linux-kernel - [BISECTED] nvme probe failure with v5.13-rc1

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-next>] [day] [month] [year] [list]

Date:   Fri, 21 May 2021 05:00:29 +0000
From:   "Verma, Vishal L" <vishal.l.verma@...el.com>
To:     "kbusch@...nel.org" <kbusch@...nel.org>,
        "linux-nvme@...ts.infradead.org" <linux-nvme@...ts.infradead.org>
CC:     "hch@....de" <hch@....de>,
        "Williams, Dan J" <dan.j.williams@...el.com>,
        "Schofield, Alison" <alison.schofield@...el.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "Weiny, Ira" <ira.weiny@...el.com>,
        "Widawsky, Ben" <ben.widawsky@...el.com>
Subject: [BISECTED] nvme probe failure with v5.13-rc1

Hi,

I ran into this failure to probe an nvme device in an emulator
(simics). It looks like there is a ~60 second wait followed by a
timeout and a failure to boot (the root device is an nvme disk) with
these messages in the log:

   [   67.174010] nvme nvme0: I/O 5 QID 0 timeout, disable controller
   [   67.175793] nvme nvme0: Removing after probe failure status: -4

I bisected this to:
   5befc7c26e5a ("nvme: implement non-mdts command limits") 

It's not immediately obvious to me what's causing the problem.
Reverting the above commit fixes it. It is easily reproducible - I'd be
happy to provide more info about the emulated device or test out
patches or theories.

It is of course possible that the emulated device is behaving in some
non spec-compliant way, in which case I'd appreciate any help figuring
out what that is.

Thanks,
-Vishal