Re: [PATCH v4 09/14] drm/tidss: oldi: Fix cleanup on a failed init

[Tomi Valkeinen]

Hi,

On 01/11/2026 16:14, Aradhya Bhatia wrote:
> Hi Tomi,
>
> On 29/10/26 10:18, Tomi Valkeinen wrote:
> > tidss_oldi_init() registers a bridge for each OLDI TX in turn. If a later
> > OLDI TX fails, the bridges registered for the earlier ones are not
> > removed.
> >
> > Fix this by calling tidss_oldi_deinit() in the error path.
> >
> > Fixes: 7246e0929945 ("drm/tidss: Add OLDI bridge support")
> > Signed-off-by: Tomi Valkeinen <tomi.valkeinen@xxxxxxxxxxxxxxxx>
> > ---
> >   drivers/gpu/drm/tidss/tidss_oldi.c | 1 +
> >   1 file changed, 1 insertion(+)
> >
> > diff --git a/drivers/gpu/drm/tidss/tidss_oldi.c b/drivers/gpu/drm/tidss/tidss_oldi.c
> > index 506507e83112..7344d86e1f13 100644
> > --- a/drivers/gpu/drm/tidss/tidss_oldi.c
> > +++ b/drivers/gpu/drm/tidss/tidss_oldi.c
> > @@ -638,5 +638,6 @@ int tidss_oldi_init(struct tidss_device *tidss)
> >   err_put_node:
> >   	of_node_put(child);
> >   	of_node_put(oldi_parent);
> > +	tidss_oldi_deinit(tidss);
> >   	return ret;
> >   }
>
> Since tidss_probe() (in tidss_drv.c) calls tidss_oldi_init(), it also
> calls tidss_oldi_deinit() in its error exit path.
Yes, but not if tidss_oldi_init() itself fails.

If we have two OLDI TXes, and the loop in tidss_oldi_init() inits the 
first one, then the second one's init fails, the code jumps to 
err_put_node. If we don't call tidss_oldi_deinit() there, the first OLDI 
TX will not be "deinited".

 Tomi