linux-kernel - Re: [PATCH 16/18] net: mpls: prevent bounds-check bypass via speculative execution

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]

Date:   Tue, 9 Jan 2018 19:07:17 -0800
From:   Alexei Starovoitov <alexei.starovoitov@...il.com>
To:     Dan Williams <dan.j.williams@...el.com>
Cc:     Linus Torvalds <torvalds@...ux-foundation.org>,
        "Eric W. Biederman" <ebiederm@...ssion.com>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        linux-arch@...r.kernel.org, Peter Zijlstra <peterz@...radead.org>,
        Netdev <netdev@...r.kernel.org>,
        Greg KH <gregkh@...uxfoundation.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        "David S. Miller" <davem@...emloft.net>,
        Elena Reshetova <elena.reshetova@...el.com>,
        Alan Cox <alan@...ux.intel.com>
Subject: Re: [PATCH 16/18] net: mpls: prevent bounds-check bypass via
 speculative execution

On Tue, Jan 09, 2018 at 06:22:09PM -0800, Dan Williams wrote:
> 
> When you came up with that tweak you noted:
> 
> "The following:
> [..]
> is generic and no speculative flows."

I meant 'no speculative control flow'
load speculation still happens.

> 
> > This macro doesn't prevent speculation.
> 
> It masks dangerous speculation. At least, I read nospec as "No
> Spectre" and it is a prefix used in the Spectre-v2 patches.

ahh. I thought 'nospec' means 'no speculation'.
I think it's too much of an honor to use bug name for the macro
that will be used in many places in the kernel.

> > I think array_access() was the best name so far.
> 
> For other usages I need the pointer to the array element, also
> array_access() by itself is unsuitable for __fcheck_files because we
> still need rcu_dereference_raw() on the element de-reference. So, I
> think it's better to get a sanitized array element pointer which can
> be used with rcu, READ_ONCE(), etc... directly rather than try to do
> the access in the same macro.

makes sense, then array_ptr() should fit ?
I'm hearing rumors that the first cpu with variant 2 and 3 fixed
will be appearing in early 2019. Which is amazing considering cpu
release cycles, but it also means that variant 1 will stay with us
for long time and we better pick clean interface and name for it.