linux-kernel - RE: [4.10][regression][PM] Oops pci_restore_msi_state() e1000e? (Re: [BUG] 4.11.0-rc1 panic on shutdown X61s)

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] [day] [month] [year] [list]

Message-ID: <309B89C4C689E141A5FF6A0C5FB2118B8C5B4FB3@ORSMSX101.amr.corp.intel.com>
Date:   Thu, 16 Mar 2017 06:52:43 +0000
From:   "Brown, Aaron F" <aaron.f.brown@...el.com>
To:     Shawn Starr <shawn.starr@...ers.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: RE: [4.10][regression][PM] Oops pci_restore_msi_state() e1000e?
 (Re: [BUG] 4.11.0-rc1 panic on shutdown X61s)

> From: Shawn Starr [mailto:shawn.starr@...ers.com]
> Sent: Tuesday, March 14, 2017 1:43 PM
> To: linux-kernel@...r.kernel.org
> Cc: Brown, Aaron F <aaron.f.brown@...el.com>
> Subject: Re: [4.10][regression][PM] Oops pci_restore_msi_state() e1000e?
> (Re: [BUG] 4.11.0-rc1 panic on shutdown X61s)
> 
> This isn't just Lenovo,
> 
> My Dell laptop experiences same problem, there is bugzilla:
> 
> https://bugzilla.kernel.org/show_bug.cgi?id=194801

Yes, I had noticed a Dell laptop in one of the reports in addition to the Thinkpads.  As it turns out one of the systems I have been trying to reproduce it on is a Lenova with an 82677LM, a newer chipset than the one Vito Caputo is seeing the problem on (82566) and a slightly older one than the one Boris Petkov reported (82579LM.)  I also have been trying with a SuperMicro that has an 82579LM that matches Boris's lspci output very closely and still have not managed to trigger it.  I used the config Boris provided to build a kernel and tried it on those systems (and a number of other ones) and still have not succeeded in making it break.  Given I've tried with the same config as one that's breaking and the range of chipsets it is appearing on I'm suspicious it's more to do with the system environment than the exact kernel config or e1000e chipset, but really don't know.  If I try to pursue it more I guess I'll try installing a Debian build on some of the systems.

Regardless, Sasha Neftin weighed in on it agreeing it needs to be reverted and the upstream people were included, so hopefully it will get reverted in one of the next dot releases.

-Aaron