Age | Commit message (Collapse) | Author |
|
allows one to specify a safe (bound checked) array
filled with immediates. it works just like a const
array and declares much like our current immediates.
|
|
|
|
llvmpipe can create a large number of shader variants for a single shader
(which are quite big), and they were only ever deleted if the shader itself
was deleted. This is especially apparent in things like glean
blendFunc where a new variant is created for every different subtest, chewing
up all memory.
This change limits the numbers of fragment shader variants (for all shaders)
which are kept around to a fixed number. If that would be exceeded a fixed
portion of the cached variants is deleted (since without tracking the used
variants this involves flushing we don't want to delete only one).
Always the least recently used variants (from all shaders together) are
deleted.
For now this is all per-context.
Both the number of how many variants are cached (1024) as well as how many
will be deleted at once (1/4 of the cache size) are just rough guesses and
subject to further optimization.
|
|
Early Z test (=ZTOP) must be disabled before a query is started,
otherwise the GPU is dead. The order of emitted registers matters more
than you might think.
This fixes hardlocks in sauerbraten.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
fixes a crash.
|
|
spotted by Vinson
|
|
A platform is already used to mean a window system in EGL. No need to
use a different term.
|
|
A platform means a native window system in EGL. Use OS that follows
Gallium instead.
|
|
Target egl_fbdev_swrast.so needs sw/fbdev/libfbdev.a.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
We *did* reach an agreement on this a few months ago, and now the docs
reflect it. However, we never got around to UV and Z...
|
|
I'm not sure I picked the best voice here. I might come back to this
later.
|
|
|
|
|
|
|
|
we kept overwriting the first attribute of a vertex in a single-stream-
out-buffer case
|
|
|
|
|
|
register masks, multiple output buffers, multiple primitives,
non-linear vertices (elts) and stride semantics.
|
|
|
|
Yay for old code.
|
|
|
|
|
|
|
|
|
|
Need to just add the resolve, then go switch to new DRM and test.
|
|
I could *not* let this slide since I'm on a 78-char-wide terminal.
|
|
|
|
|
|
|
|
|
|
|
|
When building OSMesa and xlib GL, the resulting OSMesa would be linked
against libGL instead of the internal mesa libraries. However, when
building with -fvisibility=hidden, some of the internal functions used
in OSMesa could not be resolved through libGL.
Instead, always build OSMesa standalone without linking against libGL.
This has the advantage that OSMesa is always built the same way, but it
means that disk space is wasted when libGL is installed since both
libraries will contain the internal objects.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Tested-by: Tom Fogal <tfogal@alumni.unh.edu>
|
|
|
|
|
|
|
|
|
|
|
|
If the default framebuffer is bound to <target>, then
<attachment> must be one of FRONT_LEFT, FRONT_RIGHT, BACK_LEFT,
BACK_RIGHT, AUXi, DEPTH_BUFFER, or STENCIL_BUFFER, identifying a
color buffer, the depth buffer, or the stencil buffer, and
<pname> may be FRAMEBUFFER_ATTACHMENT_OBJECT_TYPE or
FRAMEBUFFER_ATTACHMENT_OBJECT_NAME.
as well as these <pname> values
FRAMEBUFFER_ATTACHMENT_RED_SIZE,
FRAMEBUFFER_ATTACHMENT_GREEN_SIZE,
FRAMEBUFFER_ATTACHMENT_BLUE_SIZE,
FRAMEBUFFER_ATTACHMENT_ALPHA_SIZE,
FRAMEBUFFER_ATTACHMENT_DEPTH_SIZE,
FRAMEBUFFER_ATTACHMENT_STENCIL_SIZE,
FRAMEBUFFER_ATTACHMENT_COMPONENT_TYPE, or
FRAMEBUFFER_ATTACHMENT_COLOR_ENCODING.
https://bugs.freedesktop.org/show_bug.cgi?id=28551
|