On Fri, Jun 10, 2022 at 10:00:03AM +0200, Johan Hovold wrote:
> On Thu, Jun 09, 2022 at 02:45:18PM -0700, Jared Kangas wrote:
> > gbaudio_dapm_free_controls() iterates over widgets using the
> > list_for_each_entry*() family of macros from <linux/list.h>, which
> > leaves the loop cursor pointing to a meaningless structure if it
> > completes a traversal of the list. The cursor was set to NULL at the end
> > of the loop body, but would be overwritten by the final loop cursor
> > update.
> >
> > Because of this behavior, the widget could be non-null after the loop
> > even if the widget wasn't found, and the cleanup logic would treat the
> > pointer as a valid widget to free.
> >
> > To fix this, introduce a temporary variable to act as the loop cursor
> > and copy it to a variable that can be accessed after the loop finishes.
> > Due to not removing any list elements, use list_for_each_entry() instead
> > of list_for_each_entry_safe() in the revised loop.
> >
> > This was detected with the help of Coccinelle.
> >
> > Fixes: 510e340efe0c ("staging: greybus: audio: Add helper APIs for dynamic audio modules")
> > Cc: stable@xxxxxxxxxxxxxxx
> > Reviewed-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
> > Reviewed-by: Johan Hovold <johan@xxxxxxxxxx>
> > Signed-off-by: Jared Kangas <kangas.jd@xxxxxxxxx>
> > ---
> >
> > Changes since v1:
> > * Removed safe list iteration as suggested by Johan Hovold <johan@xxxxxxxxxx>
> > * Updated patch changelog to explain the list iteration change
> > * Added tags to changelog based on feedback (Cc:, Fixes:, Reviewed-by:)
>
> Apparently Greg applied this to staging-next before we had a change to
> look at it. You should have received a notification from Greg when he
> did so.
>
> https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging.git/commit/?h=staging-next&id=80c968a04a381dc0e690960c60ffd6b6aee7e157
>
> It seems unlikely that this would cause any issues in real life, but
> there's still a chance it will be picked up by the stable team despite
> the lack of a CC stable tag.
>
> I've just sent a follow-up patch to replace the list macro.
>
> Johan
Sorry about that - I got a notification but thought it was still
revisable. In hindsight, it makes sense that once it gets applied to
a public branch, changes should be done in additional patches. Thanks to
both you and Dan for taking the time to review and catch my mistakes.
Jared
