[RFC PATCH v2 1/2] PCI: rockchip-host: Retry link training on failure without PERST#

[Geraldo Nascimento]

On Tue, Jun 10, 2025 at 01:44:49PM -0500, Bjorn Helgaas wrote:
> This *looks* like it could be strictly a move of
> rockchip_pcie_set_vpcie(), without changing it at all.

Hi Bjorn,

yes, a move of rockchip_pcie_set_vpcie() is needed for re-enabling
power rails, considering patch 2/2 was also applied and the regulator
is not always-on anymore.

> If that's the case, please make the move a separate patch so it's more
> obvious what the interesting changes that actually make a difference
> are.

Good idea, I'll send v3.

Thank you!
Geraldo Nascimento