libdrm: Changes to 'refs/tags/libdrm-2.3.0-3'
Tag 'libdrm-2.3.0-3' created by David Nusinow <dnusinow@debian.org> at 2007-04-15 17:39 +0000
Tagging upload of libdrm-2.3.0-3 to experimental.
Changes since libdrm-2.3.0-1:
Aapo Tahkola (4):
Sync from r300_reg.h from Mesa.
Add support for texture cache flushes (R300_TX_CNTL)
Add general-purpose packet for manipulating scratch registers (r300)
ia64 support for r300_scratch. (not tested)
Adam Jackson (36):
Bug #979: Don't include <sys/sysmacros.h> on linux, it occasionally emits
Shared libdrm work:
Add xf86drm.h to the install target.
Add a Savage3D PCI ID
Add drm.h to the install target.
Final pass of libdrm.so work:
momentary revert. libdrm build may be broken now, but no one uses it yet...
Bug #2567: Fix reversed memset arguments. (David Krause)
Add skeletal imagine driver (but don't build it yet).
autoconfiscate libdrm
distcheck fixes
Fix the pkgconfig info. (Donnie Berkholz)
version bump
Add r300_reg.h. Bump to 1.0.2.
Fix silly install issue by moving the header install rules for shared-core
Skeleton nv drm driver, to enable DMA in EXA. (Lars Knoll, minor updates by
Better pkgconfig-fu: -ldrm in Libs:
Remove bogus Xlib dependency.
via and r300 still need installed reg headers.
Remove the remaining references to Xlib. libdrm is totally independent now.
Bump to 1.0.4.
Bug #328: Silence read-from-uninitialized warnings in DRM_CAS on Alpha.
Bump to 1.0.5 for new via header
remove xf86drmCompat.*, unused since about XFree86 4.2.
Error checking for drmStrdup (Tilman Sauerbeck)
Remove redundant #ifdef stanza
Add TODO based on my email from a while ago
Resync from Xorg head (Bug #3815, GNU/kFreeBSD hack)
Resync from Xorg head: conditional include of xorg-config.h
Bump package and DSO numbers to 2.0 to reflect 32/64 ABI change
Formatting cleanup, dead code removal. Remove N() namespacing macro,
Avoid walking off the end of the hash table. (Coverity report #465)
Bump driver date to reflect airlied's last fix.
Disable building static libraries. Bump to 2.0.2 for header updates.
Make sure busmastering gets disabled on module unload.
Delete the pre-core DRM code with extreme prejudice.
Alan Hourihane (113):
Include new updated gamma support.
Fix define of PCI_DEVICE_ID_3DLABS_GAMMA when not available in the kernel
Import of XFree86 4.0.1
Merge XFree86 4.0.1
Final misc changes to bring upto 4.0.1
Merged tdfx-2-1-branch
axp cast fix.
Import of XFree86 4.0.1d-pre
Import of XFree86 4.0.1d-pre (2)
commit xfree86 4.0.1d-pre update
file gamma.h was initially added on branch mga-1-0-0-branch.
file gamma_drm.c was initially added on branch mga-1-0-0-branch.
file gamma_context.c was initially added on branch mga-1-0-0-branch.
Don't try and setup the MTRR for AGP when AGP not available. Check
fix that last patch to initialize the MTRR when AGP available.
allow dristat to find out whether AGP is write-combined or not.
surround agp calls in drm_memory with __REALLY_HAVE_AGP instead of
file agpgart.diff was initially added on branch gamma-2-0-0-branch.
merge in 2.4.3 kernel change.
make 2.4.2 -> 2.4.3 change conditional. works with older than 2.4.3 kernels
include 2.4.2
Handle drivers that don't have __HAVE_SG defined.
search /lib/modules/<version>/build/include first, then the others....
file Makefile was initially added on branch bsd-2-0-0-branch.
file drm_os_linux.h was initially added on branch bsd-2-0-0-branch.
file drm_os_freebsd.h was initially added on branch bsd-2-0-0-branch.
file drm_sysctl.h was initially added on branch bsd-2-0-0-branch.
file drm_vm.h was initially added on branch bsd-2-0-0-branch.
fix build of i810 kernel driver for 2.4.3 or greater kernels
remove deprecated file
Make SiS driver compile with the new templated format. Not tested. minor
add new file
file drm_agpsupport.h was initially added on branch bsd-2-0-0-branch.
file drm_linux.h was initially added on branch bsd-2-0-0-branch.
add some commented enclosures around the BUS_BASE calls for Alpha
Make the SiS module work again. At least glxinfo reports it's working, yet
new multihead code was missing Voodoo3 2000 and Voodoo4 support.
No one's maintaining 2.2.x support - so remove all the cruft.
file convert.c was initially added on branch bsd-2-0-0-branch.
Use CCE for 2D acceleration (Gerd Knorr)
bumped the DRM versions (already done in ddx and client side drivers).
remove Linux 2.3.x cruft.
merge with 2.4.10 kernel
commit Abraham vd Merwe fix.
merge kernel 2.4.13-pre6.
wrap the MODULE_LICENSE definition.
merge with linux kernel 2.4.15
first pass at merging mesa 4.0 kernel drivers into new bsd-3-0-0 branch.
i830 & mga contain minor changes from 4.2.0 for mesa 4.0 bsd merge
fixup the radeon driver (not tested)
missing file
fixups for *BSD
file trident.h was initially added on branch trident-0-0-1-branch.
file trident_drv.c was initially added on branch trident-0-0-1-branch.
same udelay fixes
Remove some older Linux 2.3.99 code
gamma_alloc -> DRM(alloc)
merged bsd-3-0-0-branch
remove files missed by merge
remove obsolete files
Import of XFree86 4.2.99.2
final part of XFree86 4.2.99.2 merge
XFree86 4.3.0 merge
linux merge for drm
fix gamma headers
remove unused variable
change PREINSTALL/POSTINSTALL/UNINSTALL irq code to real functions as per
DRM_READ/WRITEMEMORYBARRIER was given an argument in the bsd tree, fix for
file savage_state.c was initially added on branch savage-1_0_0-branch.
file savage_dma.c was initially added on branch savage-1_0_0-branch.
resolve merge conflicts
linux drm fixes
bsd drm fixes
compatibility layer still uses agp (not gart)
post merge fix
Changes to DRM(irq_install)...... wrap dev->dma usage with __HAVE_DMA in
disable PCI DMA ioctls as they are not used currently until SAVAGE_CMD_DMA
fix build problem
remove Imakefile
drm_pciids.h should only 'cleaned' when the shared/ directory exists. Which
Add i915GM support Add resume functionality (must be used with later DDX)
Fix a cut & paste problem
Fix cut & paste problem
Don't know who set the i915GM PCI ID incorrectly but it should be 0x2592
Re-implement the power management.
Actually check for drm_fb_loaded before trying to initialize the sysdev
Fix copyrights
Force AGP always for Intel chipsets.
Fix suspend problem when i915 private hasn't been initialized yet.
Fix the sysdev approach for power management.
Save state of registers for suspend/resume.
Check dev_priv
Check for 2.6.12 suspend/resume changes (pm_message_t)
Reverse the pm_message_t patch for now, it appears that the 2.6.12 release
Comment out the .resume function as without programming their dependent
Given that BenH says using the sysdev approach for DRM is bogus, I'll yank
Move to linux specific directory
silence warning
add i915_ioc32.c
add remaining calls
remove i915_pm code as it causes too many issues with current software
remove the README in the linux-2.6 directory
Don't set MTRR's for intel hardware
Fix bug #4908 for now. Alternative is highlighted in this bug report as the
add Intel i945GM support
Fix CMDBUFFER path, add heap destroy and flesh out sarea for rotation
put back some of Dave's code related to vblank support - accidentally
other part of revert
fix an error message typo Unkown -> Unknown
add missing name, desc, date, major, minor and patchlevel entries
Fix compilation problem on 2.6.9 kernels (bug #6211)
Add support for Intel i965G chipsets.
Fix bug #8839 - a comment
Anish Mistry (1):
bug 7092 : add pci ids for mach64 in Dell poweredge 4200
Brian Paul (36):
Merged glxmisc-1-0-0
file vm.c was initially added on branch bsd-1-0-0-branch.
file tdfx_drv.h was initially added on branch bsd-1-0-0-branch.
file tdfx_context.c was initially added on branch bsd-1-0-0-branch.
file sysctl.c was initially added on branch bsd-1-0-0-branch.
file proc.c was initially added on branch bsd-1-0-0-branch.
file memory.c was initially added on branch bsd-1-0-0-branch.
file lock.c was initially added on branch bsd-1-0-0-branch.
file lists.c was initially added on branch bsd-1-0-0-branch.
file ioctl.c was initially added on branch bsd-1-0-0-branch.
file init.c was initially added on branch bsd-1-0-0-branch.
file fops.c was initially added on branch bsd-1-0-0-branch.
file drmstat.c was initially added on branch bsd-1-0-0-branch.
file drawable.c was initially added on branch bsd-1-0-0-branch.
file dma.c was initially added on branch bsd-1-0-0-branch.
file context.c was initially added on branch bsd-1-0-0-branch.
file bufs.c was initially added on branch bsd-1-0-0-branch.
file auth.c was initially added on branch bsd-1-0-0-branch.
file gamma_dma.c was initially added on branch bsd-1-0-0-branch.
file gamma_drv.c was initially added on branch bsd-1-0-0-branch.
file gamma_drv.h was initially added on branch bsd-1-0-0-branch.
file tdfx_drv.c was initially added on branch bsd-1-0-0-branch.
file drm.h was initially added on branch bsd-1-0-0-branch.
file Imakefile was initially added on branch bsd-1-0-0-branch.
file Makefile.bsd was initially added on branch bsd-1-0-0-branch.
file drmP.h was initially added on branch bsd-1-0-0-branch.
define VM_DONTCOPY to zero if undefined
Merged glxmisc-3-0-0
don't draw if window is unmapped, other updates (Jeff Hartmann)
applied Jeff's xf86cvs-I810copy.patch
include linux/sisfb.h, per David Gaarenstroom
replaced max() macro with conditional expression
merge from mesa-4-1-branch to get cube-map registers. bumped version to 1.7
added missing return fd (Alexander Stohr)
fix mem leak in HashHash() (bug 5171)
check for __FreeBSD_kernel__ (bug 3810)
Daryll Strauss (9):
Initial revision
Import XFree 3.9.16
First DRI release of 3dfx driver.
Move Mesa to xc/extras Update to the latest Mesa 3.2 code Fix the Q3Demo
Rename the device to be /dev/dri instead of /dev/dri to avoid a conflict.
Add a drm_poll function to the tdfx driver. This fixes the problem with
Remove -g from build Fix lost code from tdfx-1-1 merge in tdfx_priv.c Lower
Only enable DRI for 16bpp mode. Allow GL_BLEND if color is 0,0,0,1 Wrap
Merged the tdfx-2-0 branch onto the trunk
Dave Airlie (306):
add page flipping support to the DRM, up version number to 1.3.0...
fix pitch compile error
DA: fix for bug 484 in Bugzilla, originally from me, reworked by David
DA: code cleanups for i810_dma.c from 2.4 kernel
DA: patch from Matthew upgraded to latest DRI head to solve issue with i810
DA: loads of whitespace .. some from Linus, some from me
initial shell script to create linux kernel drm from the DRI one
make sure in DRM toplevel
copy correct Makefile
patch from lk
fixes from Linux kernel
patch from Linux kernel 2.6.5
more 2.4 compat fns
2.6 patches for larger dev_t
long dev_t patches from 2.6 tree
2.6 sysfs patches + stubs in drmP.h for 2.4 compatibility
big whitespace .. this aligns all the whitespace in this file with the bk
update from linux kernel for ia64
remove unused code
align code with Linux kernel.
patch from Andrew Morton tree from Arjan van de Ven fixes some oopses seen
update gamma_dma with patch from davej in 2.6
include highmem.h
white space changes to align with kernel
Add mach64 to the trunk
more files for mach64
centralise pci ids into one place and use scripts to generate files for
bug from Linux kernel list caught by checker
add new files to generate pci ids
add sparc ffb files from kernel just to keep things in sync
Add __user annotations from kernel
add another tdfx
better device class support from Jon Smirls patches..
fix 2.4 build
use drm.. not dri for class...
make drm pciids
Commit sysfs and drm PCI changes for 2.6 kernel
do some real testing and fix the DRM initialising and unloading
2.4 compat fix
another 2.4 fix
fixes from kernel: Make users of page->count use the provided macros
2.4 compat
The dev->devname being passed to request_irq in drm_irq.h is null. With the
gamma_dma_priority and gamma_dma_send_buffers both deref d->send_indices
dirty hack to make mach64 work, (don't worry I'll get around to fixing it
fix bug with pci_disable_device in the wrong place (Paul Mackerras) remove
whitespace align with kernel
fixes from kernel for 0 vs NULL - mika
align with kernel
fix issue in 2.4 kernels with returning NULL from this function
split out backwards compat into a separate file makes it easier for merging
allow O= usage for Linux 2.6 kernel building in another directory
Fix reference counting for stub for new Linux PCI probe
sparse cleanups from kernel: Al Viro
attempt to clean up stub_register, register_chrdev isn't good enough to
fix some more NULLs from kernel
add x86_64 architecture defines from kernel (leave AMD64 defines in for
Kconfig update add i915 bring over some stuff from kernel
Add NULLs instead of 0 for i915
first set of __user annotations from kernel (Al Viro)
add some more debugging fix ++ and --
ATI Rage 128 and Radeon DRM unconditionally depend on PCI
another logic error returns 0 or greater for success
Patch from Tom Arbuckle for missing bus_address
if the driver has already register don't do another intermodule register
whitespace merge with kernel
sync up with current 2.6 kernel bk tree - mostly __user annotations
check for __user if not there define it
initial port of i915 to BSD, not finished doesn't work.. no idea why...
change to agp not pci ... still not working
add read/write 16
athe patch below optimises the drm code to not do put_user() on memory the
fixes for using userspace pointers found by sparse utility
Patch picked up from Redhat xorg release:
patch from RH xorg-x11 tree ported to drm.h
forgot to check these in.. thanks to Jon for reminding me ..
fix for drm in /proc - from Jon Smirl
bring over fix from i865-agp branch, it now probes the driver, X hangs
2.4 hotplug compat
2.4 compat
Patch from Jon Smirl to add attribute field to the pciids, and use this for
minor patch from Jon Smirl : sets up some things for later use
2.6.8.1 has changed the links in /lib/modules
preparation patch for radeon permanent mapping registers/framebuffer makes
Merged drmfntbl-0-0-1
set pointers to NULL after freeing, remove some extra debugging
addmap-base-2 patch from Jon Smirl:
Merged drmfntbl-0-0-2
__NO_VERSION__ hasn't been needed since 2.3 days ditch it...
run i915 through lindent
fixup issue caused by fntbl-2 merge
Drop GAMMA DRM from a great height ...
remove hacky context thing that was gamma only
fix up some small things ffb (no idea if it works or not ..)
add context include so fn are defined
drm-memory patch, cleans up alloc/free and makes calloc look more libc like
implement drm_core_check_feature and use it .. looks lots nicer
Apply patch from Bryan Stine bugzilla #1227
oops called ctor instead of dtor.. found this on the kernel merge
file drmcore_exports.c was initially added on branch drmlib-0-0-1-branch.
file drm_proc.c was initially added on branch drmlib-0-0-1-branch.
file drm_memory_debug.c was initially added on branch drmlib-0-0-1-branch.
file drm_memory.c was initially added on branch drmlib-0-0-1-branch.
file drm_headers.h was initially added on branch drmlib-0-0-1-branch.
file drm_core_memory.h was initially added on branch drmlib-0-0-1-branch.
file drm_agpsupport.c was initially added on branch drmlib-0-0-1-branch.
file drm_core.h was initially added on branch drmlib-0-0-1-branch.
doh.. that makes no sense.. thinko in removal of OS_HAS_AGP
Fixup OS_HAS_AGP/OS_HAS_MTRR along lines of patches going to kernel, as
make the AMD64 check a compat thing
bad code copy for alpha.. fix the member names
merge back bunch of whitespace and misc changes from kernel
missed fix as part of last checkin
another fix after the macro stuff
remove HAVE_COUNTERS
Patch for Kconfig for making i830/i915 not build together
remove unused dma remnants that were gamma only - these could cause an oops
Lindent the via stuff so I can include it in kernel
fix up whitespacing in Kconfig
cleanup VIA driver to look a bit like others before kernel merge
Vladimir requested support so we can at least load r300 microcode for
Apply radeon r300 microcode patch to non-core
fix pfn vs page for older kernels (2.6.9-rc kernels many not work..)
actually 2.6.10 introduced pfn range so it should work now..
fix inter module put/get
fix for 2.4 build
respect cc and cflags
add some more r300 pci ids
patch from bug 1803 - will try and push to kernel soon
Make 1-bit fields be unsigned (no sign bit :). sparse complains about them:
Use wbinvd macro instead of assembly for it,
i810/i830 bug with Jon's file operations changes
due to backways compat crap for IRQ_HANDLED etc.. this wouldn't build on
The patch makes drmAddBufs/drmMapBufs can handle buffers in video memory
This table is in radeon_drv.h along with the version number so remove it
remove duplicate drm_sysfs.o
hopefully fix drm_core setversion ioctl... pointed out by Eric on irc
use libdrm after it was updated
fix incorrect PCI id for ATI radeon
fix radeon drv
make more functions static in i810 and fix pageflip cleanup
update i830 similiar to the kernel, add statics
make functions static in i915, remove unused functions
cleanup patch from Adrian Bunk <bunk@stusta.de>
fix agp detection on linux
change DRIVER_ to CORE_ makes things look nicer, also change it so the
Invalid bound check of driver defined ioctls in drm_ioctl
add some idct type 3 packets for reference
minor pointer cast typo
Fix for bug 2673 from Egbert Eich - memset the version
For 2.6.12 and 2.6.11.x:
3dfx DRM depends on PCI
fix bug with XFree86 4.3 on core drm
verify_area is deprecated, replaced by access_ok. Seems I missed this one
fix up AGP multi-head support for kernel 2.6.12
use linux-core for 2.6 not this stuff..
From Xorg CVS realign the i915_drv.h
Some ia64 platforms may not support write combining on all type of memory,
Description: Rather than use custom code in DRM_WAIT_ON() to do exactly
mirror changes made in main tree.. just happened to be doing this myself
Revert last commit, it affect via things
wrap config.h include with defined for KERNEL
Add i945G pci ids to drm
add radeon registers from VHA code these are the "unknown" registers
Remove warnings about code mixed with declerations..
misc cleanup patch from Adrian Bunk
fix some issues with radeon interrupt handling
add dragonfly #define from Xorg CVS
fix up drm.h for C++ include as this can be included in user programs
fix up drm_alloc_agp to take a dev arg and not pass crappy agpgart around
Add drm and radeon 32/64-bit compat code from Paul Mackerras
Add compat to Makefile
add compat code from Paul Mackerras
Bug in conversion from old DRM to core DRM....
these don't need reclaim buffers their release functions handle it
add mga and r128 32/64 bits
make r128/mga compile properly on sparc cross-compiler
fix some warnings from cross compiler
Egbert noticed this issue, but changing the IOW to IOWR mess up
up the version
revert last change due to me bring up too early
Fix bug in return to userspace resctx code
remove bus address
fix ioctl in r128 drm direction from Egbert Eich.
make some functions static in the savage drm driver
add Egberts 32/64 bit patch (its in kernel already...)
commit toggle switch to make Linux drm_handle_t unsigned int
revert reversion of a part of a patch from Jon, that I did last night while
add agp buffer token
add powerpc64 to the list of archs that this test doesn't make sense on
remove checks that make ppc64 not work properly... also fix ppc64 check..
add x86_64 to the list as well
We don't need to install all the internal headers files only drm.h and
allocating the PCIE table from GFP_DMA seems to stop it NMI'ing
add options for x86_64 and amd will small letters
check is the map containing the lock
convert ioctl flags to use flags instead of separate ints
fix makefile bug
Add GART in FB support for ati pcigart, and PCIE support for r300
back out change as gart is now in framebuffer for PCIE
make PCI Express work on 64-bit machines, thanks to Alex Deucher (agd5f)
add some pci express chips
update autogen from xserver tree
compat for pci_pretty_name
remove pci_find_class use for alpha
use linux kernel macros don't make our own
add __ATTR
hopefully fix server recycling on PCIE
Add Bens fix for radeon maps on ppc
fixup bens fix so it works..
fix pci overriding from userspace
Add support to turn writeback off via radeon module option
now I've lindented kernel it might be useful again
fix header this is now a c file
remove version not used anymore
Add X300 RV370
the old init path needs to set WAGP_ENABLE by default
dma access also needs some work
fix G550 cards
we don't use this stuff anymore .. remove 1k buffer from driver ..
cleanup / whitespace align with kernel
Fix cpu_to_le32 same as kernel not sure it is correct for ppc
cleanup ioctl/max_ioctl to use header file for extern symbols
whitespace align closer with kernel
convert to use __set_current_state, align some header includes
A bunch of create_proc_dir_entry() calls creating directories had crept in
remove extra spaces
realign whitespace with kernel
copy correct Kconfig
mixing breaks in r128 driver.. fixed in kernel
whitespace align with kernel
fix up radeon whitespace
i810 should be fine with i810.. no idea what this crack is ..
RV200 < R200
enable i915 32/64 bit ioctls
fixes from Egbert for 64bit
Fix from Hugh Dickins for consistent mapping
fixup xf86drm.c for new headers
Fix from SuSE for issue with context creation failure
add texrect support for r300
Lindent some via files
fix radeon memory mapping from Ben Herrenschmidt
step back one reported bugs against that patch
Add vblank support to i915 DRM..
add radeon card type get param so userspace can avoid walking PCI
Realign via driver with changes in Linux kernel (mainly whitespace)
complete fix for attribution
complete fix for historic attribution
some fixes from linux kernel
whitespace cleanup/aligment with kernel
align spacing with kernel
The radeon DRM wasn't passing sparse checking in the kernel, this fixes it
sparse cleanups
use common read/write routines
use drm_cards_limit instead of cards_limit
make max minor more rational
nothing from VIA or S3 in this
Fixup test for memory at end of memory space
fix card unload
major realigment of DRM CVS with kernel code, makes integration much easier
fix build wrong function call
add benh's memory management patch
add proper checking for bitblt multi
clear i915 interrupts sources on server exit
always enable and set master on pci device
fix brace placement
make some functions static from the kernel
missed a piece of benh patch
experimental PCI DMA fixes use proper Linux interfaces
use coherent memory for PCI allocations with COMP flag
set dma pages reserved
fix some use before NULL check
Fix bug I reintroduced
make some functions static in via driver
fix issue in mga from kernel
radeon fix up the PCI ids for new memory map like the kernel one.. not
make add context non master...
remove stupid init and exit flags..
coverity bugfix from the kernel
Fix from Benh for ppc r300 scratch
fixup GFP_COMP for older kernels and get_page/put_page for newer
add consts to radeon microcode.
rip out unneeded back compat code
switch drm to use Linux mutexes instead of semaphore.
remove incorrect exit marking on cleanup pci as this is called from other paths
drm: fixup whitespace and style for Linux kernel import
drm: remove extra whitespace from drm_mm.c
drm: whitespace cleanup in new files
drm: ati_pcigart cleanup
clean up script to create an lk drm directory
cleanup some whitespace issue and move compat code to compat header
i965 code and Linux coding style < 0
remove all TRUE/FALSE no need for this in the drm, use 0 or 1
remove all TRUE instances as well
drm: i810_dma.c: fix pointer arithmetic for 64-bit target
fixup some of the comments in drm_context.c
remove some DRM_ARRAY_SIZE from linux core code
remove local copies of pci domain/bus/slot/num
fix const pointer warnings with file_operations
add static function, and remove bad attributions
drm: lots of small cleanups and whitespace issues fixed up
drm: use radeon specific names for radeon flags
drm: put domain number back to 0, domain support is seriously fubar..
remove config.h from build no longer exists kbuild does it
libdrm: add support for server side functionality in libdrm
libdrm: add drmOpenOnce + drmCloseOnce to libdrm
David Dawes (42):
Initial revision
Import of XFree86 4.0.1b
first part of 4.0.1b merge
Import of XFree86 4.0.1d
merge with 4.0.1d
Import of XFree86 4.0.1f
Import of XFree86 4.0.1g
Import of XFree86 4.0.99.1
Initial XFree86 4.0.99.1 merge.
merge/build fix
Move #include <linux/config.h> inside the __linux__ #ifdef, so that it
- Move xf86ConfigDRI declaration from xf86Priv.h to xf86.h.
Import -f XFree86 4.0.99.2
First pass of XFree86 4.0.99.2 merge.
Use the linux version of xf86drm.c.
put back BUS_BASE
Import of XFree86 4.0.99.3
Initial merge for XFree86 4.0.99.3 import
finish struct stat type cleanup Fix a build dependencies for the tdfx drm
Import of XFree86 4.1.0
First pass of 4.1.0 merge.
Merge the multihead-1-0-0 branch into the trunk, with the exception of the
Import of XFree86 4.1.99.1
First part of XFree86 4.1.99.1 merge.
- Remove the rest of the unneeded client-side libraries.
Initial revision
Import of XFree86 4.2.0
First pass merge of XFree86 4.2.0 import.
First pass of mesa-4-0 branch merge into trunk.
- Fix various bugs in the DRI support for the i830 and i845G (Keith
enums should be ints
file i830_irq.c was initially added on branch mesa-4-0-4-branch.
Fix DRM module build on 2.5.41 and later kernels (tqueue -> workqueue).
Bring some drm module changes over from the XFree86 trunk:
Rework the Linux drm kernel module build to leverage off the standard
Make Config.in look more like a recent 2.4.x kernel version.
Add a Kconfig file as used in recent 2.5.x kernels.
break long line
Single/dual rasterizer quiescence patch for the glint/gamma DRI driver
Targets for building dristat and drmstat.
Support for building DRM module on OpenBSD (Wilbern Cobb, Matthieu Herrb).
DRM part of Radeon DRI suspend/resume support (Charl Botha).
David Nusinow (3):
Copy packaging over, update changelog, and remove old version
* Add myself to uploaders
Merge from unstable back to experimental
Doug Rabson (37):
file i810_drm.h was initially added on branch bsd-1-0-0-branch.
file mga_drm.h was initially added on branch bsd-1-0-0-branch.
file drawable.c was initially added on branch bsd-1-0-0-branch.
file auth.c was initially added on branch bsd-1-0-0-branch.
file bufs.c was initially added on branch bsd-1-0-0-branch.
file dma.c was initially added on branch bsd-1-0-0-branch.
file ioctl.c was initially added on branch bsd-1-0-0-branch.
file context.c was initially added on branch bsd-1-0-0-branch.
file sysctl.c was initially added on branch bsd-1-0-0-branch.
file lists.c was initially added on branch bsd-1-0-0-branch.
file proc.c was initially added on branch bsd-1-0-0-branch.
file init.c was initially added on branch bsd-1-0-0-branch.
file fops.c was initially added on branch bsd-1-0-0-branch.
file drmstat.c was initially added on branch bsd-1-0-0-branch.
file vm.c was initially added on branch bsd-1-0-0-branch.
file memory.c was initially added on branch bsd-1-0-0-branch.
file lock.c was initially added on branch bsd-1-0-0-branch.
file gamma_drv.c was initially added on branch bsd-1-0-0-branch.
file gamma_dma.c was initially added on branch bsd-1-0-0-branch.
file gamma_drv.h was initially added on branch bsd-1-0-0-branch.
file tdfx_drv.h was initially added on branch bsd-1-0-0-branch.
file Makefile was initially added on branch bsd-1-0-0-branch.
file tdfx_context.c was initially added on branch bsd-1-0-0-branch.
file tdfx_drv.c was initially added on branch bsd-1-0-0-branch.
file agpsupport.c was initially added on branch bsd-1-0-0-branch.
file r128_drm.h was initially added on branch bsd-1-0-0-branch.
Merged bsd-1-0-0
file ctxbitmap.c was initially added on branch bsd-1-0-1-branch.
file mga_drm.h was initially added on branch bsd-1-0-1-branch.
file mga_context.c was initially added on branch bsd-1-0-1-branch.
file mga_bufs.c was initially added on branch bsd-1-0-1-branch.
file mga_drv.c was initially added on branch bsd-1-0-1-branch.
file mga_dma.c was initially added on branch bsd-1-0-1-branch.
file Makefile was initially added on branch bsd-1-0-1-branch.
file mga_drv.h was initially added on branch bsd-1-0-1-branch.
file mga_state.c was initially added on branch bsd-1-0-1-branch.
Merged bsd-1-0-1
Erdi Chen (3):
Add PCI id entry for VIA CN400 (UnichromePro) chip. XFree86 bug: Reported
This patch adds three new ioctl's to the VIA Unichrome/Pro DRM driver:
Fix copy_from_user return value ignored warning at via_dma.c:168.
Eric Anholt (282):
file radeon_state.c was initially added on branch bsd-3-0-0-branch.
file radeon_drv.h was initially added on branch bsd-3-0-0-branch.
file radeon_cp.c was initially added on branch bsd-3-0-0-branch.
file radeon.h was initially added on branch bsd-3-0-0-branch.
file r128_state.c was initially added on branch bsd-3-0-0-branch.
file r128_drv.h was initially added on branch bsd-3-0-0-branch.
file r128_cce.c was initially added on branch bsd-3-0-0-branch.
file r128.h was initially added on branch bsd-3-0-0-branch.
file mga_warp.c was initially added on branch bsd-3-0-0-branch.
file mga_ucode.h was initially added on branch bsd-3-0-0-branch.
file mga_state.c was initially added on branch bsd-3-0-0-branch.
file mga_drv.h was initially added on branch bsd-3-0-0-branch.
file mga_dma.c was initially added on branch bsd-3-0-0-branch.
file mga.h was initially added on branch bsd-3-0-0-branch.
file gamma.h was initially added on branch bsd-3-0-0-branch.
file i810.h was initially added on branch bsd-3-0-0-branch.
file i810_drv.c was initially added on branch bsd-3-0-0-branch.
file i810_dma.c was initially added on branch bsd-3-0-0-branch.
file radeon_drv.c was initially added on branch bsd-3-0-0-branch.
file i810_drv.h was initially added on branch bsd-3-0-0-branch.
file i830_drv.h was initially added on branch bsd-3-0-0-branch.
file i830_drv.c was initially added on branch bsd-3-0-0-branch.
file i830.h was initially added on branch bsd-3-0-0-branch.
file i830_dma.c was initially added on branch bsd-3-0-0-branch.
file tdfx.h was initially added on branch bsd-3-0-0-branch.
file mga_drv.c was initially added on branch bsd-3-0-0-branch.
file r128_drv.c was initially added on branch bsd-3-0-0-branch.
file drm_os_netbsd.h was initially added on branch bsd-3-0-0-branch.
file r128_drm.h was initially added on branch bsd-3-0-0-branch.
file r128.h was initially added on branch bsd-3-0-0-branch.
file r128_cce.c was initially added on branch bsd-3-0-0-branch.
file r128_drv.h was initially added on branch bsd-3-0-0-branch.
file r128_state.c was initially added on branch bsd-3-0-0-branch.
file radeon_cp.c was initially added on branch bsd-3-0-0-branch.
file mga_drm.h was initially added on branch bsd-3-0-0-branch.
file radeon_drv.h was initially added on branch bsd-3-0-0-branch.
file mga_state.c was initially added on branch bsd-3-0-0-branch.
file mga.h was initially added on branch bsd-3-0-0-branch.
file mga_warp.c was initially added on branch bsd-3-0-0-branch.
file mga_ucode.h was initially added on branch bsd-3-0-0-branch.
file mga_drv.h was initially added on branch bsd-3-0-0-branch.
file mga_dma.c was initially added on branch bsd-3-0-0-branch.
file radeon_drm.h was initially added on branch bsd-3-0-0-branch.
file radeon.h was initially added on branch bsd-3-0-0-branch.
file radeon_state.c was initially added on branch bsd-3-0-0-branch.
Increase the linux-compatibility max ioctl.
Try to open the /dev/dri/cardX device once, then only if it fails check the
Remove drm_linux.h, move the two useful defines into drm_drv.h (the only
Include non-radeon modules in the build.
Remove drm_linux.h, it's no longer used.
Remove i8x0 files from the BSD side. These were not actually ported, and
Remove compat functions for FreeBSD for drivers which weren't available in
Remove this one too: it'll be used from the linux version (if ever)
Remove some extra symlinking for kernel module building that hasn't been
Fix one warning.
BSD vblank framework.
Fix up BSD irq handling.
Fix some of my silliness in DRM_WAIT_ON (not returning EINTR).
warnings cleanup
Make our set_bit and friends match linux better (cleans up some warnings)
Remove some noise being spit to the console.
Fix reversed test that broke vblank waiting on FreeBSD.
Catch up to -current.
s/udelay/DRM_UDELAY/
Use bus_alloc_resource/bus_release_resource more properly: save the rid
Kernel support for vblank syncing on Rage 128 and Matrox.
Add vblank signal code for BSD DRM. Untested so far, but working with a
Fix radeon for BSD.
Fix build on -current: Provide M_WAITOK define.
file mach64.h was initially added on branch mach64-0-0-6-branch.
file mach64_dma.c was initially added on branch mach64-0-0-6-branch.
file mach64_drv.h was initially added on branch mach64-0-0-6-branch.
file mach64_state.c was initially added on branch mach64-0-0-6-branch.
file mach64_drm.h was initially added on branch mach64-0-0-6-branch.
Merge from bsd-4-0-0-branch.
Fix build on NetBSD.
Remove a paste-o in DRM_SPINUNINIT and add a volatile to the compatibility
Update mmap handling for FreeBSD-current (Based on r1.3 of FreeBSD CVS).
Put cdevsw initialization in line with FreeBSD-current. (From r1.10,r1.11
Remove the vbl signal code because it's untested (and has lock issues on
Make dma_addr_t an unsigned long not a uint32, don't try to use memrange
Merge back from FreeBSD-current, adding FreeBSD ID tags to aid future
buf->filp is a pointer, so make printf format args be %p not %d
Add DRMFILE definitions and supply filp for BSD in the
Remove dead vma code and remove the unused devstate struct definition.
Spelling fixes in comments.
Whitespace and remove a dead commented line.
file drm_linux_list.h was initially added on branch mach64-0-0-6-branch.
file Makefile was initially added on branch mach64-0-0-6-branch.
file mach64_drv.c was initially added on branch mach64-0-0-6-branch.
Make DRM(read) and DRM(poll) stubs and remove DRM(write) and
Remove DRM_DMA_HISTOGRAM and associated code.
Move one definition to drm_drv.h and remove the rest of drm_init.h which
Remove a bunch of dead code and fix spelling of a couple of comments.
Remove the ioctl_count variable from the device. A reference is held to the
Move some common code from addbufs_<type> to addbufs. Make buf_alloc be
Remove more gamma DMA code. This isn't all of it, but it's a major portion.
Remove more gamma DMA infrastructure. Most of this code was copied straight
Clean up the style of the linux-compat code and use ioctl() directly rather
Clean up the DRM_COPY_TO_USER()ing of DRM(infobufs), making it more
Merge from FreeBSD-current.
Disable MTRRs on FreeBSD-stable. Without this, it hangs on boot in the MTRR
Replace the C atomic_cmpset_int compatibility function for -stable with the
Remove #if 0'ed code.
Move the memory functions with debugging info to drm_memory_debug.h, and
Missed files in the last commit: Remove memory debugging sysctl unless
Add PCI DMA memory functions and make addbufs_pci and associated code use
MFL: Don't install irq handler unless the driver has been initialized.
Remove the map argument from DRM_*MEMORYBARRIER. Not all of the uses of
Fix formatting of hw.dri sysctl.
Use real endian conversion functions.
Fix a typo: On takedown the mtrr operation is MEMRANGE_SET_REMOVE, not
Fix FreeBSD build after IRQ changes.
Add Rage 128 pageflipping support, defaults to off. DRM version bump to
Merge from FreeBSD r1.11: We have memset in the kernel, no need to define
Document the changes in interface version 2.5.
Document change in interface version 1.9.
Whitespace cleanup from the pageflipping commit.
Remove an unnecessary #define __NO_VERSION__
Make r128_do_wait_for_idle static, as it's only used in this file.
- Remove $FreeBSD$ tags as they weren't too useful and merges are now being
Fix the debug build.
This PCI header has been living in dev/pci/ for a while now.
Update radeon PCI IDs.
Add DRM(calloc), which is convenient, used by the new sis code, and takes
Port the SiS DRM to FreeBSD. This includes the ability for the DRM to
These files were missed in the SiS DRM commit.
Update to reflect renaming of SIS ioctls.
Fix for older -stable.
Correct format in debug printfs (free is a pointer, not an int).
Fix Kconfig for SiS DRM now that it doesn't require sisfb.
Whitespace cleanup.
Whitespace cleanup (spaces before tabs or instead of tabs).
MTRR issue with SMP and -stable seems to be resolved. Re-enable MTRRs on
Allow the DRM to attach to a "drmsub" device. This will be provided by the
Wrap sys/endian.h usage with __FreeBSD_version >= 480000.
Mostly whitespace cleanups and style(9) fixes focused on "if(" -> "if ("
Axe more old gamma DMA infrastructure.
Add an MIT-style copyright, assigned to myself, to these files. I think
Stylistic preparation for SMPng locking work: DRM_LOCK/DRM_UNLOCK have side
Some code cleanups done while working on locking. Reduces always-true
Debug printf format fix.
Try that again. It's a long.
- Converted Linux drivers to initialize DRM instances based on PCI IDs, not
- Move IRQ functions from drm_dma.h to new drm_irq.h and disentangle them
Fix probing on 2.5+ kernels, which require that drivers have .id_table set.
Clean up extra zeroing of dev->dma, and use calloc to take advantage of
- SMPng lock the DRM. This is only partial -- there are a few code paths
Clean up BSD MTRR handling. The NetBSD code is untested, but it's my best
Fix the possibility of sleeping with locks held in sysctls by copying the
Fix warning about static DRM(bufs_info) defined but not used in the
- Add DRM_GET_PRIV_WITH_RETURN macro. This can be used in shared code to
- Introduce a new ioctl, DRM_IOCTL_SET_VERSION. This ioctl allows the
The SiS300 pci id also covers the 305, so make the name reflect that.
Missed the sis.h header in Makefile.linux in the move to shared/.
Move to "old-style" probing as documented in linux/Documentation/pci.txt.
Move the REALLY_HAVE_AGP endif above the mtrr functions. Broke tdfx module.
Don't try to use dev->dma_lock unless dma is initialized (dev->dma != NULL)
Add a DRM_PUT_USER_UNCHECKED, which will be used by an upcoming radeon
Add i852/i855 PCI ID. Also fix whitespace in the other ID definitions.
Remove buf_alloc which is unused since the locking commit.
Repo-copy linux/drm/kernel/drm.h to shared/drm/kernel/drm.h and use it on
__linux__ is spelled with a lowercase 'l'
Args for the BSD DRM_PUT_USER_UNCHECKED were swapped.
Use int64_t instead of s64 -- fixes FreeBSD compile, works on linux.
- Tie the DRM to a specific device: setunique no longer succeeds when given
Remove unused variable.
Return EBUSY when attempting to addmap a DRM_SHM area with a lock in it if
Fix a locking nit, and add asserts in some things that should be called
Add a collection of Radeon and R128 PCI IDs, including the IGP chipsets.
Don't ioremap the framebuffer area. The ioremapped area wasn't used by
Fix some misuse of NULL where 0 is intended.
drm_hw_lock_t is now defined in drm.h, remove from here.
No /bin/bash on BSD.
Remove after repocopy to Makefile, now that we don't have to worry about
Catch up with sis's DRM tag change.
- Link in shared files as necessary and clean them up.
Warning fixes.
Add .cvsignore files.
Add mach64 DRM module for BSD (untested, but compiles).
Add missing DRM_ERR()s.
Merge from FreeBSD-current. Mostly 64-bit cleanliness fixes, but a few
MFC as of 20040705: dev_t -> struct cdev * change.
Fix module loading on alpha by not referencing MTRR symbols on
Enable MTRR usage on AMD64, and use DELAY() instead of rolling our own
Fix apparent copy'n'paste-o of the card attributes commit that broke the
Hopefully proper fix for corrupted driver name in memcontrol list.
Add a "dev" argument to DRIVER_CTX_[CD]TOR. This will be used in an
Remove unused pcigart/sg header stuff from i915 driver.
Add *.flags to cvsignore.
Commit WIP of BSD conversion to core model. Compiles for r128, radeon, but
Remove the vestiges of the memory-debug code.
Commit first pieces of port to OpenBSD, done by Martin Lexa (martin at
Add file missed in last commit: Commit first pieces of port to OpenBSD,
Hook the debug output up to a sysctl, so you can choose to enable at
Connect up r128_ioctls in driver config.
Move the lock/unlock ioctls to a more logical place, in drm_lock.c.
Get r128 basically working: Hook up the driver's dma ioctl, use the proper
Remove some core stuff that ended up being unnecessary.
Convert more drivers for bsd-core, moving the ioctl definitions to shared
Add the drm Makefile and update .cvsignores.
Now that the memory debug code is gone, and all 3 BSDs have M_ZERO, stop
Don't link in files which no longer exist.
Refine the locking of the DRM. Most significant is covering the driver
Use SYSCTL_ADD_OID macro instead of calling function directly.
Add detection of whether the device is AGP by walking the capabilities
Cast user data to correct type in radeon_surface_free's copyin.
Declare r128_do_cleanup_pageflip static since it's only used here and its
- Implement drm_initmap, and extend it with the resource number to help
Remove some annoying trailing whitespace.
Restore a debugging message to DRM_DEBUG instead of DRM_ERROR.
Close a race which could allow for privilege escalation by users with DRI
Add the first bits necessary for a port of savage to FreeBSD. More to
Fix bad copy'n'pastage of copyrights -- don't disclaim anything for VA
Initialize kbuf to NULL to quiet the compiler about uninit variables
Use the proper API to get PCI vendor/device number for a dev.
Use fuword32 for DRM_GET_USER_UNCHECKED when available. May help on 64-bit
Merge r1.26 from FreeBSD: Now that mem(4) is a kernel module, we need to
Fix a couple of gcc warnings in DRM_INFOs.
Fix build on FreeBSD-current, thanks to jhb@.
Use /*- to begin license blocks in BSD code to reduce diffs against FreeBSD
Fix a panic on startup with non-initmapping drivers by assigning the
Use msleep instead of tsleep to drop the DRM device lock and avoid a
Convert BSD code to mostly use bus_dma, the dma abstraction for dealing
Clean up some const qualifier cast warnings.
Convert NOMAN to the new preferred spelling NO_MAN to quiet warnings.
Bugzilla #3217: The size field in the new dma handle structure was
Bugzilla #3217: Create a new __drm_pci_free which is used internally in
Add a few more bits of Tonnerre's NetBSD port (Still need to deal with the
- Remove drm_initmap and replace its usage with drm_addmap. This reduces
Get the BSD DRM compiling again after MGA and mapping changes. Note that
Add .cvsignore file.
Add latest r300 support from r300.sf.net CVS. Patch submitted by volodya,
Link in the savage files from shared-core.
Correct a couple of descriptions of files in comments (were just
Fix issues with buffer aging when more than one dma buffer is discarded in
Add a debugging error message from testing new MGA code on BSD.
Fix the MGA driver on BSD by passing in the proper chipset flags to the
Whitespace fixups.
Rename the driver hooks in the DRM to something a little more
Fix build after linux-side checkin of master/root-only split. Still only
Port the VIA DRM to FreeBSD. Original patch by Jake, with some cleanup by
Remove the long-stale BSD non-core bits.
Include appropriate CFLAGS to find X headers, needed to build libdrm.
Add missing .cvsignore files for autotoolery.
[1] Fix BSD DRM for the nonroot changes. [2] Don't attempt to acquire the
Fix a lack of parenthesis in macro usage that showed up with INVARIANTS
Previously, drm_get_resource_start() and drm_get_resource_len() would
- Don't try to allocate mappings of less than a PAGE_SIZE in MGA DMA code.
Fix the spelling of DRM_AUTH so that the bsd core stuff builds again. Next
Re-disable the via driver while it's broken on BSD.
Correct a recursion on non-recursive mutex in drm_addmap from radeon's
Catch FreeBSD up to the pcie gart changes. Required minor modification to
Fix FreeBSD DRM for latest MGA changes to agp support, which cleans things
Correct a LOR related to the PCI resource allocations by simply moving the
Correct a LOR on FreeBSD by allocating a temporary buffer and doing a
Initial port of savage to FreeBSD for the AGP and !ShadowStatus case. Adds
Correct another LOR issue with resource allocation. This leaves the
Correct a LOR issue on FreeBSD by allocating temporary space and doing a
Fix breakage from the move of driver ioctl externs to header files.
Fix Savage DRI without ShadowStatus NO by allowing the shadow area to be
Fix compiling and reenable build of i915 driver.
Fix AGP support guessing: Implement the same bridge checking in the MGA
Assert an MIT copyright on sis_drm.h, since one was lacking and I created
FreeBSD PR kern/85479: Restore the enabling of debugging by default by the
Finish the last bits necessary to get the i915 DRM port working on FreeBSD
Add RV410 X700PRO PCI ID
Remove driver.device_is_agp code duplicated in drm_device_find_capability,
Merge patch from jhb to catch up with FreeBSD-current vgapci master device
Initialize sv.drm_dd_minor to "don't care" along with dd_major, to appease
Compile fixes for FreeBSD.
Revert a change that accidentally went in with whitespace changes from
Use __LP64__ instead of checking the linux-specific BITS_PER_LONG.
Reorder the DRM_*_AGP enum to match linux's numbers (oops). Fixes i915
Err, use "ifndef" rather than "if !", to avoid compiler warning.
Add the workaround that's in the kernel to suppress GCC's warning about
Add the bits for vblank support on FreeBSD, which most importantly avoids
Set entry->virtual for sg maps, fixing ATI PCI/PCIE GART support.
Add a typedef for u64.
Put the PCI device/vendor id in the drm_device_t.
Use the DRM_INIT_WAITQUEUE argument (needed on Linux) to avoid a warning.
Add drm_u64_t typedef on non-linux to fix libdrm build.
Felix Kuehling (22):
Conditionally add definition of list_for_each_entry_safe for kernel
Merged the Savage DRM driver from the savage-2-0-0-branch into the trunk.
Use DO_MUNMAP_4_ARGS macro defined in Makefile.linux (Steve Holland).
Create permanent maps of framebuffer, aperture and MMIO registers. Added
Allow drivers to override reclaim_buffers in an OS-independent way by
Added a new DRM map type _DRM_CONSISTENT for consistent PCI memory. It uses
Completeley rewritten Savage DRM which can be considered secure (modulo
Fixed the DMA buffer age test. Should fix occasional "soft" lockups. Bumped
3D scissor regs are now managed by the DRM to iterate over clip rects
Fixed off-by-one error in savage_bci_wait_fifo_shadow.
Improved workaround for Savage3D DMA lockup to emit NOPs only before the
Only try to find the agp_buffer_map if dma_type is AGP. This is all that's
Use virt_to_page instead of vmalloc_to_page in drm_do_vm_shm_nopage for
Setup MTRRs for frame buffer and aperture manually on Savage3D and
Fixed a bug that prevented the driver from ever emitting triangle strips or
Corrected some confusion of vb_stride and vtx_size
Removed one bogus Savage3D PCI ID. Corrected another one. Restored
Use wrap counter to extend 16-bit hardware event tags to 32-bit logical
Added support for command DMA on Savage4-based hardware. Unfortunately
Tracked down random lockups related to command DMA that occurred in Quake3
Command DMA optimizations:
Savage doesn't require AGP any more. Enable build even without CONFIG_AGP.
Felix Kuhling (1):
bug 5942: add stubs for drm_mtrr_add/del for non-MTRR configured linux
Felix Kühling (2):
drm_rmdraw: Declare id and idx as signed so testing for < 0 works as intended.
drm_rmdraw: Declare id and idx as signed so testing for < 0 works as intended.
Gareth Hughes (76):
Fixed WARP pipe hangs in Quake 3. Fullscreen or window at (0,0) seems to
Merging the ati-4-1-0 branch onto the trunk.
Update register definitions to match r128_reg.h
Fix for agpgart module detection. If the kernel module has been configured
Sync with 2.4.0-test8-pre5 kernel.
Merge of tdfx branch undid the changes from the 2.4.0-test8-pre5 kernel
Sync with 2.4.0-test8 kernel.
file r128_cce.c was initially added on branch ati-4-1-1-branch.
file r128_state.c was initially added on branch ati-4-1-1-branch.
Sync with 2.4.0-test9-pre4 kernel.
Correct sync with 2.4.0-test9-pre4 kernel.
file mach64_context.c was initially added on branch mach64-0-0-0-branch.
file mach64_bufs.c was initially added on branch mach64-0-0-0-branch.
file mach64_drv.c was initially added on branch mach64-0-0-0-branch.
file mach64_drv.h was initially added on branch mach64-0-0-1-branch.
file driver_tmp.h was initially added on branch mach64-0-0-1-branch.
file context_tmp.h was initially added on branch mach64-0-0-1-branch.
Merged ati-4-1-1-branch into trunk.
file mach64_drm.h was initially added on branch mach64-0-0-1-branch.
file mach64_state.c was initially added on branch mach64-0-0-1-branch.
file mach64_dma.c was initially added on branch mach64-0-0-1-branch.
- Fix nasty depth span bug. Drawable offset was not being added to pixel
file bufs_tmp.h was initially added on branch mach64-0-0-1-branch.
Update date information.
Remove debug messages.
- Misc cleanups.
file drm_lists.h was initially added on branch mga-1-0-0-branch.
file drm_dma.h was initially added on branch mga-1-0-0-branch.
file drm_auth.h was initially added on branch mga-1-0-0-branch.
file drm_agpsupport.h was initially added on branch mga-1-0-0-branch.
file drm_drawable.h was initially added on branch mga-1-0-0-branch.
file drm_context.h was initially added on branch mga-1-0-0-branch.
file drm_drv.h was initially added on branch mga-1-0-0-branch.
file drm_ioctl.h was initially added on branch mga-1-0-0-branch.
file drm_init.h was initially added on branch mga-1-0-0-branch.
file drm_fops.h was initially added on branch mga-1-0-0-branch.
file drm_proc.h was initially added on branch mga-1-0-0-branch.
file drm_bufs.h was initially added on branch mga-1-0-0-branch.
file drm_memory.h was initially added on branch mga-1-0-0-branch.
file drm_lock.h was initially added on branch mga-1-0-0-branch.
file drm_vm.h was initially added on branch mga-1-0-0-branch.
file mga.h was initially added on branch mga-1-0-0-branch.
file mga_drm.c was initially added on branch mga-1-0-0-branch.
file mga_ucode.h was initially added on branch mga-1-0-0-branch.
file mga_warp.c was initially added on branch mga-1-0-0-branch.
Client-side updates for drmRadeonClear(...), fixes color/depth buffer
Fix depth clears properly this time. Update all instances of
Corresponding sync with PCI GART updates.
file tdfx_drm.c was initially added on branch mga-1-0-0-branch.
file tdfx.h was initially added on branch mga-1-0-0-branch.
file r128_drm.c was initially added on branch mga-1-0-0-branch.
file r128.h was initially added on branch mga-1-0-0-branch.
file radeon.h was initially added on branch mga-1-0-0-branch.
file i810.h was initially added on branch mga-1-0-0-branch.
Merge mga-1-0-0-branch into trunk.
- Fix up merge.
- Clean up the way customization of the templates is done.
Disable DRIVER_DMA_QUIESCENT() in r128, radeon drivers. Prevents deadlock
Please try this if you experience lockups...
Make the hardware bug workaround configurable. That way, people with newer
Merge tdfx-3-1-0 branch.
Change error message to debug message when client dies while holding the
Fix ring space calculations, tests. Based on patch by Bruce Stockwell.
Try that again...
Rest of MGA dualhead patch.
__REALLY_HAVE_MTRR, vmalloc_32 fixes from Jeff Wiedemeier.
Try this...
Reply to: