| Age | Commit message (Collapse) | Author |
|---|
|
|
|
Sauerbeck).
|
|
|
|
hopefully confuse bug triagers a little less.
|
|
various problems up to system hangs on X server startup have been
reported with that and DRI enabled.
|
|
#5273 <https://bugs.freedesktop.org/attachment.cgi?id=5273>:
Add cases for FOURCC_RGBA32 and FOURCC_RGB24. This prevents clients from
receiving incorrect data sizes with these formats. (Jan Schmidt)
|
|
the problem in the comment only concerned pre-production cards that Hui
tested apparently. Also set back the default value to what it was
initially, fixes some more cards that don't have a proper table (Mac
Mini for example)
|
|
#5536 <https://bugs.freedesktop.org/attachment.cgi?id=5536>:
Change HostDataBlit interface to take dst_offset_pitch and coordinates
instead of just a destination pointer, as the latter is not sufficient
with tiling. Also, use HW clipping to avoid overwriting destination
data outside of the specified width.
Adapt to new HostDataBlit interface. This fixes corruption with
UploadToScreen to the front buffer (from exaPutImage).
|
|
#4988 <https://bugs.freedesktop.org/attachment.cgi?id=4988>: Fix Option
"DisplayPriority" "HIGH" for R300 family.
|
|
the next generation.
|
|
switching to console with 3D clients running.
|
|
is double-buffered by default, which is desirable in a number of
situations, in particular when running a GLX compositing manager with
AIGLX.
|
|
with RB3D_DSTCACHE_*. The RB2D versions are read-only mirrors of the
RB3D. This is masked when DRI is enabled because the DRM uses the right
registers in its flushing.
|
|
aperture size. Otherwise it tries to overmap things if they are up near
the end of memory.
|
|
necessary for testing upcoming general component alpha acceleration
assistance in EXA. Tested with rendercheck on an r100.
|
|
instead of MergedFB. This was broken due to bogus manipulations of a
register between the 2 entities. This fix is a big "violent" in that it
always enables the analog output regardless of what is plugged. It
should work though (at the expense of some additional power
consumption). A proper fix should be easy to do once Alex gets his
output mapping rework in.
|
|
|
|
suspend/resume work a lot better on PCIE cards.
|
|
redirected video.
|
|
- Add ConstantDPI option to force a particlar dpi across mode changes Both
based on Thomas Winischhofer's sis code.
|
|
RADEONRestoreMemMapRegisters() must not clear CRTC_OFFSET_CNTL and
friends unconditionally but only when there is a map change or it will
clear them after the DRI updates the AGP location, thus turning tiling
back off on the CRTC. Also remove a duplicate call to
RADEONInitCommonRegisters().
|
|
|
|
instead of after. I can't find any good reason why we did it backward
until now and it definitely causes problems as the workaround for VT
switch in RADEONRestoreCommonRegisters() will break output to the
secondary head in various situations. Also does an unrelated minor
update to a comment.
|
|
|
|
symbols directly. The xf86* versions aren't supposed to be used
directly.
Drop libc wrapper; don't include xf86_ansic.h and add includes now missing.
|
|
CRTCs and try to make that code more reliable. Doesn't disable the
CRTCs beforehand.
|
|
right for requesting drm versions, replace the use of some numbers with
the respective macro defines in radeon_video.c, and add some more macro
defines. None of that really matters.
|
|
RV410 to the list of new generation cards for the memory map setup and
fix a bogus message.
|
|
MergedFB). Fixes Xinerama for me.
|
|
experiencing with the driver when setting the memory map. Some of the
magic delays are a bit dodgy but they seem to work, I suppose I can't
do better now without help from ATI. Also removed some really too noisy
debug messages.
|
|
the normal CloseScreen callback chain at a spot that matches the call
to DRIFinishScreenInit in order to unwrap at the right time.
|
|
older server with older exa is installed.
The way in which this problem was treated by the actually related
developers explains a lot about this driver.
|
|
old radeon
|
|
detection. Now split PreInitConfig into PreInitChipType and PreInitVRAM
so that PreInitDRI can be done just in between. Restores proper AGPx4
functionality (when enabled)
|
|
in atipciids.h so we can update the ATI driver independently of the
server when new chips are added
|
|
info->depthMoves, since they are only defined if building with DRI.
|
|
for textures).
|
|
module version early, in order to properly limit the framebuffer size
when a module that doesn't support the new map is loaded. That will
improve backward compatibility. I also made the minimum required kernel
module version 1.3 which hopefully shouldn't be a problem for anybody,
and cleaned up the ordering in which we test for DRM version to enable
features
|
|
Bugzilla #5766 <https://bugs.freedesktop.org/show_bug.cgi?id=5766> Patch
#4636 <https://bugs.freedesktop.org/attachment.cgi?id=4636>
- Acknowledge that RN50 only has one CRTC, and use this to distinguish it
from RV100.
- Fix detection of RN50 memory type and bus width.
- Model RN50 memory bandwidth limits by capping the pixel clock range based
on memory clock, bpp and memory bus width. (ATI Technologies Inc.)
|
|
moving all the accel and card members into the driver ptr, filling in
the exa_major/ exa_minor fields, and always using LoadSubModule so we
can check the module version up front. Only tested on ATI.
|
|
OpenGL textures with EXA.
|
|
the function works in Kb not bytes... Ooops.
|
|
and also limit the CPU accessible memory to the size of the PCI BAR
size (yeah, it was passed to the driver, I just got blind for a while
it seems)
|
|
aligned to its own size. Also add back limit of 128Mb of mapped fb
space for now until I fully sort out what's up with dual function
cards. I would really need to access the PCI BAR sizes here but that
information seem to be missing from the PCI info passed to the driver.
|
|
ati wrapper too (probably not required for it to work as everything
should default to radeon anyway there).
|
|
before XAAInit(), which breaks ShadowFB.
|
|
entries which probably don't really exist (cancelled cards and such),
leave them as-is. Fix the name of some entries, mostly based on the
.inf file of the newest catalyst driver. Use own family id for rv410
and rs400, though there is no different code (yet?).
|
|
detects available & accessible video memory and initializes the card
internal memory map. Along the way, I re-ordered some of the stuffs in
pre-init in order to remove some dodgy codepath, fixed various init,
exit and vt switch issues, and overall made the driver more robust. The
full memory map fix requires an updated DRM. The X driver no longer
tries to position the AGP aperture, this becomes entirely DRM
responsibility.
|
|
addresses are setup, thus causing the card to try to bus master over
the system bus to some stale location (usually 0 but could depend on
whatever was used before X). Remove that useless init bit, and always
setup some good enough engine source & destination pointers in case we
have some other spurrious engine activity going on.
|
|
<john@deater.net>
|
