Hi Hans, > -----Original Message----- > From: Hans Verkuil [mailto:hansverk@xxxxxxxxx] > Sent: Friday, February 10, 2017 8:38 PM > To: Zheng, Jian Xu <jian.xu.zheng@xxxxxxxxx>; hans.verkuil@xxxxxxxxx; > mchehab@xxxxxxxxxxxxxxx; Laurent.pinchart@xxxxxxxxxxxxxxxx > Cc: linux-media@xxxxxxxxxxxxxxx; Qiu, Tian Shu <tian.shu.qiu@xxxxxxxxx>; > Zhang, Lei L <lei.l.zhang@xxxxxxxxx>; Hu, Jerry W <jerry.w.hu@xxxxxxxxx>; > Sakari Ailus <sakari.ailus@xxxxxxxxxxxxxxx> > Subject: Re: media device drivers for Intel hardware platform incoming > > On 02/10/17 13:26, Zheng, Jian Xu wrote: > > Dear maintainers & reviewers, > > > > First, thank you so much for your time to reading this email. > > > > Recently we are about to upstream several media drivers on Intel hardware > platform, especially one ISP HW with hybrid imaging processing pipeline and its > driver may be the first Intel ISP driver trying to upstream in linux-media. > > Also 2 sensor drivers and VCM driver are included. > > > > Patch sets will likely reach at mailing list in several weeks. So I'm sending ice- > breaking email for a smooth communication in the next months. > > Looking forward to work with the most active and respected kernel driver > component. Also contributing to the Open Source community is our honor and > goal in the progress. > > > > If you're interested in, here are some introductions about those drivers > (details will be in the cover letter of patch set) > > 1. MIPI CSI-2 device is part of ISP and is one PCI device. It would pass > on data from MIPI bus to the DDR memory(without image processing). > > MIPI CSI-2 driver registers 1 v4l2 subdev and 1 video node. > Buffers are managed via vb2. > > > > 2. Imaging processing unit is another part of ISP. It's one PCI device and > a hardware with one hybrid imaging processing hardware pipeline. > > This HW consumes DDR RAW input and produces the YUV output > after running imaging processing pipeline. > > So it's abstracted as one memory to memory device with 5 > output pins(main output, vf output, 3A statistics, DVS statistics, LACE), > > 2 input pins(RAW image, ISP parameters). Certain video nodes > are created for buffer looping. Buffers are managed via vb2. > > > > 3. Camera sensors are i2c devices and drivers register 2 v4l2 subdevs > and media entities as Pixel Array and Binner (binning/subsample). > > Exposure/gain/hflip/vflip/vblank/hblank settings are set via > certain V4L2_CID_* IOCTLs. > > > > 4. VCM is i2c device and registers as a v4l2 subdev. Lens position is > > set through V4L2_CID_FOCUS_ABSOLUTE > > Great news! Thank you for working on this. > > Just one question for my understanding: Sakari, is this something you've been > involved in or is this from a different Intel team? > > If Sakari has not been involved in this, then it is probably a good idea to work > with him to do a preliminary code review. > > Looking forward to the patch series, Thanks a lot for your prompt response. Sakari and I are from the same team and comments from him are under addressing. :-) > > Hans -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
