Re: [PATCH 04/11] PCI: tegra194: Do not skip no-link Root Port remove cleanup

[Manikanta Maddireddy]

On 14/10/26 7:51 pm, Manivannan Sadhasivam wrote:
> On 2026-09-23 12:52 +0530, Manikanta Maddireddy wrote:
> > tegra_pcie_config_rp() deinitializes the controller and disables runtime PM
> > when link training fails, then probe treats -ENOMEDIUM as success so the
> > driver can remain bound without a link.
>
> What is the use of keeping the driver then? If the controller is deinitialized,
> the it can't do hotplug either.
>
> This patch is fine though.
>
> - Mani


This behaviour dates back to the original Tegra194 driver review. Bjorn 
recommended decoupling host-bridge probe success from downstream link 
state, so Root Ports remain accessible for enumeration/debugging and 
possible hotplug. However, as you point out, deinitializing the 
controller defeats those benefits. The existing no-link policy deserves 
a separate review; this patch only fixes the cleanup skipped when 
removing that already-bound driver.

Ref: https://lkml.indiana.edu/1907.2/06835.html

- Manikanta

--
nvpublic