linux-kernel - Re: [PATCH 1/2 v2] kprobe: Do not use uaccess functions to access kernel memory that can fault

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: <CAHk-=wh9XrOykA5J9RQ7zaBio-S_D+1AE+rGnBYWSd==pCXh+w@mail.gmail.com>
Date:   Mon, 18 Feb 2019 10:23:44 -0800
From:   Linus Torvalds <torvalds@...ux-foundation.org>
To:     Steven Rostedt <rostedt@...dmis.org>
Cc:     Andy Lutomirski <luto@...capital.net>,
        Linux List Kernel Mailing <linux-kernel@...r.kernel.org>,
        Ingo Molnar <mingo@...nel.org>,
        Andrew Morton <akpm@...ux-foundation.org>,
        stable <stable@...r.kernel.org>,
        Changbin Du <changbin.du@...il.com>,
        Jann Horn <jannh@...gle.com>,
        Kees Cook <keescook@...omium.org>,
        Andy Lutomirski <luto@...nel.org>
Subject: Re: [PATCH 1/2 v2] kprobe: Do not use uaccess functions to access
 kernel memory that can fault

On Mon, Feb 18, 2019 at 9:58 AM Linus Torvalds
<torvalds@...ux-foundation.org> wrote:
>
> So I've pulled the existing pull request, because it papers over one
> particular issue, but the real fix would require:
>
>  - knowing whether it's kernel or user space you access

Actually, it would be good to split the "kernel space" access into
finer granularities, because it would be good to limit those addresses
by what you expect to happen.

For example, on x86-64, we have a separate virtual mapping for kernel
text. If you're following a data pointer, you probably shouldn't even
look at that area, and just disallow it up-front.

And the vmalloc space, we should probably look up the vmalloc
descriptor for the address, to make sure we don't go into ioremap'ed
areas (but we'd generally *do* want to follow pages for module data,
or for vmap'ed stacks).

And fixmap and/or percpu pointers may or may not be something that
we'd want to follow.

Etc.

So it would be good to not just say "user or kernel", but actually say
what *kind* of kernel access it expects.

               Linus