Re: [PATCH 1/6] nand_wait_ready timeout fix

[Artem Bityutskiy <dedekind1@gmail.com>]

On Tue, 2011-06-28 at 17:09 +0200, Matthieu CASTET wrote:
> >>  	struct nand_chip *chip = mtd->priv;
> >> -	unsigned long timeo = jiffies + 2;
> >> +	unsigned long timeo = jiffies + (2 * HZ) / 1000;
> > 
> > I agree that the code is buggy, but your fix is strange: if HZ = 100, (2
> > * HZ) / 1000 will be zero?
> > 
> > I think you should instead know for how many msecs you want to wait, and
> > use msecs_to_jiffies().
> > 
> Your right, I assume the code was written for HZ=100, and the code should wait
> for 20 ms :
> unsigned long timeo = jiffies + (20 * HZ) / 1000;
> 
> this will match with the
> static int nand_wait(struct mtd_info *mtd, struct nand_chip *chip)
> {
> 
>     unsigned long timeo = jiffies;
>     int status, state = chip->state;
> 
>     if (state == FL_ERASING)
>         timeo += (HZ * 400) / 1000;
>     else
>         timeo += (HZ * 20) / 1000;
> [...]
> }

I think we should use msecs_to_jiffies() in nand_wait() as well. Why
relying on weird calculations with HZ?

-- 
Best Regards,
Artem Bityutskiy