Re: [PATCH 0/9] Wrap up ILK-style atomic watermarks

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Wed, Nov 25, 2015 at 09:04:11AM -0800, Matt Roper wrote:
> On Wed, Nov 25, 2015 at 07:00:11PM +0200, Ville Syrjälä wrote:
> > On Wed, Nov 25, 2015 at 08:48:26AM -0800, Matt Roper wrote:
> > > We've been chasing a regression[1] that prevented us from merging the last
> > > couple patches of the ILK-style atomic watermark series.  We've finally
> > > identified the culprit --- if we fail to reconstruct the BIOS FB, our internal
> > > driver state was left in an inconsistent state which caused confusion for the
> > > watermark code.  Here's a series that collects that fix (along with a couple
> > > other bugfixes that were found while debugging), a few debugging enhancements,
> > > and the completion of the ILK atomic watermark series.
> > > 
> > >  * The first patch in this series addresses that issue and
> > >    ensures that we cleanly turn off the primary plane if we're unable to inherit
> > >    the framebuffer.  
> > >  * Patches 2-4 just clarify/expand some of the information dumped by the
> > >    driver while debugging.
> > >  * Patches 5 and 6 fix two additional bugs that were discovered while
> > >    searching for the cause of the regression.
> > >  * Patch 7 just adds some extra paranoia and invariant checking to the
> > >    watermark code.
> > >  * Patches 8 and 9 are the two final patches from the original atomic watermark
> > >    series; with the fixes earlier in the series, we've confirmed that they no
> > >    longer cause regressions on Jani's machine.
> > > 
> > > [1] http://lists.freedesktop.org/archives/intel-gfx/2015-October/077113.html
> > 
> > Quickly skimmed through the series to see if there was a fix to
> > the "merging optimal wms instead of merging active wms like we
> > should" bug we have in the current code right now. Nothing
> > relevant caught my eye at least.
> 
> I'm not sure I'm aware of this problem; is there a bugzilla or mailing
> list thread that describes the issue?

Maybe not. The problem is that ilk_wm_merge() and co. use the optimal
wms when merging from all pipes. They should use the active wms
instead.

> 
> 
> Matt
> 
> > 
> > > 
> > > Matt Roper (9):
> > >   drm/i915: Disable primary plane if we fail to reconstruct BIOS fb
> > >   drm/i915: Dump in-flight plane state while dumping in-flight CRTC
> > >     state
> > >   drm/i915: Clarify plane state during CRTC state dumps (v2)
> > >   drm/i915: Dump pipe config after initial FB is reconstructed
> > >   drm/i915: Setup clipped src/dest coordinates during FB reconstruction
> > >     (v2)
> > >   drm/i915: Convert hsw_compute_linetime_wm to use in-flight state
> > >   drm/i915: Add extra paranoia to ILK watermark calculations
> > >   drm/i915: Sanitize watermarks after hardware state readout (v2)
> > >   drm/i915: Add two-stage ILK-style watermark programming (v7)
> > > 
> > >  drivers/gpu/drm/drm_atomic_helper.c  |   1 +
> > >  drivers/gpu/drm/i915/i915_drv.h      |   5 +
> > >  drivers/gpu/drm/i915/intel_atomic.c  |   1 +
> > >  drivers/gpu/drm/i915/intel_display.c | 195 +++++++++++++++++++++++++++++++++--
> > >  drivers/gpu/drm/i915/intel_drv.h     |  31 +++++-
> > >  drivers/gpu/drm/i915/intel_pm.c      | 186 ++++++++++++++++++++++++---------
> > >  6 files changed, 359 insertions(+), 60 deletions(-)
> > > 
> > > -- 
> > > 2.1.4
> > > 
> > > _______________________________________________
> > > Intel-gfx mailing list
> > > Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
> > > http://lists.freedesktop.org/mailman/listinfo/intel-gfx
> > 
> > -- 
> > Ville Syrjälä
> > Intel OTC
> 
> -- 
> Matt Roper
> Graphics Software Engineer
> IoTG Platform Enabling & Development
> Intel Corporation
> (916) 356-2795

-- 
Ville Syrjälä
Intel OTC
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
http://lists.freedesktop.org/mailman/listinfo/intel-gfx




[Index of Archives]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]
  Powered by Linux