[PATCH] drm/fb-helper: Remove drm_fb_helper_unprepare() from drm_fb_helper_fini()

Javier Martinez Canillas javierm at redhat.com
Tue Feb 21 12:31:54 UTC 2023


Thomas Zimmermann <tzimmermann at suse.de> writes:

> Hi
>
> Am 21.02.23 um 11:27 schrieb Javier Martinez Canillas:
>> Thomas Zimmermann <tzimmermann at suse.de> writes:
>> 
>>> Move drm_fb_helper_unprepare() from drm_fb_helper_fini() into the
>>> calling fbdev implementation. Avoids a possible stale mutex with
>>> generic fbdev code.
>>>
>>> As indicated by its name, drm_fb_helper_prepare() prepares struct
>>> drm_fb_helper before setting up the fbdev support with a call to
>>> drm_fb_helper_init(). In legacy fbdev emulation, this happens next
>>> to each other. If successful, drm_fb_helper_fini() later tear down
>>> the fbdev device and also unprepare via drm_fb_helper_unprepare().
>>>
>>> Generic fbdev emulation prepares struct drm_fb_helper immediately
>>> after allocating the instance. It only calls drm_fb_helper_init()
>>> as part of processing a hotplug event. If the hotplug-handling fails,
>>> it runs drm_fb_helper_fini(). This unprepares the fb-helper instance
>>> and the next hotplug event runs on stale data.
>>>
>>> Solve this by moving drm_fb_helper_unprepare() from drm_fb_helper_fini()
>>> into the fbdev implementations. Call it right before freeing the
>>> fb-helper instance.
>>>
>>> Fixes: 4825797c36da ("drm/fb-helper: Introduce drm_fb_helper_unprepare()")
>> 
>> I think this should be Fixes: 032116bbe152 ("drm/fbdev-generic: Minimize
>> client unregistering") instead? Because commit 4825797c36da just added a
>> wrapper function for mutex_destroy(&fb_helper->lock), but it was commit
>> 032116bbe152 that made drm_fbdev_cleanup() to call that helper function.
>
> Good point. After looking through the recent fbdev commits, I'll use 
> commit 643231b28380 ("drm/fbdev-generic: Minimize hotplug error 
> handling") for the tag. This is the one that added the call to 
> drm_fb_helper_fini() to the client's hotplug handler. And _fini() 
> currently does the _unprepare(), when it shouldn't.
>

Ah, much better indeed.

-- 
Best regards,

Javier Martinez Canillas
Core Platforms
Red Hat



More information about the amd-gfx mailing list