> > On Wed, 2022-01-12 at 14:35 +0100, Lorenzo Bianconi wrote: > > > > > > > > @@ -302,6 +316,15 @@ struct mt7915_dev { > > > > u8 table_mask; > > > > u8 n_agrt; > > > > } twt; > > > > + > > > > + struct reset_control *rstc; > > > > + void __iomem *dcm; > > > > + void __iomem *sku; > > > > + > > > > + struct { > > > > + bool is_7975; > > > > + bool is_7976; > > > > + } adie[MT7986_MAX_ADIE_NUM]; > > > > > > do we really need it? Can we just read data from chip when necessary? > > > it is not > > > access in the hot-path, right? > > > I think it is easier and more readable. > > > > Adie registers are inaccessible after wfsys power on so we need some > > places to store chip configuration. > > We will modify it to make it more readable. > > + u32 adie; > > what I mean is just continue reading these values after the first time (that is > in common for both approaches). ok, sorry I misread your reply. Anyway it seems to me we need these values just in mt7986_wmac_enable() (or in routines called by mt7986_wmac_enable). Right? If so maybe we can just store them in a local struct in mt7986_wmac_enable(). What do you think? Regards, Lorenzo > > Regards, > Lorenzo > > > > > thanks, > > Peter > >
Attachment:
signature.asc
Description: PGP signature