| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
| |
Spoted by Alex Diomin
Signed-off-by: Jerome Glisse <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
r600g is up to a point where all small CPU cycle matter and pb* turn
high on profile. It's mostly because pb try to be generic and thus
trigger unecessary check for r600g driver. To avoid having too much
abstraction & too much depth in the call embedded everythings into
r600_bo. Make code simpler & faster. The performance win highly depend
on the CPU & application considered being more important on slower CPU
and marginal/unoticeable on faster one.
Signed-off-by: Jerome Glisse <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
R6XX GPU doesn't like to have two partial flush writting
back to memory in row without a prior flush of the pipeline.
Add PS_PARTIAL_FLUSH to flush all work between the CP and
the ES, GS, VS, PS shaders.
Thanks a lot to Alban Browaeys (prahal on irc) for investigating
this issue.
Signed-off-by: Alban Browaeys <[email protected]>
Signed-off-by: Jerome Glisse <[email protected]>
|
|
|
|
| |
Signed-off-by: Jerome Glisse <[email protected]>
|
| |
|
|
|
|
| |
Signed-off-by: Jerome Glisse <[email protected]>
|
|
|
|
| |
On lightsmark on my r500 this drop the bufmgr allocations of the sysprof.
|
|
|
|
| |
Signed-off-by: Alex Deucher <[email protected]>
|
|
|
|
| |
Are these functions actually used anywhere?
|
|
|
|
| |
Add explicit EVENT_TYPE field
|
|
|
|
|
|
|
|
| |
Use fetch shader instead of having fetch instruction in the vertex
shader. Allow to restrict shader update to a smaller part when
vertex buffer input layout changes.
Signed-off-by: Jerome Glisse <[email protected]>
|
|
|
|
| |
6xx-evergreen
|
|
|
|
|
|
|
| |
Occlusion query on evergreen need the event index field to be
set otherwise we endup locking up the GPU.
Signed-off-by: Jerome Glisse <[email protected]>
|
|
|
|
| |
This should fix the remaining buffer alignment issues in r600g.
|
| |
|
|
|
|
|
|
|
|
| |
For driver performance analysis it usefull to be able to
disable as much as possible the GPU interaction so that
one can profile the userspace only.
Signed-off-by: Jerome Glisse <[email protected]>
|
|
|
|
| |
Fixes crash in piglit depthrange-clear.
|
| |
|
|
|
|
|
|
|
| |
These were previously being left in the default (D3D) mode. This mean
that triangles were drawn slightly incorrectly, but also because this
state is relied on by the u_blitter code, all blits were half a pixel
off.
|
|
|
|
|
|
|
|
|
| |
This opens the question of what interface the winsys layer should
really have for talking about these concepts.
For now I'm using the existing gallium resource usage concept, but
there is no reason not use terms closer to what the hardware
understands - eg. the domains themselves.
|
|
|
|
| |
Signed-off-by: Tilman Sauerbeck <[email protected]>
|
|
|
|
|
|
| |
That way assert(map_count >= 0) can actually fail when we screwed up.
Signed-off-by: Tilman Sauerbeck <[email protected]>
|
|
|
|
| |
Signed-off-by: Tilman Sauerbeck <[email protected]>
|
|
|
|
|
|
|
|
| |
This ensures that we increase bo->map_count when radeon_bo_map_internal()
returns successfully, which in turn makes sure we don't decrement
bo->map_count below zero later.
Signed-off-by: Tilman Sauerbeck <[email protected]>
|
|
|
|
| |
Signed-off-by: Tilman Sauerbeck <[email protected]>
|
|
|
|
|
|
|
|
| |
radeon_bo_destroy() will want to read the list field. Without this patch,
we'd end up evaluating the list pointers before they have been properly
set up when we destroyed the newly created bo if it cannot be mapped.
Signed-off-by: Tilman Sauerbeck <[email protected]>
|
| |
|
|
|
|
| |
we need to know if the back is tiled so we can blit from it properly.
|
| |
|
|
|
|
|
|
|
| |
Would try to destroy radeon->cman, radeon->kman both which were still
NULL.
Signed-off-by: Dave Airlie <[email protected]>
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
Avoid having object ending up in dead list of dirty object.
Signed-off-by: Jerome Glisse <[email protected]>
|
|
|
|
| |
Signed-off-by: Jerome Glisse <[email protected]>
|
|
|
|
| |
Signed-off-by: Jerome Glisse <[email protected]>
|
|
|
|
|
|
|
| |
Since flush rework there could be only one relocation per
register in a block.
Signed-off-by: Jerome Glisse <[email protected]>
|
|
|
|
| |
Got a speed up by tracking the dirty blocks in a seperate list instead of looping through all blocks. This version should work with block that get their dirty state disabled again and I added a dirty check during the flush as some blocks were already dirty.
|
|
|
|
|
|
|
|
|
| |
Flush read cache before writting register. Track flushing inside
of a same cs and avoid reflushing same bo if not necessary. Allmost
properly force flush if bo rendered too and then use as a texture
in same cs (missing pipeline flush dunno if it's needed or not).
Signed-off-by: Jerome Glisse <[email protected]>
|
|
|
|
|
|
|
| |
Allow fast lookup of relocation information & id which
was a CPU time consumming operation.
Signed-off-by: Jerome Glisse <[email protected]>
|
|
|
|
|
|
| |
if we know the bo has gone not busy, no need to add another bo wait
thanks to Andre (taiu) on irc for pointing this out.
|
|
|
|
| |
since we plan on using dx10 constant buffers everywhere.
|
|
|
|
| |
we don't use this since constant buffers are now being used on all gpus.
|
|
|
|
|
|
| |
When we go to do a lot of bos in one draw like constant bufs we need
to avoid bouncing off the busy ioctl, this mitigates by backing off
on busy bos for a short amount of times.
|
|
|
|
|
| |
this just keeps a list of bos submitted together, and uses them to decide
bo busy state for the whole group.
|
| |
|
|
|
|
| |
Signed-off-by: Jerome Glisse <[email protected]>
|
|
|
|
| |
Signed-off-by: Jerome Glisse <[email protected]>
|