Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1702542 > unrolled thread

Re: [PATCH] zram: Fix buffer size passed to strlcpy()

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: [PATCH] zram: Fix buffer size passed to strlcpy() Doug Anderson <dianders@chromium.org> - 2017-08-03 01:00 +0200
    Re: [PATCH] zram: Fix buffer size passed to strlcpy() Minchan Kim <minchan@kernel.org> - 2017-08-03 01:50 +0200
      Re: [PATCH] zram: Fix buffer size passed to strlcpy() Matthias Kaehlcke <mka@chromium.org> - 2017-08-03 02:00 +0200

#1702542 — Re: [PATCH] zram: Fix buffer size passed to strlcpy()

Hi,

On Fri, Jul 28, 2017 at 10:12 AM, Matthias Kaehlcke <mka@chromium.org> wrote:
> comp_algorithm_store() passes the size of the source buffer to strlcpy()
> instead of the destination buffer size, fix this.

This was introduced in commit 415403be37e2 ("zram: use crypto api to
check alg availability"), but probably don't need a "Fixes" since
there's not really a bug (see below)

> Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
> ---
>  drivers/block/zram/zram_drv.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c
> index 856d5dc02451..7d2ddffad361 100644
> --- a/drivers/block/zram/zram_drv.c
> +++ b/drivers/block/zram/zram_drv.c
> @@ -327,7 +327,7 @@ static ssize_t comp_algorithm_store(struct device *dev,
>                 return -EBUSY;
>         }
>
> -       strlcpy(zram->compressor, compressor, sizeof(compressor));
> +       strlcpy(zram->compressor, compressor, sizeof(zram->compressor));

As far as I can tell the two sizes are identical.  In struct zram:

char compressor[CRYPTO_MAX_ALG_NAME];

Locally here:

char compressor[CRYPTO_MAX_ALG_NAME];

...so there is no bug per say unless there's a hidden "#undef".
...but your change does make it a little clearer, plus if someone ever
changed one of these arrays it would be safer.  Thus:

Reviewed-by: Douglas Anderson <dianders@chromium.org>


I suppose another option would be to define the local array based on
the size of the structure.  AKA locally in the function:

  char compressor[ARRAY_SIZE(zram->compressor)];

...if you did that you could replace the strlcpy() below with a simple
strcpy() since you'd be guaranteed that there's be enough space.
...but I'm probably overthinking it too much. ;-P


-Doug

[toc] | [next] | [standalone]

#1702572

Hi Doug,

On Wed, Aug 02, 2017 at 03:54:32PM -0700, Doug Anderson wrote:
> Hi,
> 
> On Fri, Jul 28, 2017 at 10:12 AM, Matthias Kaehlcke <mka@chromium.org> wrote:
> > comp_algorithm_store() passes the size of the source buffer to strlcpy()
> > instead of the destination buffer size, fix this.
> 
> This was introduced in commit 415403be37e2 ("zram: use crypto api to
> check alg availability"), but probably don't need a "Fixes" since
> there's not really a bug (see below)
> 
> > Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
> > ---
> >  drivers/block/zram/zram_drv.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c
> > index 856d5dc02451..7d2ddffad361 100644
> > --- a/drivers/block/zram/zram_drv.c
> > +++ b/drivers/block/zram/zram_drv.c
> > @@ -327,7 +327,7 @@ static ssize_t comp_algorithm_store(struct device *dev,
> >                 return -EBUSY;
> >         }
> >
> > -       strlcpy(zram->compressor, compressor, sizeof(compressor));
> > +       strlcpy(zram->compressor, compressor, sizeof(zram->compressor));
> 
> As far as I can tell the two sizes are identical.  In struct zram:
> 
> char compressor[CRYPTO_MAX_ALG_NAME];
> 
> Locally here:
> 
> char compressor[CRYPTO_MAX_ALG_NAME];
> 
> ...so there is no bug per say unless there's a hidden "#undef".
> ...but your change does make it a little clearer, plus if someone ever
> changed one of these arrays it would be safer.  Thus:
> 
> Reviewed-by: Douglas Anderson <dianders@chromium.org>
> 
> 
> I suppose another option would be to define the local array based on
> the size of the structure.  AKA locally in the function:
> 
>   char compressor[ARRAY_SIZE(zram->compressor)];
> 
> ...if you did that you could replace the strlcpy() below with a simple
> strcpy() since you'd be guaranteed that there's be enough space.
> ...but I'm probably overthinking it too much. ;-P

First of all, Thanks for the patch, Matthias. You are correct and you
patch doesn't have any problem. However, I think Doug's suggestion
looks better. Could you mind resending?

Thanks.

[toc] | [prev] | [next] | [standalone]

#1702581

El Thu, Aug 03, 2017 at 08:44:37AM +0900 Minchan Kim ha dit:

> Hi Doug,
> 
> On Wed, Aug 02, 2017 at 03:54:32PM -0700, Doug Anderson wrote:
> > Hi,
> > 
> > On Fri, Jul 28, 2017 at 10:12 AM, Matthias Kaehlcke <mka@chromium.org> wrote:
> > > comp_algorithm_store() passes the size of the source buffer to strlcpy()
> > > instead of the destination buffer size, fix this.
> > 
> > This was introduced in commit 415403be37e2 ("zram: use crypto api to
> > check alg availability"), but probably don't need a "Fixes" since
> > there's not really a bug (see below)
> > 
> > > Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
> > > ---
> > >  drivers/block/zram/zram_drv.c | 2 +-
> > >  1 file changed, 1 insertion(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c
> > > index 856d5dc02451..7d2ddffad361 100644
> > > --- a/drivers/block/zram/zram_drv.c
> > > +++ b/drivers/block/zram/zram_drv.c
> > > @@ -327,7 +327,7 @@ static ssize_t comp_algorithm_store(struct device *dev,
> > >                 return -EBUSY;
> > >         }
> > >
> > > -       strlcpy(zram->compressor, compressor, sizeof(compressor));
> > > +       strlcpy(zram->compressor, compressor, sizeof(zram->compressor));
> > 
> > As far as I can tell the two sizes are identical.  In struct zram:
> > 
> > char compressor[CRYPTO_MAX_ALG_NAME];
> > 
> > Locally here:
> > 
> > char compressor[CRYPTO_MAX_ALG_NAME];
> > 
> > ...so there is no bug per say unless there's a hidden "#undef".
> > ...but your change does make it a little clearer, plus if someone ever
> > changed one of these arrays it would be safer.  Thus:
> > 
> > Reviewed-by: Douglas Anderson <dianders@chromium.org>
> > 
> > 
> > I suppose another option would be to define the local array based on
> > the size of the structure.  AKA locally in the function:
> > 
> >   char compressor[ARRAY_SIZE(zram->compressor)];
> > 
> > ...if you did that you could replace the strlcpy() below with a simple
> > strcpy() since you'd be guaranteed that there's be enough space.
> > ...but I'm probably overthinking it too much. ;-P
> 
> First of all, Thanks for the patch, Matthias. You are correct and you
> patch doesn't have any problem. However, I think Doug's suggestion
> looks better. Could you mind resending?

Sure, I can rework the patch.

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web