Re: [PATCH RFC 0/2] PCI: rockchip-host: Retry training and hot-reset a JMB585 on RK3399

[Igor Velkov]

On 2026/10/27, Shawn Lin wrote:
> I have no idea about it. The PCIE_CLIENT_BASIC_STATUS0 says:
>
> 14:12  	RO  max_payload_size
>
> So you are saying different MPS in different cases, not linking status.

You're right, thanks. Our two reads came from different points: the
slow-link one before host_init_port() sets the MPS to 256 bytes, the
other after it, so bit 12 says nothing about the link. The link-up
time after the hot reset remains the only signal we have: 2.9-4.5 ms
fine, 10.9-35.4 ms aborts.