| Age | Commit message (Collapse) | Author |
|---|
|
|
|
Before, changing any of these sampler values triggered generation
of new JIT code. Added a new flag for the special case of
min_lod == max_lod which is hit during auto mipmap generation.
|
|
Allows disabling various operations (mainly texture-related, but
will grow) to try & identify bottlenecks.
Unlike LP_DEBUG, this is active even in release builds - which is
necessary for performance investigation.
|
|
|
|
|
|
|
|
'llvm' suffix unnecessary now that the C sampling version disappeared.
|
|
Finally a substantial performance improvement: framerates of apps using
texturing tripled, and furthermore, enabling/disabling texturing only
affects around 15% of the framerate, which means the bottleneck is now
somewhere else.
Generated texture sampling code is not complete though -- we always
sample from the base level -- so final figures will be different.
|
|
translation.
|
|
This is more a short term experiment than a long term commitment, as we'll
need to support higher precision textures too, as this will all be
be replaced by runtime generated code.
With this change most Mesa demos fps increased around 10%. Not a huge
improvement, but not a negligible one either.
|
|
|
|
|
|
The lp_tile_cache is often called repeatedly to look up the same
tile. Add a cache (to the cache) of the single tile most recently
retreived and make a quick inline check to see if this matches the
subsequent request.
Add a tile_address bitfield struct to make this check easier.
|
|
The tile cache is a utility, it shouldn't know anything about the
entity which is making use of it (ie llvmpipe).
Remove llvmpipe parameter to all the tilecache function calls, and
also remove the need to keep a llvmpipe pointer in the sampler structs.
|
|
|
