Re: [PATCH v2 1/2] PCI: dwc: Add .post_deinit counterpart to endpoint .pre_init callback

[Marek Vasut]

Hello Den-san,

On 8/20/26 10:09 AM, Koichiro Den wrote:
> On Thu, Aug 20, 2026 at 05:57:55AM +0200, Marek Vasut wrote:
> > In case the .pre_init() callback fails, it is mandatory to correctly
>
> I think this should be "If .pre_init() succeeds but subsequent initialization
> fails,". Otherwise LGTM.

This is fixed in V3, thank you.

> I did some quick tests on an S4 Spider, covering the normal path and several
> injected failure paths. I don't have V4H hardware, though.
> Please feel free to pick up either or both tags if that helps.
>
> Reviewed-by: Koichiro Den <den@xxxxxxxxxxxxx>
> Tested-by: Koichiro Den <den@xxxxxxxxxxxxx>

I am only following the code execution paths in the endpoint case .

My endpoint setup for V4H is still pending assembly, I finally have the 
oculink adapters and cable here, but I did not assemble it yet. I still 
need to find out how to handle clocking, whether one V4H PCIe controller 
(clock source) has to send the PCIe clock to the other V4H PCIe 
controller (clock sink) which would require board modification, or 
whether I can manage the oculink RC-to-EP setup without board modification.

Thank you for your help with testing on the S4 !

--
Best regards,
Marek Vasut