netdev - Re: [PATCH 15/15] x86: use non-set

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: <20200507051213.GB4501@lst.de>
Date:   Thu, 7 May 2020 07:12:13 +0200
From:   Christoph Hellwig <hch@....de>
To:     Linus Torvalds <torvalds@...ux-foundation.org>
Cc:     Christoph Hellwig <hch@....de>,
        the arch/x86 maintainers <x86@...nel.org>,
        Alexei Starovoitov <ast@...nel.org>,
        Daniel Borkmann <daniel@...earbox.net>,
        Masami Hiramatsu <mhiramat@...nel.org>,
        Andrew Morton <akpm@...ux-foundation.org>,
        linux-parisc@...r.kernel.org,
        linux-um <linux-um@...ts.infradead.org>,
        Netdev <netdev@...r.kernel.org>, bpf@...r.kernel.org,
        Linux-MM <linux-mm@...ck.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 15/15] x86: use non-set_fs based maccess routines

On Wed, May 06, 2020 at 12:01:32PM -0700, Linus Torvalds wrote:
> Oh, absolutely. I did *NOT* mean that you'd use "unsafe_get_user()" as
> the actual interface. I just meant that as an implementation detail on
> x86, using "unsafe_get_user()" instead of "__get_user_size()"
> internally both simplifies the implementation, and means that it
> doesn't clash horribly with my local changes.

I had a version that just wrapped them, but somehow wasn't able to
make it work due to all the side effects vs macros issues.  Maybe I
need to try again, the current version seemed like a nice way out
as it avoided a lot of the silly casting.


> Btw, that brings up another issue: so that people can't mis-use those
> kernel accessors and use them for user addresses, they probably should
> actually do something like
> 
>         if ((long)addr >= 0)
>                 goto error_label;
> 
> on x86. IOW, have the "strict" kernel pointer behavior.
> 
> Otherwise somebody will start using them for user pointers, and it
> will happen to work on old x86 without CLAC/STAC support.
> 
> Of course, maybe CLAC/STAC is so common these days (at least with
> developers) that we don't have to worry about it.

The actual public routines (probe_kernel_read and co) get these
checks through probe_kernel_read_allowed, which is implemented by
the x86 code.  Doing this for every 1-8 byte access might be a little
slow, though.  Do you really fear drivers starting to use the low-level
helper?  Maybe we need to move those into a different header than
<asm/uaccess.h> that makes it more clear that they are internal?

> But here you see what it is, if you want to. __get_user_size()
> technically still exists, but it has the "target branch" semantics in
> here, so your patch clashes badly with it.

The target branch semantics actually are what I want, that is how the
maccess code is structured.  This is the diff I'd need for the calling
conventions in your bundle:


diff --git a/arch/x86/include/asm/uaccess.h b/arch/x86/include/asm/uaccess.h
index 765e18417b3ba..d1c8aacedade1 100644
--- a/arch/x86/include/asm/uaccess.h
+++ b/arch/x86/include/asm/uaccess.h
@@ -526,14 +526,8 @@ do {									\
 #define HAVE_ARCH_PROBE_KERNEL
 
 #define arch_kernel_read(dst, src, type, err_label)			\
-do {									\
-        int __kr_err;							\
-									\
 	__get_user_size(*((type *)dst), (__force type __user *)src,	\
-			sizeof(type), __kr_err);			\
-        if (unlikely(__kr_err))						\
-		goto err_label;						\
-} while (0)
+			sizeof(type), err_label);			\
 
 #define arch_kernel_write(dst, src, type, err_label)			\
 	__put_user_size(*((type *)(src)), (__force type __user *)(dst),	\