkos@gmail.com>
|
|
This patch makes render emit size prediction count the corect maximum emit size
for state.
Signed-off-by: Pauli Nieminen <suokkos@gmail.com>
|
|
Patch adds prediction functionthat tries to predict emit size to the smallest
possible values that is quarenteed to be higher than worst case scenario in
rendering pipeline.
State emit size prediction code is in place but fix for emit sizes is included
in next patch.
Signed-off-by: Pauli Nieminen <suokkos@gmail.com>
|
|
Signed-off-by: Pauli Nieminen <suokkos@gmail.com>
|
|
This fixes bug that xdemos/manywin would segfault if it was run with command
./manywin 2. Demo is tring to call glXSwapBuffers while another context was
bind using glXMakeCurrent.
Fix is simple makes drawable and readable persistent untill they change or
context is destroyed.
I found a logic error when same dri context is used for multiple drawables
which caused readable and drawable to fall out of sync in special case. Fix
is simple just updating drawables more often than in original patch.
Signed-off-by: Pauli Nieminen <suokkos@gmail.com>
|
|
The real difference is that the driver suffix is now appended. This
also fixes an annoying bug that EGL_DRIVER could not specify the path to
a driver because a suffix was always appended.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
|
|
libX11 is used to determine the screen number, which is in turned used
to determine the DRI driver. However, the sysfs interface for
determining the DRI driver is gone, and no working driver depends on
this mechanism.
Display string parsing is moved to a new function,
_eglSplitDisplayString.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
|
|
An unlinked context is destroyed after _eglMakeCurrent. Flushing such
context would cause segfault.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
|
|
The display may be NULL when checking a handle.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
|
|
|
|
|
|
The commit "ARP prog parser: Implement the spec, not what makes sense"
broke the parsing of scalar constants. This commit reverts that part
of that commit. Now vp_swizzle.c passes.
|
|
switches more state handling to common code. We need
should be more fine grained with the state atoms
eventually.
|
|
Slowing migrating to atom based system like
the other radeon drivers.
|
|
We probably want to go finer grained eventually, but
this is a good start.
|
|
|
|
|
|
|
|
Let eg. sp_get_samples_rect be hooked directly in as the tgsi sampler
routine.
Add a field to determine whether this is a vertex or fragment sampling
call, and massage parameters to match the tgsi call.
|
|
|
|
|
|
Fixes two issues - firstly for mipmap levels with one or more
dimensions smaller than tilesize, the code was sampling off the edge
of the texture (but still within the tile).
Secondly, in the linear_mipmap_linear case, both the default code and
new fastpath were incorrect. This change fixes the fastpath and adds
a comment to the default path, which still needs to be fixed.
Basically the issue is that the coordinates in the smaller texture
level are/were being computed by just dividing thecoordinates from the
larger texture level by two, as in:
x0[j] /= 2;
y0[j] /= 2;
x1[j] /= 2;
y1[j] /= 2;
The issues with this are signficant. Initially x1 is most often equal
to x0+1, but after this, it will likely be equal to x0, so we will not
actually be performing the linear blend within the smaller mipmap.
The fastpath code avoided this (recalculated x1), but was still using
the weighting factors from the larger mipmap level (xw, yw), which
were incorrect.
Change the fastpath code to do two full, independent linear samples of
the two mipmap levels before blending. The default code needs to do
the same thing.
|
|
Fall back to interpreter for now. This doesn't happen very often.
|
|
SOA dependencies can happen when a register is used both as a source and
destination and the source is swizzled. For example:
MOV T, T.yxwz; would expand into:
MOV t0, t1;
MOV t1, t0;
MOV t2, t3;
MOV t3, t2;
The second instruction will produce the wrong result since we wrote to t0
in the first instruction. We need to use an intermediate temporary to fix
this.
This will take more work to fix for all TGSI instructions. This seems to
happen with MOV instructions more than anything else so fix that case now
and warn on others.
Fixes piglit glsl-vs-loop test (when not using SSE). See bug 23317.
|
|
Users of the parser can make use of this.
|
|
At least its better then hardcoded to i915
|
|
|
|
Fall back to interpreter for now. This doesn't happen very often.
|
|
SOA dependencies can happen when a register is used both as a source and
destination and the source is swizzled. For example:
MOV T, T.yxwz; would expand into:
MOV t0, t1;
MOV t1, t0;
MOV t2, t3;
MOV t3, t2;
The second instruction will produce the wrong result since we wrote to t0
in the first instruction. We need to use an intermediate temporary to fix
this.
This will take more work to fix for all TGSI instructions. This seems to
happen with MOV instructions more than anything else so fix that case now
and warn on others.
Fixes piglit glsl-vs-loop test (when not using SSE). See bug 23317.
|
|
Users of the parser can make use of this.
|
|
fixes corruption issues with apps like teapot and
geartrain.
|
