[Bug 71701] Regression, Dual-Head, i915 mode-setting driver, "pipe B stuck"

[bugzilla-daemon at bugzilla.kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=71701

--- Comment #25 from Jani Nikula <jani.nikula at intel.com> ---
(In reply to Kris Karas from comment #24)
> The new drm-intel-nightly works!  I'm seeing 100 good out of 100 boots.  New
> AUX code is dealing with dongle delays just fine.

Great! ...although I am not sure I understand why. The code should be similar
to what we had in the driver. :o

> However, this appears to be new code suitable for mainline kernels 3.15 or
> even 3.16.

Please try 3.15-rc5 for completeness. Now that you have the git tree, it should
be simple.

> We need a simple patch to address the regression in the old code
> from 3.13.6 onwards.  And from what I gather, this new code is not likely to
> be backported into those stable trees.  Hence, the patch which I proposed
> above, in comment #19 is, I think, the simplest fix for those old trees, and
> really should be included by Greg KH; failing that, the ideologically
> correct fix would be to revoke the regression
> (87c3c227f580e2965feceae5f8973e9a38644963) until such time as a
> non-regressing patch were applied.

I'm afraid we can't backport your fix to stable, because it's not upstream, and
really there's not even a similar fix upstream. We need to figure out what it
is about the current upstream that fixes your issue, and see if we can backport
a similar fix.

-- 
You are receiving this mail because:
You are on the CC list for the bug.