linux-stable.git/drivers/char/agp, branch linux-4.1.y

agp/intel: Flush all chipset writes after updating the GGTT

2018-05-23T01:33:47+00:00

[ Upstream commit 8516673a996870ea0ceb337ee4f83c33c5ec3111 ]

Before accessing the GGTT we must flush the PTE writes and make them
visible to the chipset, or else the indirect access may end up in the
wrong page. In commit 3497971a71d8 ("agp/intel: Flush chipset writes
after updating a single PTE"), we noticed corruption of the uploads for
pwrite and for capturing GPU error states, but it was presumed that the
explicit calls to intel_gtt_chipset_flush() were sufficient for the
execbuffer path. However, we have not been flushing the chipset between
the PTE writes and access via the GTT itself.

For simplicity, do the flush after any PTE update rather than try and
batch the flushes on a just-in-time basis.

References: 3497971a71d8 ("agp/intel: Flush chipset writes after updating a single PTE")
Signed-off-by: Chris Wilson 
Cc: Tvrtko Ursulin 
Cc: Mika Kuoppala 
Cc: drm-intel-fixes@lists.freedesktop.org
Reviewed-by: Joonas Lahtinen 
Link: https://patchwork.freedesktop.org/patch/msgid/20171208214616.30147-1-chris@chris-wilson.co.uk
Signed-off-by: Sasha Levin

agp/intel: Fix typo in needs_ilk_vtd_wa()

2015-08-03T16:29:18+00:00

commit 8b572a4200828b4e75cc22ed2f494b58d5372d65 upstream.

In needs_ilk_vtd_wa(), we pass in the GPU device but compared it against
the ids for the mobile GPU and the mobile host bridge. That latter is
impossible and so likely was just a typo for the desktop GPU device id
(which is also buggy).

Fixes commit da88a5f7f7d434e2cde1b3e19d952e6d84533662
Author: Chris Wilson 
Date:   Wed Feb 13 09:31:53 2013 +0000

    drm/i915: Disable WC PTE updates to w/a buggy IOMMU on ILK

Reported-by: Ting-Wei Lan 
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91127
References: https://bugzilla.freedesktop.org/show_bug.cgi?id=60391
Signed-off-by: Chris Wilson 
Cc: Daniel Vetter 
Reviewed-by: Daniel Vetter 
Signed-off-by: Jani Nikula 
Signed-off-by: Greg Kroah-Hartman

Merge branch 'drm-intel-next' of git://anongit.freedesktop.org/drm-intel into drm-next

2015-02-05T00:32:44+00:00

drm-intel-next-2015-01-30:
- chv rps improvements from Ville
- atomic state handling prep work from Ander
- execlist request tracking refactoring from Nick Hoath
- forcewake code consolidation from Chris&Mika
- fastboot plane config refactoring and skl support from Damien
- some more skl pm patches all over (Damien)
- refactor dsi code to use drm dsi helpers and drm_panel infrastructure (Jani)
- first cut at experimental atomic plane updates (Matt Roper)
- piles of smaller things all over, as usual

* 'drm-intel-next' of git://anongit.freedesktop.org/drm-intel: (102 commits)
  drm/i915: Remove bogus locking check in the hangcheck code
  drm/i915: Update DRIVER_DATE to 20150130
  drm/i915: Use pipe_config's cpu_transcoder for reading encoder hw state
  drm/i915: Fix a use-after-free in intel_execlists_retire_requests
  drm/i915: Split shared dpll setup out of __intel_set_mode()
  drm/i915: Don't do posting reads on getting forcewake
  drm/i915: Do uncore early sanitize after domain init
  drm/i915: Handle CHV in vlv_set_rps_idle()
  drm/i915: Remove nested work in gpu error handling
  drm/i915/documentation: Add intel_uncore.c to drm.tmpl
  drm/i915/dsi: remove intel_dsi_cmd.c and the unused functions therein
  drm/i915/dsi: move dpi_send_cmd() to intel_dsi.c and make it static
  drm/i915/dsi: remove old read/write functions in favor of new stuff
  drm/i915/dsi: make the vbt panel driver use mipi_dsi_device for transfers
  drm/i915/dsi: add drm mipi dsi host support
  drm/i915/dsi: switch to drm_panel interface
  drm/i915/skl: Enabling PSR on Skylake
  Revert "drm/i915: Fix mutex->owner inspection race under DEBUG_MUTEXES"
  drm/i915: Be consistent on printing seqnos
  drm/i915: Display current hangcheck status in debugfs
  ...

agp: change agp_free_page_array to use kvfree

2015-01-29T02:37:41+00:00

Change agp_free_page_array to use kvfree function,
remove the duplicated code.

Signed-off-by: Yalin Wang 
Signed-off-by: Dave Airlie

agp/intel: Serialise after GTT updates

2015-01-28T09:15:28+00:00

An interesting bug occurs on Pineview through which the root cause is
that the writes of the PTE values into the GTT is not serialised with
subsequent memory access through the GTT (when using WC updates of the
PTE values). This is despite there being a posting read after the GTT
update. However, by changing the address of the posting read, the memory
access is indeed serialised correctly.

Whilst we are manipulating the memory barriers, we can remove the
compiler :memory restraint on the intermediate PTE writes knowing that
we explicitly perform a posting read afterwards.

v2: Replace posting reads with explicit write memory barriers - in
particular this is advantages in case of single page objects. Update
comments to mention this issue is only with WC writes.

Testcase: igt/gem_exec_big #pnv
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=88191
Tested-by: huax.lu@intel.com (v1)
Signed-off-by: Chris Wilson 
Cc: Daniel Vetter 
Signed-off-by: Daniel Vetter

agp: Fix up email address & attributions in AGP MODULE_AUTHOR tags

2014-12-23T10:03:10+00:00

- Remove soon-to-be-dead @redhat address.
- Jeff Hartmann wrote the bulk of the original backend code, and should
  at least get a mention in the MODULE_AUTHOR for backend.o
- Various people at Intel have done a lot more work than myself on the
  intel-* drivers, so again, mention that.

Signed-off-by: Dave Jones 
Signed-off-by: Dave Airlie

agp/intel-gtt: Remove get/put_pages

2014-09-19T12:43:11+00:00

If a page isn't allocated as __GFP_MOVEABLE it won't move around, so
no need to grab a reference to lock it into place.

Discovered while reviewing page allocation handling in i915 gem.

Acked-by: Chris Wilson 
Signed-off-by: Daniel Vetter

agp: remove read/write stubs

2014-08-01T20:43:51+00:00

The VFS layer handles those in the very same way, if unset. No need for
additional stubs.

Signed-off-by: Mathias Krause 
Cc: Alexander Viro 
Signed-off-by: Dave Airlie

agp: info leak in agpioc_info_wrap()

2014-05-06T20:05:00+00:00

On 64 bit systems the agp_info struct has a 4 byte hole between
->agp_mode and ->aper_base.  We need to clear it to avoid disclosing
stack information to userspace.

Signed-off-by: Dan Carpenter 
Cc: David Airlie 
Cc: Daniel Vetter 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

drivers/char: delete non-required instances of include

2014-02-07T23:10:19+00:00

None of these files are actually using any __init type directives
and hence don't need to include .  Most are just a
left over from __devinit and __cpuinit removal, or simply due to
code getting copied from one driver to the next.

Cc: David Airlie 
Cc: Matt Mackall 
Cc: Herbert Xu 
Cc: Kukjin Kim 
Cc: Corey Minyard 
Cc: Chris Metcalf 
Cc: Arnd Bergmann 
Cc: Greg Kroah-Hartman 
Cc: Peter Huewe 
Cc: Ashley Lai 
Cc: Marcel Selhorst 
Signed-off-by: Paul Gortmaker 
Signed-off-by: Greg Kroah-Hartman