linux-stable.git/drivers/gpu/drm/vc4, branch v4.7.2

drm/vc4: Make pageflip completion handling more robust.

2016-06-06T20:00:40+00:00

Protect both the setup of the pageflip event and the
latching of the new requested displaylist head pointer
by the event lock, so we can't get into a situation
where vc4_atomic_flush latches the new display list via
HVS_WRITE, then immediately gets preempted before queueing
the pageflip event, then the page-flip completes in hw and
the vc4_crtc_handle_page_flip() runs and no-ops due to
lack of a pending pageflip event, then vc4_atomic_flush
continues and only then queues the pageflip event - after
the page flip handling already no-oped. This would cause
flip completion handling only at the next vblank - one
frame too late.

In vc4_crtc_handle_page_flip() check the actual DL head
pointer in SCALER_DISPLACTX against the requested pointer
for page flip to make sure that the flip actually really
completed in the current vblank and doesn't get deferred
to the next one because the DL head pointer was written
a bit too late into SCALER_DISPLISTX, after start of
vblank, and missed the boat. This avoids handling a
pageflip completion too early - one frame too early.

According to Eric, DL head pointer updates which were
written into the HVS DISPLISTX reg get committed to hardware
at the last pixel of active scanout. Our vblank interrupt
handler, as triggered by PV_INT_VFP_START irq, gets to run
earliest at the first pixel of HBLANK at the end of the
last scanline of active scanout, ie. vblank irq handling
runs at least 1 pixel duration after a potential pageflip
completion happened in hardware.

This ordering of events in the hardware, together with the
lock protection and SCALER_DISPLACTX sampling of this patch,
guarantees that pageflip completion handling only runs at
exactly the vblank irq of actual pageflip completion in all
cases.

Background info from Eric about the relative timing of
HVS, PV's and trigger points for interrupts, DL updates:

https://lists.freedesktop.org/archives/dri-devel/2016-May/107510.html

Tested on RPi 2B with hardware timing measurement equipment
and shown to no longer complete flips too early or too late.

Signed-off-by: Mario Kleiner 
Reviewed-by: Eric Anholt

drm/vc4: Fix ioctl permissions for render nodes.

2016-06-06T20:00:30+00:00

Contrary to other flags to DRM_IOCTL_DEF_DRV(), which restrict usage,
the flag for render node is an enabler (the IOCTL can't be used from
render node if it's not present).  So DRM_RENDER_ALLOW needs to be
added to all the flags that were previously 0.

Signed-off-by: Herve Jourdain 
Reviewed-by: Eric Anholt 
Fixes: 0cd3e2747662 ("drm/vc4: Add missing render node support")

drm/vc4: Return -EBUSY if there's already a pending flip event.

2016-05-31T22:15:55+00:00

As per the documentation in drm_crtc.h, atomic_commit should return
-EBUSY if an asynchronous update is requested and there is an earlier
update pending.

v2: Rebase on the s/async/nonblock/ change.

Signed-off-by: Robert Foss 
Reviewed-by: Eric Anholt

drm/vc4: Fix drm_vblank_put/get imbalance in page flip path.

2016-05-31T22:15:35+00:00

The async page flip path was missing drm_crtc_vblank_get/put
completely. The sync flip path was missing a vblank put, so async
flips only reported proper pageflip completion events by chance,
and vblank irq's never turned off after a first vsync'ed page flip
until system reboot.

Tested against Raspian kernel 4.4.8 tree on RPi 2B.

Signed-off-by: Mario Kleiner 
Reviewed-by: Eric Anholt 
Fixes: b501bacc6060 ("drm/vc4: Add support for async pageflips.")

drm/vc4: Fix get_vblank_counter with proper no-op for Linux 4.4+

2016-05-31T22:15:29+00:00

get_vblank_counter hooked up to drm_vblank_count() which alway was
non-sensical but didn't hurt in the past. Since Linux 4.4 it
triggers a WARN_ON_ONCE in drm_update_vblank_count on first vblank
irq disable, so fix it by hooking to drm_vblank_no_hw_counter().

Tested against Raspian kernel 4.4.8 tree on RPi 2B.

Signed-off-by: Mario Kleiner 
Reviewed-by: Eric Anholt 
Fixes: c8b75bca92cb ("drm/vc4: Add KMS support for Raspberry Pi.")

Merge branch 'drm-next' of git://people.freedesktop.org/~airlied/linux

2016-05-23T18:48:48+00:00

Pull drm updates from Dave Airlie:
 "Here's the main drm pull request for 4.7, it's been a busy one, and
  I've been a bit more distracted in real life this merge window.  Lots
  more ARM drivers, not sure if it'll ever end.  I think I've at least
  one more coming the next merge window.

  But changes are all over the place, support for AMD Polaris GPUs is in
  here, some missing GM108 support for nouveau (found in some Lenovos),
  a bunch of MST and skylake fixes.

  I've also noticed a few fixes from Arnd in my inbox, that I'll try and
  get in asap, but I didn't think they should hold this up.

  New drivers:
   - Hisilicon kirin display driver
   - Mediatek MT8173 display driver
   - ARC PGU - bitstreamer on Synopsys ARC SDP boards
   - Allwinner A13 initial RGB output driver
   - Analogix driver for DisplayPort IP found in exynos and rockchip

  DRM Core:
   - UAPI headers fixes and C++ safety
   - DRM connector reference counting
   - DisplayID mode parsing for Dell 5K monitors
   - Removal of struct_mutex from drivers
   - Connector registration cleanups
   - MST robustness fixes
   - MAINTAINERS updates
   - Lockless GEM object freeing
   - Generic fbdev deferred IO support

  panel:
   - Support for a bunch of new panels

  i915:
   - VBT refactoring
   - PLL computation cleanups
   - DSI support for BXT
   - Color manager support
   - More atomic patches
   - GEM improvements
   - GuC fw loading fixes
   - DP detection fixes
   - SKL GPU hang fixes
   - Lots of BXT fixes

  radeon/amdgpu:
   - Initial Polaris support
   - GPUVM/Scheduler/Clock/Power improvements
   - ASYNC pageflip support
   - New mesa feature support

  nouveau:
   - GM108 support
   - Power sensor support improvements
   - GR init + ucode fixes.
   - Use GPU provided topology information

  vmwgfx:
   - Add host messaging support

  gma500:
   - Some cleanups and fixes

  atmel:
   - Bridge support
   - Async atomic commit support

  fsl-dcu:
   - Timing controller for LCD support
   - Pixel clock polarity support

  rcar-du:
   - Misc fixes

  exynos:
   - Pipeline clock support
   - Exynoss4533 SoC support
   - HW trigger mode support
   - export HDMI_PHY clock
   - DECON5433 fixes
   - Use generic prime functions
   - use DMA mapping APIs

  rockchip:
   - Lots of little fixes

  vc4:
   - Render node support
   - Gamma ramp support
   - DPI output support

  msm:
   - Mostly cleanups and fixes
   - Conversion to generic struct fence

  etnaviv:
   - Fix for prime buffer handling
   - Allow hangcheck to be coalesced with other wakeups

  tegra:
   - Gamme table size fix"

* 'drm-next' of git://people.freedesktop.org/~airlied/linux: (1050 commits)
  drm/edid: add displayid detailed 1 timings to the modelist. (v1.1)
  drm/edid: move displayid validation to it's own function.
  drm/displayid: Iterate over all DisplayID blocks
  drm/edid: move displayid tiled block parsing into separate function.
  drm: Nuke ->vblank_disable_allowed
  drm/vmwgfx: Report vmwgfx version to vmware.log
  drm/vmwgfx: Add VMWare host messaging capability
  drm/vmwgfx: Kill some lockdep warnings
  drm/nouveau/gr/gf100-: fix race condition in fecs/gpccs ucode
  drm/nouveau/core: recognise GM108 chipsets
  drm/nouveau/gr/gm107-: fix touching non-existent ppcs in attrib cb setup
  drm/nouveau/gr/gk104-: share implementation of ppc exception init
  drm/nouveau/gr/gk104-: move rop_active_fbps init to nonctx
  drm/nouveau/bios/pll: check BIT table version before trying to parse it
  drm/nouveau/bios/pll: prevent oops when limits table can't be parsed
  drm/nouveau/volt/gk104: round up in gk104_volt_set
  drm/nouveau/fb/gm200: setup mmu debug buffer registers at init()
  drm/nouveau/fb/gk20a,gm20b: setup mmu debug buffer registers at init()
  drm/nouveau/fb/gf100-: allocate mmu debug buffers
  drm/nouveau/fb: allow chipset-specific actions for oneinit()
  ...

drm: Nuke ->vblank_disable_allowed

2016-05-20T21:03:31+00:00

This was added in

commit 0a3e67a4caac273a3bfc4ced3da364830b1ab241
Author: Jesse Barnes 
Date:   Tue Sep 30 12:14:26 2008 -0700

    drm: Rework vblank-wait handling to allow interrupt reduction.

to stay backwards-compatible with old UMS code that didn't even tell
the kernel when it did a modeset, so that the kernel could
save/restore vblank counters. At worst this means vblanks will be
somewhat funky on a setup that very likely no one still runs.

So let's just nuke it.

Plan B would be to set it unconditionally in drm_vblank_init for kms
drivers, instead of in each driver separately. So if this patch breaks
anything please only restore the hunks in drmP.h and drm_irq.c, plus
add a check for DRIVER_MODESET in drm_vblank_init.

Stumbled over this in a discussion on irc with Chris.

Cc: Chris Wilson 
Cc: Alex Deucher 
Cc: Liviu Dudau 
Cc: Russell King 
Cc: Thierry Reding 
Cc: Eric Anholt 
Cc: Laurent Pinchart 
Cc: Inki Dae 
Cc: Tomi Valkeinen 
Cc: Mark Yao 
Cc: Sascha Hauer 
Cc: Philipp Zabel 
Signed-off-by: Daniel Vetter 
Reviewed-by: Alex Deucher 
Acked-by: Liviu Dudau 
Acked-by: Laurent Pinchart 
Tested-by: Laurent Pinchart 
Signed-off-by: Dave Airlie

drm: Drop plane argument from __drm_atomic_helper_plane_destroy_state

2016-05-17T06:56:55+00:00

It's unused, and really this helper should only look at the state
structure and nothing else.

v2: Fix commit message (Laurent).

v3: Rebase onto mtk driver merge.

Cc: Maarten Lankhorst 
Cc: Thierry Reding 
Cc: Eric Anholt 
Cc: Laurent Pinchart 
Cc: Inki Dae 
Cc: Tomi Valkeinen 
Cc: Mark Yao 
Acked-by: Thierry Reding 
Acked-by: Maarten Lankhorst 
Acked-by: Laurent Pinchart 
Acked-by: Eric Anholt 
Signed-off-by: Daniel Vetter 
Signed-off-by: Daniel Vetter 
Link: http://patchwork.freedesktop.org/patch/msgid/1462804451-15318-2-git-send-email-daniel.vetter@ffwll.ch

drm: Drop crtc argument from __drm_atomic_helper_crtc_destroy_state

2016-05-17T06:56:41+00:00

It's unused, and really this helper should only look at the state
structure and nothing else.

v2: Rebase on top of rockchip changes

v3: Drop unrelated hunk, spotted by Laurent.

v4: Rebase onto mtk driver merge.

Cc: Maarten Lankhorst 
Cc: Thierry Reding 
Cc: Eric Anholt 
Cc: Mark Yao 
Acked-by: Thierry Reding 
Acked-by: Maarten Lankhorst 
Acked-by: Laurent Pinchart 
Acked-by: Eric Anholt 
Signed-off-by: Daniel Vetter 
Signed-off-by: Daniel Vetter 
Link: http://patchwork.freedesktop.org/patch/msgid/1462804451-15318-1-git-send-email-daniel.vetter@ffwll.ch

drm: Remove unused drm_device from drm_gem_object_lookup()

2016-05-17T06:47:30+00:00

drm_gem_object_lookup() has never required the drm_device for its file
local translation of the user handle to the GEM object. Let's remove the
unused parameter and save some space.

Signed-off-by: Chris Wilson 
Cc: dri-devel@lists.freedesktop.org
Cc: Dave Airlie 
Cc: Daniel Vetter 
[danvet: Fixup kerneldoc too.]
Signed-off-by: Daniel Vetter