mtd: cfi: Fixed endless loop problem in CFI when value was written but corrupted.

There was an endless loop in CFI Flash driver when a value was written
incorrectly. In such case chip_ready returns true but chip_good returns
false and we never get out of the loop.

The solution was to break the loop in 2 cases, either device is ready or
device is not ready and timeout elapsed. The correctness of the write is
checked after the loop ended. That way we ensure the loop always ends.

Signed-off-by: Przemyslaw Sobon <psobon@amazon.com>
---
 drivers/mtd/chips/cfi_cmdset_0002.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

On Wed, 2019-01-16 at 00:32 +0000, Przemyslaw Sobon wrote:
> 
> 
> There was an endless loop in CFI Flash driver when a value was written
> incorrectly. In such case chip_ready returns true but chip_good returns
> false and we never get out of the loop.
> 
> The solution was to break the loop in 2 cases, either device is ready or
> device is not ready and timeout elapsed. The correctness of the write is
> checked after the loop ended. That way we ensure the loop always ends.
> 
> Signed-off-by: Przemyslaw Sobon <psobon@amazon.com>


hmm, current code was introduced by Tokunori Ikegami <ikegami@allied-telesis.co.jp> to address another problem he had.
See 
   mtd: cfi_cmdset_0002: Change write buffer to check correct value
and
   mtd: cfi_cmdset_0002: Change erase functions to check chip good only

I wonder if you need to wrap an extra loop with retries around chip_good to adress the problem Tokunori had.

Tokunori, what do you think ?

Jocke
> ---
>  drivers/mtd/chips/cfi_cmdset_0002.c | 11 +++++++----
>  1 file changed, 7 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/mtd/chips/cfi_cmdset_0002.c b/drivers/mtd/chips/cfi_cmdset_0002.c
> index 72428b6bfc47..6cc31d2057e9 100644
> --- a/drivers/mtd/chips/cfi_cmdset_0002.c
> +++ b/drivers/mtd/chips/cfi_cmdset_0002.c
> @@ -1879,15 +1879,18 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip,
>                 if (time_after(jiffies, timeo) && !chip_ready(map, adr))
>                         break;
> 
> -               if (chip_good(map, adr, datum)) {
> -                       xip_enable(map, chip, adr);
> -                       goto op_done;
> -               }
> +               if (chip_ready(map, adr))
> +                       break;
> 
>                 /* Latency issues. Drop the lock, wait a while and retry */
>                 UDELAY(map, chip, adr, 1);
>         }
> 
> +       if (chip_good(map, adr, datum)) {
> +               xip_enable(map, chip, adr);
> +               goto op_done;
> +       }
> +
>         /*
>          * Recovery from write-buffer programming failures requires
>          * the write-to-buffer-reset sequence.  Since the last part
> --
> 2.16.5
> 
> 
> ______________________________________________________
> Linux MTD discussion mailing list
> https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Flists.infradead.org%2Fmailman%2Flistinfo%2Flinux-mtd%2F&amp;data=02%7C01%7Cjoakim.tjernlund%40infinera.com%7C35b64c743938427ffa7208d67b4a2a60%7C285643de5f5b4b03a1530ae2dc8aaf77%7C1%7C1%7C636831955813341722&amp;sdata=TgI7aw8Qv57MY%2B62KWS87kyfte2A8qQY5OFjQc9Vhwc%3D&amp;reserved=0

On Wed, 2019-01-16 at 08:33 +0000, Joakim Tjernlund wrote:
> CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.
> 
> 
> On Wed, 2019-01-16 at 00:32 +0000, Przemyslaw Sobon wrote:
> > 
> > There was an endless loop in CFI Flash driver when a value was written
> > incorrectly. In such case chip_ready returns true but chip_good returns
> > false and we never get out of the loop.
> > 
> > The solution was to break the loop in 2 cases, either device is ready or
> > device is not ready and timeout elapsed. The correctness of the write is
> > checked after the loop ended. That way we ensure the loop always ends.
> > 
> > Signed-off-by: Przemyslaw Sobon <psobon@amazon.com>
> 
> hmm, current code was introduced by Tokunori Ikegami <ikegami@allied-telesis.co.jp> to address another problem he had.

Seems like Tokunori Ikegami's email address is invalid now ...

-----Original Message-----
From: Joakim Tjernlund <Joakim.Tjernlund@infinera.com> 
Sent: Wednesday, January 16, 2019 12:33 AM
To: linux-mtd@lists.infradead.org; computersforpeace@gmail.com; ikegami@allied-telesis.co.jp; Sobon, Przemyslaw <psobon@amazon.com>; dwmw2@infradead.org; richard@nod.at; marek.vasut@gmail.com
Subject: Re: [PATCH] mtd: cfi: Fixed endless loop problem in CFI when value was written but corrupted.

> On Wed, 2019-01-16 at 00:32 +0000, Przemyslaw Sobon wrote:
> > 
> > 
> > There was an endless loop in CFI Flash driver when a value was written 
> > incorrectly. In such case chip_ready returns true but chip_good 
> > returns false and we never get out of the loop.
> > 
> > The solution was to break the loop in 2 cases, either device is ready 
> > or device is not ready and timeout elapsed. The correctness of the 
> > write is checked after the loop ended. That way we ensure the loop always ends.
> > 
> > Signed-off-by: Przemyslaw Sobon <psobon@amazon.com>
> 
> 
> hmm, current code was introduced by Tokunori Ikegami <ikegami@allied-telesis.co.jp> to address another problem he had.
> See 
>    mtd: cfi_cmdset_0002: Change write buffer to check correct value and
>    mtd: cfi_cmdset_0002: Change erase functions to check chip good only
> 
> I wonder if you need to wrap an extra loop with retries around chip_good
> to adress the problem Tokunori had.
If we add "time_after" loop and write is complete but wrong value was written we would have
to wait specific amount of time anyway. Example: we try to write value 4 at address 0x100, the write itself
is done (chip is in ready state) after 10 us but value written was 3 (wrong value). In my proposal the loop
will end after 10 us and we would check if value written is correct and if not we would return error. The
execution time of the loop would be 10 us then. If we surround chip_good with a loop and consider above
situation we will wait whatever the loop is set to e.g. 1ms even though the write was done after 10 us.
This is because we always read value 3 and we retry until timeout elapses.
> 
> Tokunori, what do you think ?
> 
> Jocke
> > ---
> >  drivers/mtd/chips/cfi_cmdset_0002.c | 11 +++++++----
> >  1 file changed, 7 insertions(+), 4 deletions(-)
> > 
> > diff --git a/drivers/mtd/chips/cfi_cmdset_0002.c 
> > b/drivers/mtd/chips/cfi_cmdset_0002.c
> > index 72428b6bfc47..6cc31d2057e9 100644
> > --- a/drivers/mtd/chips/cfi_cmdset_0002.c
> > +++ b/drivers/mtd/chips/cfi_cmdset_0002.c
> > @@ -1879,15 +1879,18 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip,
> >                 if (time_after(jiffies, timeo) && !chip_ready(map, adr))
> >                         break;
> > 
> > -               if (chip_good(map, adr, datum)) {
> > -                       xip_enable(map, chip, adr);
> > -                       goto op_done;
> > -               }
> > +               if (chip_ready(map, adr))
> > +                       break;
> > 
> >                 /* Latency issues. Drop the lock, wait a while and retry */
> >                 UDELAY(map, chip, adr, 1);
> >         }
> > 
> > +       if (chip_good(map, adr, datum)) {
> > +               xip_enable(map, chip, adr);
> > +               goto op_done;
> > +       }
> > +
> >         /*
> >          * Recovery from write-buffer programming failures requires
> >          * the write-to-buffer-reset sequence.  Since the last part
> > --
> > 2.16.5
> > 
> > 
> > ______________________________________________________
> > Linux MTD discussion mailing list
> > https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Flists.
> > infradead.org%2Fmailman%2Flistinfo%2Flinux-mtd%2F&amp;data=02%7C01%7Cj
> > oakim.tjernlund%40infinera.com%7C35b64c743938427ffa7208d67b4a2a60%7C28
> > 5643de5f5b4b03a1530ae2dc8aaf77%7C1%7C1%7C636831955813341722&amp;sdata=
> > TgI7aw8Qv57MY%2B62KWS87kyfte2A8qQY5OFjQc9Vhwc%3D&amp;reserved=0
> 
>