Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1337041 > unrolled thread

[GIT PULL] drm/vc4: changes for drm-fixes

Started byEric Anholt <eric@anholt.net>
First post2016-02-18 06:30 +0100
Last post2016-02-19 04:00 +0100
Articles 2 — 2 participants

Back to article view | Back to linux.kernel


Contents

  [GIT PULL] drm/vc4: changes for drm-fixes Eric Anholt <eric@anholt.net> - 2016-02-18 06:30 +0100
    Re: [GIT PULL] drm/vc4: changes for drm-fixes Dave Airlie <airlied@gmail.com> - 2016-02-19 04:00 +0100

#1337041 — [GIT PULL] drm/vc4: changes for drm-fixes

FromEric Anholt <eric@anholt.net>
Date2016-02-18 06:30 +0100
Subject[GIT PULL] drm/vc4: changes for drm-fixes
Message-ID<r3pk6-MU-5@gated-at.bofh.it>
These changes have been on the list for over a week without comment,
so I think it might be an appropriate time to pull (without an active
development community, getting review will probably be hard).  I'm
also curious when your window for taking fixes would close -- the
modesetting series I sent yesterday covers what most users have been
running into.

The following changes since commit 388f7b1d6e8ca06762e2454d28d6c3c55ad0fe95:

  Linux 4.5-rc3 (2016-02-07 15:38:30 -0800)

are available in the git repository at:

  git@github.com:anholt/linux.git tags/drm-vc4-fixes-2016-02-17

for you to fetch changes up to 36cb6253f9383fd9a59ee7b8458c6232ef48577c:

  drm/vc4: Use runtime PM to power cycle the device when the GPU hangs. (2016-02-16 12:21:01 -0800)

----------------------------------------------------------------
This pull request fixes GPU reset (which was disabled shortly after
V3D integration due to build breakage) and waits for idle in the
presence of signals (which X likes to do a lot).

----------------------------------------------------------------
Eric Anholt (8):
      drm/vc4: Validate that WAIT_BO padding is cleared.
      drm/vc4: Fix the clear color for the first tile rendered.
      drm/vc4: Return an ERR_PTR from BO creation instead of NULL.
      drm/vc4: Fix -ERESTARTSYS error return from BO waits.
      drm/vc4: Drop error message on seqno wait timeouts.
      drm/vc4: Fix spurious GPU resets due to BO reuse.
      drm/vc4: Enable runtime PM.
      drm/vc4: Use runtime PM to power cycle the device when the GPU hangs.

 drivers/gpu/drm/vc4/vc4_bo.c        | 16 ++++-----
 drivers/gpu/drm/vc4/vc4_drv.h       | 13 ++++++--
 drivers/gpu/drm/vc4/vc4_gem.c       | 65 ++++++++++++++++++++++++++++---------
 drivers/gpu/drm/vc4/vc4_irq.c       |  2 +-
 drivers/gpu/drm/vc4/vc4_render_cl.c | 22 ++++++-------
 drivers/gpu/drm/vc4/vc4_v3d.c       | 48 ++++++++++++++++++++-------
 drivers/gpu/drm/vc4/vc4_validate.c  |  4 +--
 7 files changed, 118 insertions(+), 52 deletions(-)

[toc] | [next] | [standalone]


#1337851

FromDave Airlie <airlied@gmail.com>
Date2016-02-19 04:00 +0100
Message-ID<r3Jst-6IV-1@gated-at.bofh.it>
In reply to#1337041
On 18 February 2016 at 15:21, Eric Anholt <eric@anholt.net> wrote:
> These changes have been on the list for over a week without comment,
> so I think it might be an appropriate time to pull (without an active
> development community, getting review will probably be hard).  I'm
> also curious when your window for taking fixes would close -- the
> modesetting series I sent yesterday covers what most users have been
> running into.

Pretty much depends on big and serious the fixes are.
As things start to taper down to rc5/6 time I'd want only smaller ones,
it really should be regressions only, but for a new driver like vc4 some
leeway can be given.

Dave.

[toc] | [prev] | [standalone]


Back to top | Article view | linux.kernel


csiph-web