Hi Laurent, Am Donnerstag, 7. Juli 2022, 13:34:50 CEST schrieb Laurent Pinchart: > Hi Alexander, > > On Thu, Jul 07, 2022 at 12:45:01PM +0200, Alexander Stein wrote: > > Am Donnerstag, 7. Juli 2022, 02:23:55 CEST schrieb Laurent Pinchart: > > > On Wed, May 11, 2022 at 03:36:14PM +0200, Alexander Stein wrote: > > > > Hello Laurent, > > > > > > > > thanks for pushing out this huge set. > > > > > > > > Am Dienstag, 10. Mai 2022, 13:58:09 CEST schrieb Laurent Pinchart: > > > > > Hello, > > > > > > > > > > This patch series prepares the imx7-media-csi for destaging by > > > > > decoupling it from the helpers shared with the i.MX6 IPUv3. > > > > > > > > > > The strategy Paul and I have followed is to import copies of helper > > > > > code > > > > > and, refactor it within the imx7-media-csi driver, and repeat until > > > > > no > > > > > more shared helpers are used. There is still room for refactoring > > > > > and > > > > > simplification of the imx7-media-csi driver, but I believe it is now > > > > > in > > > > > a state clean enough to be moved out of staging. > > > > > > > > > > The series also includes a few fixes or improvements in supported > > > > > formats that are now made possible thanks to this refactoring. See > > > > > patches 45/50 and 46/50 for details. > > > > > > > > > > The code size has grown as a result. This is partly offset by code > > > > > in > > > > > the shared helpers that can be removed or simplified, but I haven't > > > > > starting working on that. The helpers are now used for the i.MX6 > > > > > IPUv3 > > > > > only, so I will leave this exercise to anyone who would be > > > > > interested in > > > > > destaging that driver as well. > > > > > > > > > > Some of the items in the TODO file related to the imx7-media-csi > > > > > driver > > > > > have been addressed. The two remaining items are frame interval > > > > > monitor > > > > > support and restricting the list of supported formats to the SoC > > > > > version. The former isn't a destaging blocker in my opinion, as the > > > > > feature can be added later if desired (and frame interval monitoring > > > > > should then be moved to the V4L2 core). I believe the latter could > > > > > also > > > > > be addressed after destaging the driver, but in any case, this is a > > > > > discussion for a future destaging series (which may come as soon as > > > > > this > > > > > one is accepted). > > > > > > > > > > Alexander, this also could greatly simplify your "[PATCH v3 0/8] > > > > > imx7/imx8mm media / csi patches" series. > > > > > > > > Thanks for putting me on CC. I'll try to get my system running your > > > > patches > > > > ASAP. > > > > > > As only minor changes were needed, I ended up posting v1.1 of the > > > corresponding patches instead of spamming everybody with a full v2. For > > > your convenience, I've pushed the result to > > > > > > git://linuxtv.org/pinchartl/media.git imx7/destage > > > > > > I know it's a bit of a short notice, but I'd like to send a pull request > > > in time for v5.20, when do you think you would be able to test the > > > series ? > > > > I happen to have my setup already available, so I could test your updated > > series pretty fast. > > Accessing imx327 still works for WIP config. So > > Tested-by: Alexander Stein <alexander.stein@xxxxxxxxxxxxxxx> > > Thanks a lot for the quick testing. I've now sent the pull request. > > Do you plan to submit a new version of your patches on top of this ? The patches regarding the imx[,7]-media-* are not needed anymore, they are obsolete now. There are only patches regarding DT overlay and the Vision Components FPGA and sensors I'm working on right now. Best regards, Alexander
