[PATCH] fbdev: Introduce devm_register_framebuffer()

[Thomas Weißschuh]

On 2024-08-30 12:16:46+0000, Helge Deller wrote:
> On 8/30/24 11:45, Thomas Weißschuh wrote:
> > Introduce a device-managed variant of register_framebuffer() which
> > automatically unregisters the framebuffer on device destruction.
> > This can simplify the error handling and resource management in drivers.
> > 
> > Signed-off-by: Thomas Weißschuh <linux at weissschuh.net>
> > ---
> > This is a fixed resend of [0], which was broken.
> > Thanks to Bert [1], and Chaitanya Kumar [2]
> > for reporting the issue.
> > 
> > [0] https://lore.kernel.org/lkml/20240827-efifb-sysfs-v1-3-c9cc3e052180@weissschuh.net/
> > [1] https://lore.kernel.org/lkml/20240829224124.2978-1-spasswolf@web.de/
> > [2] https://lore.kernel.org/lkml/SJ1PR11MB612925C1C533C09F8F62F7CBB9972@SJ1PR11MB6129.namprd11.prod.outlook.com/
> 
> I've applied this patch to the fbdev git tree.
> Please double check at
> https://git.kernel.org/pub/scm/linux/kernel/git/deller/linux-fbdev.git/log/?h=for-next

Looks good.
Sorry for the extra effort necessary.

> Can you please check if this fixes this new report too:
> https://marc.info/?l=linux-fbdev&m=172500784802901&w=2

I double check it and am decently sure it's the same issue.
It tries to unregister the framebuffer but then notices that it never
was registered in the first place.

> > Helge, I didn't document the function devm_unregister_framebuffer() as
> > it is only an internal helper and will ever only used by one user,
> > similar to other helpers in fbmem.c.
> 
> Ok.
> 
> Helge