| Age | Commit message (Collapse) | Author |
|---|
|
No idea why G80 doesn't hit this, but, this fixes at least one NV40 card.
|
|
Using this call in OUT_BATCH_TABLE reduces radeonEmitState cpu usage from
9% to 5% and emit_vpu goes from 7% to 1.5%. I did use calgrind to profile
gears for cpu hotspots with r500 card.
Signed-off-by: Pauli Nieminen <suokkos@gmail.com>
|
|
GCC did war about optimization not possible because possible forever loop.
Signed-off-by: Pauli Nieminen <suokkos@gmail.com>
|
|
Signed-off-by: Pauli Nieminen <suokkos@gmail.com>
|
|
|
|
Nasty, but nicer than silently not writing into the pushbuf
|
|
|
|
|
|
Noticed by vehemens on irc.
Signed-off-by: Dave Airlie <airlied@redhat.com>
|
|
Signed-off-by: Eric Anholt <eric@anholt.net>
|
|
|
|
This caches the mapping and just use mapping as a sync point
|
|
|
|
|
|
|
|
inbalances cpu_prep/cpu_finish
- The bo was mapped with sysmem == NULL, so this means cpu prep is called.
- The bo was unmapped with sysmem != NULL, so this means cpu finish is not called.
- This can lead to a non-zero "cpu writers" count in ttm_bo.
|
|
this rounds out the r6xx/r7xx stuff in radeon_drm.h
|
|
This lets us libdrm from master for r6xx/r7xx 3d
|
|
|
|
The goal of the BO cache is to keep buffers on hand for fast continuous use,
as in every frame of a game or every batchbuffer of the X Server. Keeping
older buffers on hand not only doesn't serve this purpose, it may hurt
performance by resulting in disk cache getting kicked out, or even driving
the system to swap.
Bug #20766.
|
|
this was giving me a warning on automake 1.11
|
|
|
|
Not sure what I was thinking.
|
|
If call was interrupted by signal we have to make call again.
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
|
|
Several POSIX extensions are used in the libdrm code (e.g., mknod and ffs).
Set _XOPEN_SOURCE and _GNU_SOURCE to something reasonable to ensure that
prototypes for these functions are available. This is done in configure.ac
using AC_USE_SYSTEM_MACROS. This requires autoconf 2.60 or later. Eventually
the code should check for the existance of these defines and do something
reasonable if they are not available.
Inspired by a patch by Pauli Nieminen and suggestions from Julien Cristau.
Thanks.
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
|
|
The logbase2 would overflow and wrap the size around to 0, making the code
allocate a 4kb object instead. By simplifying the code to just walk the
14-entry bucket array comparing sizes instead of indexing on
ffs(1 << logbase2(size)), we avoid silly math errors and have code of
approximately the same speed.
Many thanks to Simon Farnsworth for debugging and providing a working patch.
Bug #27365.
|
|
bug #21999
|
|
integers.
|
|
Based on patch by Pauli Nieminen. Thanks.
|
|
|
|
This ports a lot of the space checking code into a the common
library, so that the DDX and mesa can use it.
|
|
We always realloc at least 0x1000 dwords (page on most system)
when growing the cs buffer this is to avoid having to realloc
at each cs_begin.
|
|
This should use ndw not cdw, using cdw leads to realloc alignment going wrong
|
|
|
|
the DDX does this and used to handle it internally
|
|
|
|
Signed-off-by: Thomas Hellstrom <thellstrom-at-vmware-dot-com>
|
|
Match gtt map/unmap for correctness.
|
|
requires --enable-radeon-experimental-api for now
|
|
Normal map() should operate as before, and map_range()/map_flush() should
give correct results but lacking any performance difference from map().
Nothing exiting being done here yet, but the interface is a good start.
|
|
|
|
The commit 651e3dc6dd58a79c90db7513ee2fb28360a4560d, "drm: Fix
compilation on 2.6.30" broke all builds on kernels before 2.6.26.
Variadic functions cannot be inlined.
The variadic inline function is replaced with a variadic macro.
Also, the function dev_name() added by that same commit is never used,
so it is removed.
Signed-off-by: Pekka Paalanen <pq@iki.fi>
|
|
Fixes the dri1 gallium driver if the front buffer happens to be non-linear.
|
|
No idea what this reg does.. NVIDIA put 0x802 into it on some chips, the
criteria aren't too clear on when they do that however. Hopefully 0x800
will work everywhere...
|
|
This fixes DRM compilation of nouveau.ko on kernel 2.6.30.
Signed-off-by: Michael Buesch <mb@bu3sch.de>
|
|
|
|
Patch allows the backlight to be manipulated under gnome on apple powerpc
based NV30 machines. It works fine on my powerbook, and should also work
for older NV17/NV18 machines.
Note that older powerpc specific tools (pbbuttonsd) have some problems with
this implementation (because the device is not yet there at the start time
of the daemon, and the code makes incorrect assumptions about the max
brightness values). However, IMHO these things can and should be addressed
in the daemon.
Some style/warning fixes applied by Stuart Bennett <stuart@freedesktop.org>
|
|
|
|
This avoids making objects significantly bigger than they would be
otherwise, which would result in some failing at binding to the GTT.
Found from firefox hanging on:
http://upload.wikimedia.org/wikipedia/commons/b/b7/Singapore_port_panorama.jpg
due to a software fallback trying to do a GTT-mapped copy between two 73MB
BOs that were instead each 128MB, and failing because both couldn't fit
simultaneously.
The cost here is that we get no opportunity to cache these objects and
avoid the mapping. But since the objects are a significant percentage
of the aperture size, each mapped access is likely having to fault and rebind
the object most of the time anyway.
Bug #20152 (2/3)
|
|
The convention is that all APIs are per-bufmgr, so make this one the same.
Then, have it return -1 on failure so that the application can know what's
going on and do something sensible.
Signed-off-by: Keith Packard <keithp@keithp.com>
|
