Hello! On 09/17/2018 02:17 PM, Niklas Söderlund wrote: >>>> Add the R-Car V3H (AKA R8A77980) SoC support to the R-Car CSI2 driver. >>>> >>>> Signed-off-by: Sergei Shtylyov <sergei.shtylyov@xxxxxxxxxxxxxxxxxx> >>> >>> It looks good but before I add my tag I would like to know if you tested >>> this on a V3H? >> >> No, have not tested the upstream patch, seemed too cumbersome (we did >> test the BSP patch AFAIK). > > Did you find time to test this change? I think it looks good but I would Unfortunately, no. > feel a lot better about this change if I knew it was tested :-) More > often then not have we shaken a bug out of the rcar-{csi2,vin} tree by > adding support for a new SoC. > > Also if you for some reason need to resend this series could you split > the DT documentation and driver changes in two separate patches? OK. >>> In the past extending the R-Car CSI-2 receiver to a new >>> SoC always caught some new corner case :-) >>> >>> I don't have access to a V3H myself otherwise I would of course test it >>> myself. >> >> CSI-2 on the Condor board is connected to a pair of MAX9286 GMSL de-serializers >> which are connected to 4 (composite?) connectors... There's supposed to be sensor >> chip on the other side, AFAIK... > > There is experimental patches which are tested on V3M that uses the > MAX9286 GMSL setup if you feel brave and want to try and extend them to > cover V3H ;-P Wouldn't hurt having them. :-) MBR, Sergei