Re: [PATCH] mtd:clear cache_state to avoid writing to bad clocks repeatedly

[Miquel Raynal <miquel.raynal@xxxxxxxxxxx>]

Hi Xiaoming,

Xiaoming Ni <nixiaoming@xxxxxxxxxx> wrote on Mon, 30 Mar 2020 16:44:29
+0800:

> On 2020/3/30 15:53, Miquel Raynal wrote:
> > Hi Xiaoming,
> > 
> > Xiaoming Ni <nixiaoming@xxxxxxxxxx> wrote on Sat, 28 Mar 2020 21:01:17
> > +0800:
> >   
> >> The function call process is as follows:
> >> 	mtd_blktrans_work()
> >> 	  while (1)  
> ....
> 
> >> +	 *
> >> +	 * if this cache_offset points to a bad block  
> > 
> > Can you start your sentences with a capital letter please?
> > 
> > 	 * If
> >   
> >> +	 * data cannot be written to the device.
> >> +	 * clear cache_state to avoid writing to bad clocks repeatedly  
> > 
> > 	 * Clear
> > 
> > And also please break your lines à 80, not 70.
> >   
> >>   	 */
> >> -	mtdblk->cache_state = STATE_EMPTY;
> >> -	return 0;
> >> +	if (ret == 0 || ret == -EIO)
> >> +		mtdblk->cache_state = STATE_EMPTY;  
> Should I add a warning print for EIO here

I don't think so.

Thanks,
Miquèl

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/