- 11 Feb, 2011 2 commits
-
-
Jesse Barnes authored
Disable any PCH ports associated with a pipe when disabling it. This should prevent transcoder disable failures due to ports still being on. Signed-off-by:
Jesse Barnes <jbarnes@virtuousgeek.org> [ickle: introduce *_PIPE_ENABLED() macro] Signed-off-by:
Chris Wilson <chris@chris-wilson.co.uk>
-
Chris Wilson authored
... so we handle that for i915_gem_fault() in the same manner as ERESTARTSYS, or we send a SIGBUS to the faulting application. Signed-off-by:
-
- 07 Feb, 2011 10 commits
-
-
Jesse Barnes authored
We had some conversions over to the _PIPE macros, but didn't get everything. So hide the per-pipe regs with an _ (still used in a few places for legacy) and add a few _PIPE based macros, then make sure everyone uses them. [update: remove usage of non-existent no-op macro] [update 2: keep modesetting suspend/resume code, update to new reg names] Signed-off-by:
-
Chris Wilson authored
Removes some superfluous fluff from tracing... Signed-off-by: -
Chris Wilson authored
A lot of minor tweaks to fix the tracepoints, improve the outputting for ftrace, and to generally make the tracepoints useful again. It is a start and enough to begin identifying performance issues and gaps in our coverage. Signed-off-by: -
Chris Wilson authored
By returning EAGAIN upon a wedged GPU before attempting to wait, we would hit an infinite loop of repeating operation without ever progressing. Instead this needs to be EIO so that userspace knows that the GPU is truly wedged and not in the process of error recovery. Similarly, we need to handle the error recovery during i915_gem_fault. Signed-off-by: -
Jesse Barnes authored
The PCH can drive several reference clocks simultaneously, and needs to with multiple display configurations. So we can't just clobber the existing state everytime we set a mode, we need to take into account what the other CRTCs are doing at the time. Doing so fixes an issue where you'd lose the LVDS display at boot if you had an LVDS+DP config. [updated: init bools and check CRTC status correctly] Signed-off-by:
-
Jesse Barnes authored
When a transcoder is disabled, any ports pointing at it should also be disabled. If they're not, we may fail to disable the transcoder, leading to blank displays. Signed-off-by:
-
Jesse Barnes authored
These bits have a different meaning on ILK+, where planes are hardwired to pipes. Fixing this avoid some spurious assertion failures. Signed-off-by:
-
Eric Anholt authored
The specs say to do so. Signed-off-by:
Eric Anholt <eric@anholt.net> Signed-off-by:
-
Chris Wilson authored
For consistency and segregation from the noisy DRM_DEBUG(). Signed-off-by: -
Chris Wilson authored
Signed-off-by:
-
- 01 Feb, 2011 5 commits
-
-
Chris Wilson authored
With the recent SDVO fix, this is working on all the machines I have to hand - except for an 845G. Signed-off-by: -
Chris Wilson authored
Merge resume fixes from -fixes and an important fix for SDVO GMBUS.
-
Chris Wilson authored
Fortunately unreachable. For Crestline, the watermarks must always be programmed to 8... Signed-off-by: -
Chris Wilson authored
Judging by comments in the BIOS, if the SDVO LVDS option h40 is enabled, then we are supposed to query the real panel type via Int15. We don't do this and so for the Sony Vaio VGC-JS210J which has otherwise default values, we choose the wrong mode. This patch adds a driver option, i915.vbt_sdvo_panel_type, which can be used to override the value in the VBT. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=33691Signed-off-by:
-
Chris Wilson authored
Signed-off-by:
-
- 31 Jan, 2011 4 commits
-
-
Chris Wilson authored
Hugh Dickins found that characters in xterm were going missing and oft delayed. Being the curious type, he managed to associate this with the new high-precision vblank patches; disabling these he found, restored the orderliness of his characters. The oddness begins when one realised that Hugh was not using vblanks at all on his system (fvwm and some xterms). Instead, all he had to go on were warning of a pipe underrun, curiously enough at around 60Hz. He poked and found that in addition to the underrun warning, the hardware was flagging the start of a new frame, a vblank, which in turn was kicking off the pending vblank processing code. There is little we can do for the underruns on Hugh's machine, a Crestline [965GM], which must have its FIFO watermarks set to 8. However, we do not need to process the vblank if we know that they are disabled... Reported-by:
Hugh Dickins <hughd@google.com> Signed-off-by:
-
Chris Wilson authored
Signed-off-by: -
Chris Wilson authored
By setting the FB of a CRTC to NULL, we are turning off the CRTC (and so disable the unused encoders and connectors). As such we can simplify the later tests by making sure the set->mode is NULL. Setting the num_connectors to zero means that we do not need to loop over the unused connectors. All current usage appears correct, this only builds additional defense into the routine. References: https://bugzilla.kernel.org/show_bug.cgi?id=27722Tested-by:
Takashi Iwai <tiwai@suse.de> Signed-off-by:
-
Chris Wilson authored
In drm_crtc_helper_set_config, we call drm_crtc_helper_set_mode which may return early and do no operation if the crtc is to be disabled. In this case we merrily swap to the new fb, discarding the old_fb believing that it has been cleaned up. However, due to the early return, the old_fb was not presented to the backend for correct reaping, and nor was the new one - which is about to be reaped via the drm_helper_disable_unused_functions(), leading to incorrect refcounting of the pinned objects. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=27722 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=29857 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=29230Tested-by:
-
- 28 Jan, 2011 1 commit
-
-
Chris Wilson authored
Signed-off-by:
-
- 27 Jan, 2011 2 commits
-
-
Chris Wilson authored
Instead of reporting EIO upfront in the entrance of an ioctl that may or may not attempt to use the GPU, defer the actual detection of an invalid ioctl to when we issue a GPU instruction. This allows us to continue to use bo in video memory (via pread/pwrite and mmap) after the GPU has hung. Signed-off-by: -
Chris Wilson authored
Signed-off-by:
-
- 26 Jan, 2011 2 commits
-
-
Chris Wilson authored
Signed-off-by: -
Chris Wilson authored
Signed-off-by:
-
- 25 Jan, 2011 9 commits
-
-
Chris Wilson authored
Based on a patch by Takashi Iwai. Reported-by:
Matthias Hopf <mat@mshopf.de> Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=27272Tested-by:
-
Chris Wilson authored
Upon resume, like after a cold boot, we need to forcibly probe the analog connector and cannot rely on the hotplug status. Based on a patch by Takashi Iwai. Reported-by:
Stefan Dirsch <sndirsch@suse.de> Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=26952Tested-by:
-
Chris Wilson authored
Call drm_mode_config_reset() after an invalidation event to restore any cached state to unknown. Tested-by:
-
Chris Wilson authored
Iterate over the attached CRTCs, encoders and connectors and call the supplied reset vfunc in order to reset any cached state back to unknown. Useful after an invalidation event such as a GPU reset or resuming. Tested-by:
-
Chris Wilson authored
We were not pausing after detecting the response was pending and so did not allow the hardware sufficient time to complete before aborting. This lead to transient failures whilst probing SDVO devices. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=30235Reported-by:
Knut Petersen <Knut_Petersen@t-online.de> Signed-off-by:
-
Jesse Barnes authored
Rather than power cycling the panel when there are no bits to display, use the VDD AUX bit to power the panel up just enough for DP AUX transactions to work. This prevents a bit of unnecessary ugliness as mode sets occur on the panel. Signed-off-by:
-
Jesse Barnes authored
Now that we're doing the right thing elsewhere, these are no longer necessary. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=31114Signed-off-by:
-
Chris Wilson authored
Move the plane->mode config to the point of use rather than repeatedly querying the same information. Signed-off-by: -
Chris Wilson authored
Signed-off-by:
-
- 24 Jan, 2011 5 commits
-
-
Chris Wilson authored
Merge important suspend and resume regression fixes and resolve the small conflict. Conflicts: drivers/gpu/drm/i915/i915_dma.c
-
Chris Wilson authored
error_bo and pinned_bo could be used uninitialised if there were no active buffers. Caught by kmemcheck. Signed-off-by: -
Michael Karcher authored
The opregion is a shared memory region between ACPI and the graphics driver. As the ACPI mapping has been changed to cachable in commit 6d5bbf00, mapping the intel opregion non-cachable now fails. As no bus-master hardware is involved in the opregion, cachable map should do no harm. Tested on a Fujitsu Lifebook P8010. Signed-off-by:
Michael Karcher <kernel@mkarcher.dialup.fu-berlin.de> [ickle: convert to acpi_os_ioremap for consistency] Signed-off-by:
-
Chris Wilson authored
Merge with Linus to resolve conflicting fixes for the reusing the stale HEAD value during intel_ring_wait(). Conflicts: drivers/gpu/drm/i915/intel_ringbuffer.c
-
Chris Wilson authored
If the driver calls into the kernel to wait for a breadcrumb to pass, but hasn't enabled interrupts, fallback to polling the breadcrumb value. Reported-by:
Chris Clayton <chris2553@googlemail.com> Tested-by:
-
