[PATCH 00/19] drm: debugfs: Remove all files automatically on cleanup

[Noralf Trønnes]

Den 01.03.2017 15.31, skrev Daniel Vetter:
> On Fri, Jan 27, 2017 at 03:29:29PM +0100, Daniel Vetter wrote:
>> On Fri, Jan 27, 2017 at 03:23:43PM +0100, Noralf Trønnes wrote:
>>> Den 27.01.2017 08.49, skrev Daniel Vetter:
>>>> On Thu, Jan 26, 2017 at 11:56:02PM +0100, Noralf Trønnes wrote:
>>>>> This patchset removes the need for drivers to clean up their debugfs
>>>>> files on exit. It is done automatically in drm_debugfs_cleanup().
>>>>> This funtion is also called should the driver error out in it's
>>>>> drm_driver.debugfs_init callback.
>>>>>
>>>>> Two drivers still use drm_debugfs_remove_files():
>>>>> - tegra in it's connectors, not sure if I can remove it.
>>>> I read through them, and they're removed on the component device nodes
>>>> stuff. That looks somewhat fishy from a lifetime point of view, and I
>>>> think removing all that code would be better, too.
>>>>
>>>>> - qxl because it's debugfs files list is part of struct qxl_device which
>>>>>     is freed on unload before drm_minor_unregister() is called cleaning debugfs.
>>>> In -next qxl is already demidlayered and there's no longer an unload hook.
>>>> This should be all safe ... why is it not?
>>> My bad, I fetched linux-next a few days ago and figured it was
>>> up-to-date-enough. Duh, I should have known better after following drm for
>>> a year now, it is constantly changing, no pauses.
>>>
>>> Can you please ping me when you have pulled driver patches and I'll respin
>>> msm, tegra and qxl (and others if necessary), and remove the hook.
>>> It's much easier for me to do a small patchset, it's really a strain to get
>>> everything lined up correctly with big changes. I didn't have that patch
>>> juggling class when in school, so I'm late to the game :-)
>> You're doing great with the patch juggling :-) I've just made a pass
>> through the series and merge what's already reviewed/acked.
> Ok, pulled in the remaining patch (I hope, pls ping if I missed one). We
> have only a few debugfs_remove calls left, and I think it's safe to remove
> them all too. Up to do that too? Then we could remove the export, to make
> sure no new driver gets this wrong ...

Yes, they're all in now. I'll fix up what's left.

Noralf.

> Thanks a lot, Daniel