On Mon, Jul 29, 2024 at 01:00:47PM GMT, Markus Elfring wrote:
> …
> > +++ b/drivers/gpu/drm/tiny/sharp-memory.c
> > @@ -0,0 +1,684 @@
> …
> > static int sharp_memory_update_display(struct sharp_memory_device *smd,
> > + struct drm_framebuffer *fb,
> > + struct drm_rect clip,
> > + struct drm_format_conv_state *fmtcnv_state)
> > +{
> …
> > + mutex_lock(&smd->tx_mutex);
> > +
> > + /* Populate the transmit buffer with frame data */
> …
> > + mutex_unlock(&smd->tx_mutex);
> > +
> > + return ret;
> > +}
> …
>
> Under which circumstances would you become interested to apply a statement
> like “guard(mutex)(&smd->tx_mutex);”?
> https://elixir.bootlin.com/linux/v6.10.2/source/include/linux/mutex.h#L196
>
Ah, I didn't realize guarded mutexes were implemented. That's really
good to know.
I'd usually use them when it's possible to return before mutex_unlock is
called to avoid goto cleanup logic. But it's probably best to just use them
by default.
Thanks for the review. Will clean up in v4.
Best regards,
Alex
