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: <20180312154016.GA29395@redhat.com>
Date:   Mon, 12 Mar 2018 11:40:16 -0400
From:   Mike Snitzer <snitzer@...hat.com>
To:     hare@...e.de, "Martin K. Petersen" <martin.petersen@...cle.com>,
        Ewan Milne <emilne@...hat.com>, linux-scsi@...r.kernel.org
Cc:     kbuild-all@...org, linux-kernel@...r.kernel.org,
        kbuild test robot <fengguang.wu@...el.com>,
        dm-devel@...hat.com
Subject: Re: ERROR: "scsi_device_from_queue" [drivers/md/dm-multipath.ko]
 undefined!

On Sat, Mar 10 2018 at  2:29pm -0500,
kbuild test robot <fengguang.wu@...el.com> wrote:

> tree:   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
> head:   3266b5bd97eaa72793df0b6e5a106c69ccc166c4
> commit: 8d47e65948ddea4398892946d9e50778a316b397 dm mpath: remove unnecessary NVMe branching in favor of scsi_dh checks
> date:   4 days ago
> config: x86_64-randconfig-x004-201810 (attached as .config)
> compiler: gcc-7 (Debian 7.3.0-1) 7.3.0
> reproduce:
>         git checkout 8d47e65948ddea4398892946d9e50778a316b397
>         # save the attached .config to linux build tree
>         make ARCH=x86_64 
> 
> All errors (new ones prefixed by >>):
> 
> >> ERROR: "scsi_device_from_queue" [drivers/md/dm-multipath.ko] undefined!
> 
> ---
> 0-DAY kernel test infrastructure                Open Source Technology Center
> https://lists.01.org/pipermail/kbuild-all                   Intel Corporation

This error is due to CONFIG_SCSI not being set.

DM commit 8d47e6594 ("dm mpath: remove unnecessary NVMe branching in
favor of scsi_dh checks") updated dm-mpath.c to call
scsi_device_from_queue() to determine if a device is SCSI.

The following patch fixes the buildbot reported case when CONFIG_SCSI
isn't set... but it is unclear if this is the best way forward.  I don't
want to make CONFIG_DM_MULTIPATH depend on CONFIG_SCSI, as such we
either need the following patch or the equivalent in a dm-mpath.c
wrapper.

Thoughts?

diff --git a/include/scsi/scsi_device.h b/include/scsi/scsi_device.h
index 7ae177c..d6bcc96 100644
--- a/include/scsi/scsi_device.h
+++ b/include/scsi/scsi_device.h
@@ -333,7 +333,14 @@ extern int scsi_add_device(struct Scsi_Host *host, uint channel,
 extern int scsi_unregister_device_handler(struct scsi_device_handler *scsi_dh);
 void scsi_attach_vpd(struct scsi_device *sdev);
 
+#ifdef CONFIG_SCSI
 extern struct scsi_device *scsi_device_from_queue(struct request_queue *q);
+#else
+static inline struct scsi_device *scsi_device_from_queue(struct request_queue *q)
+{
+	return NULL;
+}
+#endif
 extern int __must_check scsi_device_get(struct scsi_device *);
 extern void scsi_device_put(struct scsi_device *);
 extern struct scsi_device *scsi_device_lookup(struct Scsi_Host *,



Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ