On Mon, Feb 10, 2025 at 04:15:38PM +0100, Miquel Raynal wrote: > Hello Christian, > > On 09/02/2025 at 15:54:32 +01, Christian Marangi <ansuelsmth@xxxxxxxxx> wrote: > > > With some research in some obscure old QSDK, it was possible to find the > > MASK of the last register there were still set with raw shift and > > convert them to FIELD_PREP API. > > > > This is only a cleanup and modernize the code a bit and doesn't make > > any behaviour change. > > > > Signed-off-by: Christian Marangi <ansuelsmth@xxxxxxxxx> > > --- > > drivers/mtd/nand/raw/qcom_nandc.c | 36 ++++++++++++++-------------- > > I'm fine with your two patches. I was about to apply them, but the first > one needs to go through fixes, whereas the second through next, and they > are dependent on each other. I propose the following modification: > - create patch 1/2 with the content of the cleanup done just below, but > only adapted to the very specific spot that is touched by the fix "fix > broken config...". It would be a prerequisite for the fix. > - patch 2/2 would be the content of "fix broken config..." > > And aside, a totally independent patch easy to apply on -rc1 with the > rest of this patch. > > Would that work for you? > Mhhh are they really dependent on each other? I posted them in 2 separate patch as one should have priority and be applied ASAP. The other is really a cleanup and from what I can see no delta in the patch gets affected by the fix in the other patch. In theory they should apply independently. An alternative solution might be to just delay the cleanup patch and post/merge it later in some week? Open to any suggetion to better handle this but I feel they don't conflict on each other (please confirm if I'm wrong about this) -- Ansuel
