When i915_gem_retire_request has a flush which matches an object write
domain, clear the write domain. This will move the object to the inactive
list rather than the flushing list, avoiding trouble with objects left stuck
on the flushing list.
In i915_gem_object_wait_rendering, if the object write domain is being
written by the GPU, the appropriate flushing commands are written to the
device and an additional request queued to mark that flush. Finally, the
function blocks on that new request.
The bug was that the write_domain in the object was cleared before the
function blocked.
If the wait is interrupted by a signal, the flushing commands may still be
pending. With the current write_domain information lost, the restarted
syscall will drop right through the write_domain test as that value was
lost, and so the function will not block at all. Oops.
Fixed by simply moving the write_domain clear until after the wait_request
succeeds. Note that the restarted system call will generate an additional
flush sequence and request, but that should be 'harmless', aside from a
slight performance impact.
Someday we'll track flushing more accurately and clear write_domains more
efficiently, but for now, this should suffice.
This bug was discovered in the 2d gem development by running x11perf
-copypixwin500 and noticing that the window got cleared accidentally.
Main fix is an oops that was triggered by the gtt pwrite path when we don't
have the gtt initialized. Also, settle on -EBADF for "bad object handle",
and -EINVAL for "reading/writing beyond object boundary".
This is around 3x or so speedup, since we would read wide rows at a time, and
clflush each tile 8 times as a result. We'll want code related to this anyway
when we do fault-based per-page clflushing for sw fallbacks.
this lets us debug the X server through xkb startup.
Not sure what the correct answer is, probably X needs to drop
the lock when execing stuff, with input hotplug it can get
xkb stuff at any time I believe.
This increases overhead for the large-readpixels case due to the repeated
page cache accessing, but greatly reduces overhead for the small-readpixels
case.
Caused drm_update_vblank_count() not to do its thing when called from
drm_modeset_ctl() -> drm_vblank_get().
The vblank functionality no longer needs to be suspended during a modeset, so
rename the field to vblank_inmodeset.
In my last push I forgot to convert users of drm_update_vblank_count
over to drm_vblank_get/put, since that's where any interrupt off->on
update accounting is done now. Since the modeset ioctl did something
similar (an open coded update of the counter) convert it over to using
get/put too, which saves us from having to deal with every combination
of interrupt off & on between calls.
The current code uses the hw vblank counter exclusively, which can lead
to wakeups during the active period rather than during the vblank period
if the hw counter counts displayed frames rather than vblank periods.
This change coverts the code over to using the counter while interrupts
are enabled, fixing that issue. It also includes a couple of related
changes: one to not enable the new enable/disable behavior until the
modeset ioctl is called (to preserve old client behavior) and another to
account for lost events due to mode setting with the new counter scheme.
BSD will require similar changes to its drm_irq.c code, but they should
be straightforward.
modifications to make it work correctly on my test hardware (altered the
backlight write function, made it enable the legacy backlight controller
interrupts on mobile hardware, sorted the interrupt function so we don't
get an excessive number of vblank interrupts). This lets the backlight
keys on my T61 work properly, though there's a 750msec or so delay
between the request and the brightness actually changing - this sounds
awfully like the hardware spinning waiting for a status flag to become
ready, but as far as I can tell they're all set correctly. If anyone can
figure out what's wrong here, it'd be nice to know.
Some of the functions are still stubs and just tell the hardware that
the request was successful. These can be filled in as kernel modesetting
gets integrated. I think it's worth getting this in anyway, since it's
required for backlight control to work properly on some new platforms.
Signed-off-by: Matthew Garrett <mjg@redhat.com>
One of our systems has been returning 0xffffffff from all MCHBAR reads, which
means we'll need to figure out why, or add an alternate detection method.
Various chips have exciting interactions between the CPU and the GPU's
different ways of accessing interleaved memory, so we need some kernel
assistance in determining how it works.
Only fully tested on GM965 so far.
Dave Airlie
Tomas Carnecky
Stephane Marchesin
Maarten Maathuis
Thomas Hellstrom
Dave Airlie
Eric Anholt
Keith Packard
Stefan Becker
Michel Dänzer
Jesse Barnes
Ian Romanick
Hong Liu
martin capitanio