On 2017-09-06 18:24, Weijie Gao wrote: > Yes, I have to admit operations of GigaDevice's chips are the same. > All the chips have the same read/write/erase command sequences. > Differences are plane selection bits (Micron), die selection > (Micron/Winbond), buffered read/continuous read (Winbond) and ECC > status bits. > These differences can be abstracted to different functions chosen by > the flash id. So this won't be a problem. > > As for the patches, may be I can submit patches to both generic-4.4 > and generic-4.9, and remove patches from pistachio. > So this won't be a problem when ar71xx updates linux-4.9. I think it's a bad idea to integrate new code that is based on an old framework rejected by upstream. Even if it's more work, I strongly suggest using the new code by Micron. If there's no way around using the old code for now, it should definitely not be placed in generic/ because this could potentially lead to even more people picking it up, thus creating long term porting issues.
- Felix _______________________________________________ Lede-dev mailing list Lede-dev@lists.infradead.org http://lists.infradead.org/mailman/listinfo/lede-dev
