linux-kernel - Re: [PATCH 1/2] PCI: Clear LBMS on resume to avoid Target Speed quirk

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: <42afa3ee-4429-90e4-9e98-18a0d30c0a3c@linux.intel.com>
Date: Fri, 9 Aug 2024 18:55:56 +0300 (EEST)
From: Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>
To: "Maciej W. Rozycki" <macro@...am.me.uk>
cc: Bjorn Helgaas <helgaas@...nel.org>, linux-pci@...r.kernel.org, 
    LKML <linux-kernel@...r.kernel.org>, 
    Mika Westerberg <mika.westerberg@...ux.intel.com>
Subject: Re: [PATCH 1/2] PCI: Clear LBMS on resume to avoid Target Speed
 quirk

On Fri, 9 Aug 2024, Maciej W. Rozycki wrote:

> On Wed, 7 Feb 2024, Ilpo Järvinen wrote:
> 
> > > > > Because if it is constantly picking another speed, it would mean you get 
> > > > > LBMS set over and over again, no? If that happens 34-35 times per second, 
> > > > > it should be set already again when we get into that quirk because there 
> > > > > was some wait before it gets called.
> > > > 
> > > >  I'll see if I can experiment with the hardware over the next couple of 
> > > > days and come back with my findings.
> > > 
> > > Okay thanks.
> > 
> > One point I'd be very interested to know if the link actually comes up 
> > successfully (even if briefly) because this has large implications whether 
> > the quirk can actually be invoked from the bandwidth controller code.
> 
>  That was more than a couple of days, sorry about it.  I have now been 
> able to verify that LBMS keeps getting reasserted over and over again as 
> the device goes through the infinite link training dance.  I haven't ever 
> observed the link to become active in the course.  Here's a short log of 
> commands repeatedly entered at the command prompt:
> 
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5011
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5812
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5811
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5812
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5011
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5811
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5811
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5812
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5812
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5811
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5811
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5812
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5812
> # setpci -s 02:03.0 CAP_EXP+0x12.W
> 5811
> 
> As you can see the Link Training bit oscillates as I previously reported 
> and noted in the introduction to `pcie_failed_link_retrain', and also the 
> Current Link Speed field flips between 2.5GT/s and 5GT/s.

Okay, thanks for testing. I suppose that test wasn't done in a busy loop 
(it might not be easy capture very short link up state if there was any 
such period when testing it by manually launching that command a few 
times)?


-- 
 i.