Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20190418170058.GA22985@openwall.com>
Date: Thu, 18 Apr 2019 19:00:58 +0200
From: Solar Designer <solar@...nwall.com>
To: oss-security@...ts.openwall.com
Subject: Linux kernel address leaks

Hi,

Fuqian just sent us all 13 of these in 2 days, and I guess there might
be many more to come.  Do we really want to see them in here?  And in
that many individual messages?  I doubt it - but not enough to have used
my moderator powers to outright reject the messages without discussion.

This in part depends on upstream Linux kernel's policy as to kernel
address leaks.  Are those treated as vulnerabilities or as not yet fully
completed kernel hardening effort?  I think it's the latter, in which
case such reports are better sent to the kernel-hardening mailing list.

Also, what's the current approach at dealing with them?  I haven't been
keeping track of that.  Is the kptr_restrict sysctl possibly supposed to
be sufficient, and %p format specifiers can stay intact?  If so, are
those many individual findings by Fuqian actually non-issues?

A proposed patch is generally better than a mere problem report, so if
the issues do need individual fixes, then:

I think these issues should be grouped by subsystem (one message per
subsystem) and sent in the form of proposed patches (fixing the issues
how exactly? need to discuss that first) to LKML and to proper
per-subsystem lists and addresses (per the MAINTAINERS file), CC'ing
kernel-hardening.

I'd appreciate it if someone currently involved in dealing with similar
issues comments on this in a reply-to-all to this message.

Thanks,

Alexander

On Tue, Apr 16, 2019 at 10:08:10AM +0800, Fuqian Huang wrote:
> In ascot2e_attach, dev_info will print the address of adapter to
> dmesg, sensitive kernel information will be leaked to user space.
> struct dvb_frontend *ascot2e_attach(struct dvb_frontend *fe,
>   const struct ascot2e_config *config,
>   struct i2c_adapter *i2c)
> {
>   ...
>   dev_info(&priv->i2c->dev,
>   "Sony ASCOT2E attached on addr=%x at I2C adapter %p\n",
>   priv->i2c_address, priv->i2c);
>   ...
> }

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.