[Intel-gfx] [BUG] Again panning problems on i830 chipsets

[Thomas Richter]

Hi folks,

when playing with the 4.1.0-rc5 kernel coming from the latest
intel-drm-nightly branch yesterday, I noted that this branch again
introduces problems with panning on i830 based machines. This panning
bug seems to be unrelated to the long-ago resolved mis-configuration of
the watermark registers.

>From the effects I see, I would guess that the problem is due to some
i830 registers now changing mid-screen rather than synchronized to the
top of the screen (or vertical blank). In fact, I see some tearing-type
artifacts during scrolling where the panel misses synchronization
mid-screen.

I have not yet tried to bisect the error due to the slowness of the
kernel compilation on those machines, but this seems to be a relatively
recent bug.

Greetings,
	Thomas