[PATCH v11 16/24] arm64: dts: rockchip: rk356x: Add VOP2 nodes

Heiko Stuebner heiko at sntech.de
Fri May 6 01:54:18 PDT 2022


Am Freitag, 6. Mai 2022, 09:10:28 CEST schrieb Sascha Hauer:
> On Thu, May 05, 2022 at 09:23:03AM +0200, Heiko Stübner wrote:
> > Am Donnerstag, 5. Mai 2022, 08:41:31 CEST schrieb Sascha Hauer:
> > > On Thu, May 05, 2022 at 02:28:24AM +0200, Heiko Stübner wrote:
> > > > Am Freitag, 22. April 2022, 09:28:33 CEST schrieb Sascha Hauer:
> > > > > The VOP2 is the display output controller on the RK3568. Add the node
> > > > > for it to the dtsi file along with the required display-subsystem node
> > > > > and the iommu node.
> > > > > 
> > > > > Signed-off-by: Sascha Hauer <s.hauer at pengutronix.de>
> > > > > Acked-by: Rob Herring <robh at kernel.org>
> > > > > ---
> > > > > +	vop: vop at fe040000 {
> > > > > +		reg = <0x0 0xfe040000 0x0 0x3000>, <0x0 0xfe044000 0x0 0x1000>;
> > > > > +		reg-names = "regs", "gamma_lut";
> > > > 
> > > > reg-names does not seem to be part of the dt-binding.
> > > > Though now in the vop2-driver I see that the code uses that naming.
> > > > 
> > > > I guess we might want to just go the vop1-way by using numbered
> > > > platform-resources via a follow-up patch and drop the regnames here?
> > > 
> > > Or we could document the reg-names properties in a follow-up patch.
> > > Wouldn't that be better?
> > 
> > Also a possibility. Although that of course leads us into the naming
> > bike-shed ;-) ... i.e. for reg-names I'd do "vop" + "gamme-lut" (dash instead
> > of underscore) - instead of the ones used right now.
> 
> I also like dashes better than underscores and "vop" describes it better
> than the generic "reg" name, so at least the two of us don't need to
> dive deeper into the naming bike-shed ;)

:-D ... one problem solved already

> > While using numbered resources would keep the commonality between
> > both vops
> 
> I like the named resources better, but I'll leave it up to you. You
> decide, I'll create fixup patches in either direction.

ok, then let's go with the named-resources. As we'll need an Ack from
DT-people, we should probably expedite this.

Ideally I guess, do the dt-binding fixup + resend the fixed dt-node patches
with it. From the original series those dt-nodes + the hdmi rate stuff
is the only thing missing and the rates can be a separate thing.

Heiko





More information about the linux-arm-kernel mailing list