On Thu, Apr 04, 2024 at 10:20:29AM +0200, Bartosz Golaszewski wrote:
> On Wed, Apr 3, 2024 at 3:15 PM Kent Gibson <warthog618@xxxxxxxxx> wrote:
> >
> > When adding sanitization of the label, the path through
> > edge_detector_setup() that leads to debounce_setup() was overlooked.
> > A request taking this path does not allocate a new label and the
> > request label is freed twice when the request is released, resulting
> > in memory corruption.
> >
> > Add label sanitization to debounce_setup().
> >
> > Cc: stable@xxxxxxxxxxxxxxx
> > Fixes: b34490879baa ("gpio: cdev: sanitize the label before requesting the interrupt")
> > Signed-off-by: Kent Gibson <warthog618@xxxxxxxxx>
> > ---
> > drivers/gpio/gpiolib-cdev.c | 31 +++++++++++++++++++------------
> > 1 file changed, 19 insertions(+), 12 deletions(-)
> >
> > diff --git a/drivers/gpio/gpiolib-cdev.c b/drivers/gpio/gpiolib-cdev.c
> > index fa9635610251..f4c2da2041e5 100644
> > --- a/drivers/gpio/gpiolib-cdev.c
> > +++ b/drivers/gpio/gpiolib-cdev.c
> > @@ -728,6 +728,16 @@ static u32 line_event_id(int level)
> > GPIO_V2_LINE_EVENT_FALLING_EDGE;
> > }
> >
> > +static inline char *make_irq_label(const char *orig)
> > +{
> > + return kstrdup_and_replace(orig, '/', ':', GFP_KERNEL);
> > +}
> > +
> > +static inline void free_irq_label(const char *label)
> > +{
> > + kfree(label);
> > +}
> > +
> > #ifdef CONFIG_HTE
> >
> > static enum hte_return process_hw_ts_thread(void *p)
> > @@ -1015,6 +1025,7 @@ static int debounce_setup(struct line *line, unsigned int debounce_period_us)
> > {
> > unsigned long irqflags;
> > int ret, level, irq;
> > + char *label;
> >
> > /* try hardware */
> > ret = gpiod_set_debounce(line->desc, debounce_period_us);
> > @@ -1037,11 +1048,17 @@ static int debounce_setup(struct line *line, unsigned int debounce_period_us)
> > if (irq < 0)
> > return -ENXIO;
> >
> > + label = make_irq_label(line->req->label);
>
> Now that I look at the actual patch, I don't really like it. We
> introduce a bug just to fix it a commit later. Such things have been
> frowned upon in the past.
>
> Let me shuffle the code a bit, I'll try to make it a bit more correct.
>
The debounce_setup() oversight bug is the more severe, so it makes more
sense to me to fix it first. But then I my preferred solution would be
to pull the original patch and submit a corrected patch that merges all
three, so no bugs, but I assume that isn't an option.
Cheers,
Kent.
