Not much big news here. Another point release, another few bugs
squashed. A new libdrm_intel_function lets userland tell libdrm to use
an active buffer when available, so that pixmap allocations don't have
to wait when they're just going to be written to by the GPU as
temporaries. This saves piles of memory for some workloads. Also, the
GTT fence management means that we won't send batchbuffers that will
fail to render due to needing more fence regs than are available on
pre-965 hardware.
Most of the diff here is nouveau code that remains off by default.
Ben Skeggs (15):
nv04-nv40: correct RAMHT size
nv50: support chipset NV96
nouveau: bring in new mm api definitions, without the actual mm code
nouveau: install libdrm_nouveau with libdrm
nouveau: fix some issues where buffer objects never get freed
drm/nv50: let the card handle the initial context switch
drm/nv50: correct ramfc pointer in channel header
drm/nv50: use a slightly different initial context for nv96
drm/nv50: add context info for nv98
drm/nv50: fix nv9x chipsets
libdrm: make libdrm_nouveau conditional, disabled by default
nv50: context info for chipset 0xa0
nv40: fail completely if we don't have a ctxprog for the chipset
libdrm/nouveau: fix dma debugging
libdrm/nouveau: free drmVersion after we're done with it
Dave Airlie (1):
drm: remove drmstat/dristat from linux-core build
Eric Anholt (8):
intel: Fix under-counting of fences registers required in check_aperture.
intel: don't count fences on 965 and later, as they don't use
them.
Clean up README for the current state of kernel module affairs.
intel: Clear ioctl parameter outvalues and pads that confuse valgrind.
intel: Add a new bufmgr alloc function to get BOs ready for rendering to.
Bump version to 2.4.5 for new API additions.
Fix distcheck for optional nouveau stuff.
Add libdrm_intel.pc by popular demand.
Ivan Kokshaysky (1):
Fix DRM_CAS() on Alpha. (#16549)
Jakob Bornecrantz (1):
mode: Make xfdrmMode.[c|h] not depend on drm_mode.h
Jesse Barnes (1):
intel: libdrm support for fence management in execbuf
Kristian H?gsberg (7):
No need to pin buffer anymore in modetest.c
modetest: Use a more interesting test pattern.
modetest: Handle setting modes on multiple crtcs with one fb.
modetest: Don't sleep when just dumping state, wait for key for
modeset.
modetest: Use cairo (if available) to paint a better pattern.
modetest: Add syntax to also specify a crtc when setting a mode.
modetest: Print names of properties.
Matthew Garrett (2):
nouveau: Add in-kernel backlight control support
nv40, nv50: fix backlight build for <2.6.29 kernels
Pekka Paalanen (6):
drm compat: fix euid for >=2.6.28
nouveau: fix type warnings: void* arith, un/signed
Remove i915 driver sources from linux-core
drm_compat: remove kmap_atomic_prot_pfn()
nouveau: support backlight only when kernel does
Fix fix distcheck for optional nouveau stuff.
Stephane Marchesin (1):
Remove the "nv" driver.
Stuart Bennett (3):
nouveau: no suspend support for nv50+
nouveau: don't save channel context if it has recently become invalid
nouveau: don't try to traverse non-existent lists
git tag: libdrm-2.4.5
http://dri.freedesktop.org/libdrm/libdrm-2.4.5.tar.bz2
MD5: 0d09025f069ff1a986cf3339e51908b3 libdrm-2.4.5.tar.bz2
SHA1: e27257b9464916f9483c40e48fe340d4fb3f9750 libdrm-2.4.5.tar.bz2
http://dri.freedesktop.org/libdrm/libdrm-2.4.5.tar.gz
MD5: 4d86f50bab4f36d5a3c63f20086736a5 libdrm-2.4.5.tar.gz
SHA1: 9e919c4cda6454118ebd288edc3bc65fc6779a51 libdrm-2.4.5.tar.gz
--
Eric Anholt
eric at anholt.net eric.anholt at intel.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part
Url :
http://lists.freedesktop.org/archives/xorg-announce/attachments/20090223/4b273340/attachment.pgp