Nouveau - Apr 2021 - [PATCH v3 03/20] drm/dp: Move i2c init to drm_dp_aux_init, add __must_check and fini

On Tue, 2021-04-20 at 02:16 +0300, Ville Syrj?l? wrote:
> 
> The init vs. register split is intentional. Registering the thing
> and allowing userspace access to it before the rest of the driver
> is ready isn't particularly great. For a while now we've tried to
> move towards an architecture where the driver is fully initialzied
> before anything gets exposed to userspace.
Yeah-thank you for pointing this out. Thierry - do you think there's an
alternate solution we could go with in Tegra to fix the get_device() issue
that wouldn't require us trying to expose the i2c adapter early?
> 
-- 
Cheers,
 Lyude Paul (she/her)
 Software Engineer at Red Hat

OK - talked with Ville a bit on this and did some of my own research, I
actually think that moving i2c to drm_dp_aux_init() is the right decision for
the time being. The reasoning behind this being that as shown by my previous
work of fixing drivers that call drm_dp_aux_register() too early - it seems
like there's already been drivers that have been working just fine with
setting up the i2c device before DRM registration. 

In the future, it'd probably be better if we can split up i2c_add_adapter()
into an init and register function - but we'll have to talk with the i2c
maintainers to see if this is acceptable w/ them

On Thu, 2021-04-22 at 13:18 -0400, Lyude Paul wrote:
> On Tue, 2021-04-20 at 02:16 +0300, Ville Syrj?l? wrote:
> > 
> > The init vs. register split is intentional. Registering the thing
> > and allowing userspace access to it before the rest of the driver
> > is ready isn't particularly great. For a while now we've tried
to
> > move towards an architecture where the driver is fully initialzied
> > before anything gets exposed to userspace.
> 
> Yeah-thank you for pointing this out. Thierry - do you think there's an
> alternate solution we could go with in Tegra to fix the get_device() issue
> that wouldn't require us trying to expose the i2c adapter early?
> 
> > 
> 
-- 
Cheers,
 Lyude Paul (she/her)
 Software Engineer at Red Hat

On Thu, Apr 22, 2021 at 01:18:09PM -0400, Lyude Paul
wrote:
> On Tue, 2021-04-20 at 02:16 +0300, Ville Syrj?l? wrote:
> > 
> > The init vs. register split is intentional. Registering the thing
> > and allowing userspace access to it before the rest of the driver
> > is ready isn't particularly great. For a while now we've tried
to
> > move towards an architecture where the driver is fully initialzied
> > before anything gets exposed to userspace.
> 
> Yeah-thank you for pointing this out. Thierry - do you think there's an
> alternate solution we could go with in Tegra to fix the get_device() issue
> that wouldn't require us trying to expose the i2c adapter early?
I suppose we could do it in a hackish way that grabs a reference to the
I2C adapter only upon registration. We can't do that for the regular I2C
DDC case where the I2C controller is an external one because by the time
we get to registration it could've gone again. This would make both code
paths asymmetric, so I'd prefer not to do it. Perhaps it could serve as
an stop-gap solution until something better is in place, though.

Thierry
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20210423/2b3770bc/attachment-0001.sig>