Hi Marek.
On Wed, Jul 29, 2020 at 07:02:51PM +0200, Marek Vasut wrote:
> On 7/29/20 6:56 PM, Sam Ravnborg wrote:
> [...]
> >> +static int tc358762_probe(struct mipi_dsi_device *dsi)
> >> +{
> >> + struct device *dev = &dsi->dev;
> >> + struct tc358762 *ctx;
> >> + int ret;
> >> +
> >> + ctx = devm_kzalloc(dev, sizeof(struct tc358762), GFP_KERNEL);
> >> + if (!ctx)
> >> + return -ENOMEM;
> >> +
> >> + mipi_dsi_set_drvdata(dsi, ctx);
> > I cannot see why this is needed. If not used then delete it.
> > With the above addressed:
> > Reviewed-by: Sam Ravnborg <sam@xxxxxxxxxxxx>
> >
> > Let me know if I shall apply - but we have to wait until binding is
> > reviewed. And preferably one extra set of eyes on this driver.
>
> [...]
>
> >> +static int tc358762_remove(struct mipi_dsi_device *dsi)
> >> +{
> >> + struct tc358762 *ctx = mipi_dsi_get_drvdata(dsi);
> >> +
> >> + mipi_dsi_detach(dsi);
> >> + drm_bridge_remove(&ctx->bridge);
>
> Because it's used here in the remove callback.
Thanks - missed that.
So consider the point addressed by enlightning me.
Sam
