Re: [PATCH v3 00/15] Improvements and fixes for mxsfb DRM driver

[Robert Chiras]

Hi Leonard,

On Lu, 2019-08-26 at 19:19 +0000, Leonard Crestez wrote:
> On 26.08.2019 17:35, Stefan Agner wrote:
> > 
> > On 2019-08-26 14:05, Guido GÃnther wrote:
> > > 
> > > Hi,
> > > On Wed, Aug 21, 2019 at 01:15:40PM +0300, Robert Chiras wrote:
> > > > 
> > > > This patch-set improves the use of eLCDIF block on iMX 8 SoCs
> > > > (like 8MQ, 8MM
> > > > and 8QXP). Following, are the new features added and fixes from
> > > > this
> > > > patch-set:
> > > I've applied this whole series on top of my NWL work and it looks
> > > good
> > > with a DSI panel. Applying the whole series also fixes an issue
> > > where
> > > after unblank the output was sometimes shifted about half a
> > > screen width
> > > to the right (which didn't happen with DCSS). So at least from
> > > the parts
> > > I could test:
> > > 
> > > ÂÂÂTested-by: Guido GÃnther <agx@xxxxxxxxxxx>
> > > 
> > > for the whole thing.
> > Thanks for testing! What SoC did you use? I think it would be good
> > to
> > also give this a try on i.MX 7 or i.MX 6ULL before merging.
> I did a quick test on imx6sx-sdb and it seems that [PATCH 07/15]Â
> "drm/mxsfb: Fix the vblank events" causes a hang on boot, even
> without aÂ
> panel.
> 
> If I revert that particular patch it seems to be fine: the new pixelÂ
> formats seem to work in modetest (checked with sii,43wvf1g panel).
Thanks for feedback. I tested this and, indeed there are issues on 6SX
with this particular patch. It seems that there is a race-condition
caused by the vblank_on call in enable and IRQ routine. Since this is
not happening on any of i.MX8 SoC, I suspect the axi clock usage.
I think I will just remove this patch from the patch-set and handle
this case separately.
> 
> --
> Regards,
> Leonard

Regards,
Robert