GitLab

Andi Kleen narrowed his GPU hangs on his Sugar Bay (SNB desktop) rev 09
down to the use of GPU semaphores, and we already know that they appear
broken up to Huron River (mobile) rev 08. (I'm optimistic that disabling
GPU semaphores is simply hiding another bug by the latency and
side-effects of the additional device interaction it introduces...)

However, use of semaphores is a massive performance improvement... Only
as long as the system remains stable. Enable at your peril.
Reported-by: Andi Kleen <andi-fd@firstfloor.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=33921

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Whilst the GT is powered down (rc6), writes to MMADDR are placed in a
FIFO by the System Agent. This is a limited resource, only 64 entries, of
which 20 are reserved for Display and PCH writes, and so we must take
care not to queue up too many writes. To avoid this, there is counter
which we can poll to ensure there are sufficient free entries in the
fifo.

"Issuing a write to a full FIFO is not supported; at worst it could
result in corruption or a system hang."
Reported-and-Tested-by: Matt Turner <mattst88@gmail.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=34056

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

The automatic powersaving feature is once again causing havoc, with 100%
reliable hangs on boot and resume on affected machines.
Reported-by: Francesco Allertsen <fallertsen@gmail.com>
Reported-by: Gui Rui <chaos.proton@gmail.com>
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=28582

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Early chipsets (gen2/3) used function 1 as a placeholder for multi-head.
We used to ignore these since they were not assigned to
PCI_CLASS_DISPLAY_VGA. However with 934f992c

 we attempt to bind to all
Intel PCI_CLASS_DISPLAY devices (and functions) to work in multi-gpu
systems. This fails hard on gen2/3.
Reported-by: Ferenc Wágner <wferi@niif.hu>
Tested-by: Ferenc Wágner <wferi@niif.hu>
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=28012

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org

Call drm_mode_config_reset() after an invalidation event to restore any
cached state to unknown.
Tested-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

We only have sufficient information for accurate (sub-frame) timestamping
when the modesetting is under our control.
Reported-by: Chris Clayton <chris2553@googlemail.com>
Tested-by: Chris Clayton <chris2553@googlemail.com>
Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Starting with SandyBridge (though possible with earlier hacked BIOSes),
the BIOS may initialise the IGFX as secondary to a discrete GPU. Prior,
it would simply disable the integrated GPU. So we adjust our PCI class
mask to match any DISPLAY_CLASS device.

In such a configuration, the IGFX is not a primary VGA controller and
so should not take part in VGA arbitration, and the error return from
vga_client_register() is expected.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org

In order to workaround the issue with LVDS not working on the Lenovo
U160 apparently due to using the wrong SSC frequency, add an option to
disable SSC.
Suggested-by: Lukács, Árpád <lukacs.arpad@gmail.com>
Bugzillla: https://bugs.freedesktop.org/show_bug.cgi?id=32748

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org

Cleanup several aspects of the rc6 code:
  - misnamed intel_disable_clock_gating function (was only about rc6)
  - remove commented call to intel_disable_clock_gating
  - rc6 enabling code belongs in its own function (allows us to move the
    actual clock gating enable call back into restore_state)
  - allocate power & render contexts up front, only free on unload
    (avoids ugly lazy init at rc6 enable time)
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
[ickle: checkpatch cleanup]
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Enabling RC6 implies setting a graphics context.  Make sure we do that
only after the ring has been enabled, otherwise our ring commands will
hang.
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

When bringing up new hardware, or otherwise experimenting, GPU hangs are
a way of life. However, the automatic GPU reset can do more harm than
good under these circumstances, as we may wish to capture a full trace for
debugging.

Based on a patch by Zhenyu Wang.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

We need to track the state of the switch in drivers, so that after s/r
we don't resume the card we've explicitly switched off before. Also
don't allow a userspace open to occur if we've switched the gpu off.
Signed-off-by: Dave Airlie <airlied@redhat.com>

v2: Change IS_IRONLAKE to IS_GEN5 to adapt to 2.6.37

This patch adds new functions for use by the drm core:

.get_vblank_timestamp() provides a precise timestamp
for the end of the most recent (or current) vblank
interval of a given crtc, as needed for the DRI2
implementation of the OML_sync_control extension.
It is a thin wrapper around the drm function
drm_calc_vbltimestamp_from_scanoutpos() which does
almost all the work.

.get_scanout_position() provides the current horizontal
and vertical video scanout position and "in vblank"
status of a given crtc, as needed by the drm for use by
drm_calc_vbltimestamp_from_scanoutpos().

The patch modifies the pageflip completion routine
to use these precise vblank timestamps as the timestamps
for pageflip completion events.

This code has been only tested on a HP-Mini Netbook with
Atom processor and Intel 945GME gpu. The codepath for
(IS_G4X(dev) || IS_GEN5(dev) || IS_GEN6(dev)) gpu's
has not been tested so far due to lack of hardware.
Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de>
Acked-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Add frame buffer compression on Sandybridge. The method is similar to
Ironlake, except that two new registers of type GTTMMADR must be written
with the right fence info.
Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Once we have read the value out of the GT power well, we need to remove
the FORCE WAKE bit to allow the system to auto-power down.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

As suggested by Daniel Vetter, this is a safeguard should any of the
registers cause reference to PTE entries.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

So we can remove the repeated initialisation.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

The bulk of the change is to convert the growing list of rings into an
array so that the relationship between the rings and the semaphore sync
registers can be easily computed.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

This will declare the machine wedged, but is better than truly wedging
the machine.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

This still uses the agp functions to actually reinstate the mappings
(with a gross hack to make agp cooperate), but it wires everything
up correctly for the switchover.

The call to agp_rebind_memory can be dropped because all non-kms drivers
do all their rebinding on EnterVT.

v2: Be more paranoid and flush the chipset cache after restoring gtt
mappings.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

This has proven sufficient to recover from a hang of the GPU using the
gem_bad_blit test while at the KMS console then starting X.  When
attempting the same during an X session, the timer doesn't appear to
trigger.
Signed-off-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Signed-off-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Frame buffer compression is broken on Ironlake due to buggy hardware.
Currently it is disabled through chicken bits, but it still consumes
over 1W more than if we simply never attempt to enable the FBC code
paths.
Signed-off-by: Alex Shi <alex.shi@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: stable@kernel.org

2.6.36 appears to respect the 0400 mode we assigned to the parameter
preventing it from being adjusted after loading. However, this is safe
to adjust at runtime.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=31311

Reported-by: Fernando Lemos <fernandotcl@gmail.com>
Cc: stable@kernel.org
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Eliminate the racy device unload by embedding a shrinker into each
device. Smaller, simpler code.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

The ringbuffer keeps a pointer to the parent device, so we can use that
instead of passing around the pointer on the stack.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

So remove the redundant bit in the capabilities block and
s/IS_IRONLAKE/IS_GEN5/.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Based on an original patch by Zhenyu Wang, this initializes the BLT ring for
SandyBridge and enables support for user execbuffers.

Cc: Zhenyu Wang <zhenyuw@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

The issue is that we may become stuck executing a long running shader
and continually attempt to reset the GPU. (Or maybe we tickle some bug
and need to break the vicious cycle.) So if we are detect a second hang
within 5 seconds, give up trying to programme the GPU and report it
wedged.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

So far only found registers for i830, i845, i865 and one of those has no
effect on i865!

At this moment in time, attempting to reset i8xx is a little
optimistic...
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

When the GPU is reset, the fence registers are invalidated, so release
the objects and clear them out.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Otherwise we will hit a list handling assertion when moving the object
to the inactive list.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Suspending (especially hibernating) may take a finite amount of time,
during which a hotplug event may trigger and we will attempt to handle
it with inconsistent state. Disable hotplug polling around suspend and
resume.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=30070

Reported-by: Rui Tiago Matos <tiagomatos@gmail.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

This ring buffer is used for video decoding/encoding on Sandybridge.
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Previously we only tidied up the active bo lists for chipsets were we
would attempt to reset the GPU. However, this action is necessary for
the system to continue and reclaim the dead bo for all chipsets.

Pointed out, in passing, by Owain Ainsworth.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Clear the GPU read domain for the inactive objects on a reset so that
they are correctly invalidated on reuse.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

Owain Ainsworth noticed that the reset code failed to clear the flushing
list leaving the driver in an inconsistent state following a hung GPU.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

On more recent chipsets, restoring the display is not as simple as
writing a few registers, so force a full modeset of the current
configuration in order to retrain the display link.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>