On 10/29/26 11:09 AM, Nora Schiffer wrote:
On Wed, 2026-10-28 at 16:16 +0100, Marek Vasut wrote:
On 10/28/26 11:40 AM, Nora Schiffer wrote:
On Tue, 2026-10-27 at 18:24 +0100, Marek Vasut wrote:
On 10/27/26 10:39 AM, Nora Schiffer wrote:
While a separate usbmisc driver could be added, no such driver currently
exists, and it would not necessarily be part of the PHY subsystem. Rather
than making the misc setup conditional on CONFIG_PHY, make it depend on
the existence of a fsl,usbmisc property. This will allow enabling
CONFIG_PHY on the i.MX7D for the NOP PHY driver's power domain support.
Also update the code that reads the usbmisc address to use the modern
ofnode API and fix the portnr to be determined from the usbmisc phandle
arg instead of the alias seqno.
It seems this patch does multiple things, can you please split it up ?
Okay, I can split this into two parts:
- Make the setup independent of CONFIG_PHY
- Use address/portnr from ofnode
Yes please, the clearer the separation of changes is, the better
(=easier to review). Also, please split the series by subsystems, so the
maintainers can pick it up through their trees. Thank you.
I get that these will need to go through different trees eventually, but I think
in the current stage having the full series to show the whole picture and have
something testable makes sense.
Will the ehci-mx6 patches go through the USB or the i.MX tree? I feel like
pushing them through i.MX would be less of a pain due to the dependencies
between the patches.
USB, since this is USB stuff. There do not seem to be many dependencies
between patches in this series.
