linux-kernel - Re: [PATCH AUTOSEL 5.5 094/542] s390/pci: Fix possible deadlock in recover

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: <20200217093156.GB42010@tuxmaker.boeblingen.de.ibm.com>
Date:   Mon, 17 Feb 2020 10:31:56 +0100
From:   Niklas Schnelle <schnelle@...ux.ibm.com>
To:     Sasha Levin <sashal@...nel.org>
Cc:     linux-kernel@...r.kernel.org, stable@...r.kernel.org,
        Peter Oberparleiter <oberpar@...ux.ibm.com>,
        Vasily Gorbik <gor@...ux.ibm.com>, linux-s390@...r.kernel.org
Subject: Re: [PATCH AUTOSEL 5.5 094/542] s390/pci: Fix possible deadlock in
 recover_store()

On Fri, Feb 14, 2020 at 10:41:26AM -0500, Sasha Levin wrote:
> From: Niklas Schnelle <schnelle@...ux.ibm.com>
> 
> [ Upstream commit 576c75e36c689bec6a940e807bae27291ab0c0de ]
> 
> With zpci_disable() working, lockdep detected a potential deadlock
> (lockdep output at the end).
> 
> The deadlock is between recovering a PCI function via the
> 
> /sys/bus/pci/devices/<dev>/recover
> 
> attribute vs powering it off via
> 
> /sys/bus/pci/slots/<slot>/power.
> 
> The fix is analogous to the changes in commit 0ee223b2e1f6 ("scsi: core:
> Avoid that SCSI device removal through sysfs triggers a deadlock")
> that fixed a potential deadlock on removing a SCSI device via sysfs.
[ ... snip ... ]

While technically useful on its own this commit really should go together with
the following upstream commit:

17cdec960cf776b20b1fb08c622221babe591d51
("s390/pci: Recover handle in clp_set_pci_fn()")

While the problem fixed here is independent,  writing to the power/recover
attributes will often fail due to an inconsistent function handle without the
second commit.
In particular without it a PCI function in the error state can not be
recovered or powered off.

I would recommend adding the second commit to the backports as well.

Thanks,
Niklas Schnelle
-- 
Niklas Schnelle
Linux on Z Development