Re: [PATCH 5/6] media: qcom: camss: enable vfe for Glymur

[Bryan O'Donoghue]

On 01/06/2026 14:55, Dmitry Baryshkov wrote:
> On Mon, Jun 01, 2026 at 06:31:08PM +0530, Nihal Kumar Gupta wrote:
> > On 01-06-2026 17:01, Dmitry Baryshkov wrote:
> > > Is it completely compatible? If so, we can drop the driver parts and
> > > make Glymur fallback to Hamoa.
> >
> > Not fully — Glymur lacks CSIPHY2. So, A direct fallback to x1e80100 would
> > probe a non-existent CSIPHY2 register region.
> >
> > All other resources (VFE, CSID, TPG, ICC) and version-switch paths are identical
> > to x1e80100, so only the csiphy_res table needs to remain Glymur-specific.
>
> We have all the resources in DT. Can we not use DT to determine whether
> the host has CSIPHY2 or not?

With the CSI PHY driver the answer is yes.

With legacy bindings - you depend on driver churn so no.

---
bod