Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 17 Oct 2017 17:44:33 -0400 (EDT)
From: Vladis Dronov <>
Subject: Re: CVE-2017-12190: Linux kernel: block: memory leak
 when merging small consecutive buffers in SCSI IO vectors

Hello, oss-security@,

If somebody could help me a bit with this CVE? Could you please run a reproducer
from (it does only a number of disk reads and allocates a memory):

and check if the issue reproduces on your system? The system should have a scsi
disk, so for a virtual machine a virtio disk (/dev/vda) won't work, but a virtual
scsi disk (/dev/sda) should.

Please, check a value of 'Active(anon)' in /proc/meminfo before and after running
a reproducer, it should jump up for 40Mb and should not decrease, like this:

(as root)# grep anon /proc/meminfo ; ./sgio-leak /dev/sda ; grep anon /proc/meminfo
Active(anon):      39156 kB
Inactive(anon):     8368 kB
Active(anon):      79268 kB
Inactive(anon):     8368 kB
(as root)# grep anon /proc/meminfo ; ./sgio-leak /dev/sda ; grep anon /proc/meminfo
Active(anon):      79268 kB
Inactive(anon):     8368 kB
Active(anon):     119156 kB
Inactive(anon):     8368 kB

If the issue in not reproduced, 'Active(anon)' should jump up shortly but return to
the previous before-the-run value on an idle system.

Best regards,
Vladis Dronov | Red Hat, Inc. | Product Security Engineer

Powered by blists - more mailing lists

Please check out the Open Source Software Security Wiki, which is counterpart to this mailing list.

Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.