This implements fast Z clear, Z compression, and HiZ support for r300->r500
GPUs.
It also allows cbzb clears when fast Z clears are being used for the ZB.
It requires a kernel with hyper-z support.
Thanks to Marek Olšák <maraeo@gmail.com>, who started this off, and Alex Deucher at AMD for providing lots of hints.
v2:
squashed zmask ram size fix]
squashed r300g/blitter: fix Z readback when compressed]
v3:
rebase around texture changes in master - .1 fix more bits
v4:
migrated to using u_mm in r300_texture to manage hiz/zmask rams consistently
disabled HiZ when using OQ
flush z-cache before turning hyper-z off
update hyper-z state on dsa state change
store depthclearvalue across cbzb clears and replace it afterwards.
Signed-off-by: Dave Airlie <airlied@redhat.com>
The BGNLOOP and ENDLOOP instructions are now being used correctly, which
makes break and continue possible. The deadcode pass has been modified to
handle breaks, and the compiler is more careful about which loops are
unrolled.
When we have instance divisors we don't really know which vertex
elements we'll be fetching ahead of time.
This fixes a bug in instanced drawing which was exposed by the new
draw_vbo() code because of max_index not being ~0 as often as it used
to be. The test for max_index >= DRAW_PIPE_MAX_VERTICES often hid
this problem before.
It's easily reproducible with Compiz with its Resize window mode
set to Normal (which is usually not the default mode).
https://bugs.freedesktop.org/show_bug.cgi?id=28658https://bugs.freedesktop.org/show_bug.cgi?id=29303
This is actually a workaround to prevent Compiz crashes.
Instead, a completely white titlebar might show up during resizing
transparent windows (a rare case).
The underlying cause should be fixed by someone who has more knowledge
about the code. (dri2_drawable_get_buffers should not return NULL)
Acked-By: Jakob Bornecrantz <jakob@vmware.com>
Make sure LIT fills all slot for instruction (can't do W instruction
without having the Z slot filled with at least a NOP).
ALU instruction can't access more than 4 constant, move constant to
temporary reg if we reach the limit.
Fix ALU block splitting, only split ALU after ALU with last instruction
bit sets.
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
If texture coordinates come from the vertex shader, there are always
4 components in the rasterizer input packet, but if the coordinates
are stuffed (like for point sprites), there are only 2 or 3 components
(based on GB_ENABLE) and if we rasterize more, it locks up.
