lkml.org 
[lkml]   [2020]   [Jun]   [9]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
Date
SubjectRe: [PATCH v4 12/13] gpiolib: add new ioctl() for monitoring changes in line info
wt., 9 cze 2020 o 02:23 Kent Gibson <warthog618@gmail.com> napisał(a):
>
> On Tue, Dec 24, 2019 at 01:07:08PM +0100, Bartosz Golaszewski wrote:
> > From: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> >
>
> [snip!]
>
> > +static int lineinfo_changed_notify(struct notifier_block *nb,
> > + unsigned long action, void *data)
> > +{
> > + struct gpio_chardev_data *priv = to_gpio_chardev_data(nb);
> > + struct gpioline_info_changed chg;
> > + struct gpio_desc *desc = data;
> > + int ret;
> > +
> > + if (!test_bit(desc_to_gpio(desc), priv->watched_lines))
> > + return NOTIFY_DONE;
> > +
> > + memset(&chg, 0, sizeof(chg));
> > + chg.info.line_offset = gpio_chip_hwgpio(desc);
> > + chg.event_type = action;
> > + chg.timestamp = ktime_get_real_ns();
> > + gpio_desc_to_lineinfo(desc, &chg.info);
> > +
>
> Is this call legal? It can sleep - in fact you recently changed that
> very function to move a mutex call outside of a spinlock protected section.
> Here it is being called within an RCU lock, as lineinfo_changed_notify
> is at the end of an atomic_notifier_call_chain.
>

Yeah, this is clearly wrong and lockdep would complain about invalid
wait context. I'm trying to remember why I went for an atomic notifier
chain here though... At first glance it doesn't look like the chain
could be called from atomic context anywhere.

Bart

\
 
 \ /
  Last update: 2020-06-09 09:59    [W:0.217 / U:0.512 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site