[xserver PULL for 1.19] Various modesetting and other fixes

[Hans de Goede]

Hi Adam, Keith,

Here is a pull-req consisting of a bunch of fixes written by
me reviewed by others as well as a bunch of fixes from patchwork /
the list which I've picked up and which are reviewed by me.

Note there is one unreviewed patch in here v2 of:
"glx: Always enable EXT_texture_from_pixmap for DRI swrast glx"

But I believe that Adam and I are in agreement wrt v2, so that
one should be good to go too.


The following changes since commit d0c5d205a919fc1d2eb599356090b58b1bf0176d:

   dix: Make InitCoreDevices() failures more verbose. (2016-09-21 21:11:40 +1000)

are available in the git repository at:

   git://people.freedesktop.org/~jwrdegoede/xserver for-server-1.19

for you to fetch changes up to 01189c04a073447dc4eb474ec4cee7c59be5576e:

   Simplify, statement is always true. (2016-09-23 15:55:22 +0300)

----------------------------------------------------------------
Adam Jackson (1):
       xephyr: Don't crash if the server advertises zero xv adaptors

Daniel Martin (1):
       modesetting: Consume all available udev events at once

David CARLIER (1):
       xfree86: small memory leaks fixes

Hans de Goede (8):
       modesetting: Fix reverse prime partial update issues on secondary GPU outputs
       modesetting: Fix reverse prime update lagging on secondary GPU outputs
       xf86RandR12: Move calculating of shift inside init_one_component
       xf86RandR12: Fix XF86VidModeSetGamma triggering a BadImplementation error
       glx: Always enable EXT_texture_from_pixmap for DRI swrast glx
       modesetting: Do not call drmModeSetCursor2() twice on every cursor change
       Xext: Fix a memory leak
       XF86VidMode: Fix free() on walked pointer

Keith Packard (2):
       os: Ready clients with pending output aren't flushed, so set NewOutputPending
       os: Clear saved poll events in listen so that edge triggering works

Kyle Guinn (1):
       xfree86: Fix null pointer dereference

Laszlo Ersek (1):
       xfree86: recognize primary BUS_PCI device in xf86IsPrimaryPlatform()

Maya Rashish (1):
       Simplify, statement is always true.

Michael Thayer (3):
       modesetting: only fall back to drmModeSetCursor() on -EINVAL
       modesetting: always set a hardware cursor when requested to load one.
       modesetting: allow switching from software to hardware cursors (v4).

Qiang Yu (1):
       config: fix GPUDevice fail when AutoAddGPU off + BusID

  Xext/shm.c                                       |  4 +-
  Xext/vidmode.c                                   |  2 +-
  glx/glxdriswrast.c                               |  3 +-
  hw/kdrive/ephyr/ephyrvideo.c                     |  2 +-
  hw/xfree86/common/xf86pciBus.c                   |  6 +--
  hw/xfree86/common/xf86platformBus.c              | 28 +++++++++--
  hw/xfree86/ddc/ddc.c                             |  2 +-
  hw/xfree86/drivers/modesetting/driver.c          | 29 +++++++++---
  hw/xfree86/drivers/modesetting/drmmode_display.c | 60 +++++++++++-------------
  hw/xfree86/drivers/modesetting/drmmode_display.h |  2 -
  hw/xfree86/i2c/xf86i2c.c                         |  8 ++--
  hw/xfree86/modes/xf86RandR12.c                   | 44 ++++++++++++++---
  hw/xfree86/utils/gtf/gtf.c                       |  2 +
  os/io.c                                          |  3 +-
  os/ospoll.c                                      |  8 +++-
  15 files changed, 132 insertions(+), 71 deletions(-)

Regards,

Hans