linux-kernel - Re: [PATCH] PCI: Always lift 2.5GT/s restriction in PCIe failed link retraining

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <alpine.DEB.2.21.2512080029210.49654@angie.orcam.me.uk>
Date: Mon, 8 Dec 2025 19:25:26 +0000 (GMT)
From: "Maciej W. Rozycki" <macro@...am.me.uk>
To: Matthew W Carlis <mattc@...estorage.com>
cc: ahuang12@...ovo.com, alok.a.tiwari@...cle.com, ashishk@...estorage.com, 
    Bjorn Helgaas <bhelgaas@...gle.com>, guojinhui.liam@...edance.com, 
    Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>, 
    jiwei.sun.bj@...com, linux-kernel@...r.kernel.org, 
    linux-pci@...r.kernel.org, Lukas Wunner <lukas@...ner.de>, 
    msaggi@...estorage.com, sconnor@...estorage.com, sunjw10@...ovo.com
Subject: Re: [PATCH] PCI: Always lift 2.5GT/s restriction in PCIe failed link
 retraining

On Thu, 4 Dec 2025, Matthew W Carlis wrote:

> For what its worth I have a particular Gen5 device in my lab here that gets stuck
> in an infinite link up-down loop when you force the speed to Gen2 when also combined
> with a specific downstream port... I'm sure there is another device somewhere else
> in the world that has the same issue when you force it to Gen1.

 Well, it's an erratum vs another erratum case.  Then should such a device 
at its maximum link speed trigger the workaround somehow, with this fix in 
place any temporary clamp will be lifted anyway and the link retrained, so 
it will recover from the link up-down loop.  So no functional regression.

> The kernel should assume a PCIe link will automatically train to the best
> speed that the devices can achieve & if the link fails to train then it should
> be up to the user space to decide what to do with it in my opinion.

 It might be tough where you have your rootfs device down the problematic 
link.

 Thank you for your input.

  Maciej