Re: [PATCH v2 1/4] mtd: rawnand: Add support manufacturer specific lock/unlock operatoin

[Boris Brezillon]

On Mon, 17 Feb 2020 10:01:24 +0100
Miquel Raynal <miquel.raynal@xxxxxxxxxxx> wrote:

> Hi Mason,
> 
> masonccyang@xxxxxxxxxxx wrote on Mon, 17 Feb 2020 16:14:23 +0800:
> 
> > Hi Boris,
> >   
> > >     
> > > >  /* Set default functions */
> > > >  static void nand_set_defaults(struct nand_chip *chip)
> > > >  {
> > > > @@ -5782,8 +5810,8 @@ static int nand_scan_tail(struct nand_chip     
> > *chip)  
> > > >     mtd->_read_oob = nand_read_oob;
> > > >     mtd->_write_oob = nand_write_oob;
> > > >     mtd->_sync = nand_sync;
> > > > -   mtd->_lock = NULL;
> > > > -   mtd->_unlock = NULL;
> > > > +   mtd->_lock = nand_lock;
> > > > +   mtd->_unlock = nand_unlock;
> > > >     mtd->_suspend = nand_suspend;
> > > >     mtd->_resume = nand_resume;
> > > >     mtd->_reboot = nand_shutdown;
> > > > diff --git a/include/linux/mtd/rawnand.h b/include/linux/mtd/rawnand.h
> > > > index 4ab9bcc..2430ecd 100644
> > > > --- a/include/linux/mtd/rawnand.h
> > > > +++ b/include/linux/mtd/rawnand.h
> > > > @@ -1136,6 +1136,9 @@ struct nand_chip {
> > > >        const struct nand_manufacturer *desc;
> > > >        void *priv;
> > > >     } manufacturer;
> > > > +
> > > > +   int (*_lock)(struct nand_chip *chip, loff_t ofs, uint64_t len);
> > > > +   int (*_unlock)(struct nand_chip *chip, loff_t ofs, uint64_t len);    
> > > 
> > > Please drop this _ prefix.    
> > 
> > Drop _ prefix of _lock will get compile error due to there is already 
> > defined "struct mutex lock" in struct nand_chip.  
> 
> Right!

Or maybe move all hooks to a sub-struct (struct nand_chip_ops ops). I
had planned to do that in my nand_chip_legacy refactor but never did, so
maybe now is a good time.

> 
> > 
> > What about keep this _ prefix or patch it to blocklock/blockunlock,
> > i.e.,
> > int (*blocklock)(struct nand_chip *chip, loff_t ofs, uint64_t len);
> > int (*blockunlock)(struct nand_chip *chip, loff_t ofs, uint64_t len);  
> 
> What about lock_area() unlock_area() ? Seems more accurate to me, tell
> me if I'm wrong.

Yep, definitely better.