linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Eric Biggers <ebiggers3@gmail.com>
To: Kees Cook <keescook@chromium.org>
Cc: kernel-hardening@lists.openwall.com,
	David Windsor <dave@nullcore.net>,
	linux-mm@kvack.org, linux-kernel@vger.kernel.org
Subject: Re: [kernel-hardening] [PATCH 22/23] usercopy: split user-controlled slabs to separate caches
Date: Mon, 19 Jun 2017 21:24:42 -0700	[thread overview]
Message-ID: <20170620042442.GC610@zzz.localdomain> (raw)
In-Reply-To: <1497915397-93805-23-git-send-email-keescook@chromium.org>

On Mon, Jun 19, 2017 at 04:36:36PM -0700, Kees Cook wrote:
> From: David Windsor <dave@nullcore.net>
> 
> Some userspace APIs (e.g. ipc, seq_file) provide precise control over
> the size of kernel kmallocs, which provides a trivial way to perform
> heap overflow attacks where the attacker must control neighboring
> allocations of a specific size. Instead, move these APIs into their own
> cache so they cannot interfere with standard kmallocs. This is enabled
> with CONFIG_HARDENED_USERCOPY_SPLIT_KMALLOC.
> 

This is a logically separate change which IMO should be its own patch, not just
patch 22/23.

Also, is this really just about heap overflows?  I thought the main purpose of
separate heaps is to make it more difficult to exploit use-after-frees, since
anything allocating an object from heap A cannot overwrite freed memory in heap
B.  (At least, not at the SLAB level; it may still be done at the page level.)

> diff --git a/include/linux/gfp.h b/include/linux/gfp.h
> index a89d37e8b387..ff4f4a698ad0 100644
> --- a/include/linux/gfp.h
> +++ b/include/linux/gfp.h
> @@ -45,6 +45,7 @@ struct vm_area_struct;
>  #else
>  #define ___GFP_NOLOCKDEP	0
>  #endif
> +#define ___GFP_USERCOPY		0x4000000u
>  /* If the above are modified, __GFP_BITS_SHIFT may need updating */
>  
>  /*
> @@ -83,12 +84,17 @@ struct vm_area_struct;
>   *   node with no fallbacks or placement policy enforcements.
>   *
>   * __GFP_ACCOUNT causes the allocation to be accounted to kmemcg.
> + *
> + * __GFP_USERCOPY indicates that the page will be explicitly copied to/from
> + *   userspace, and may be allocated from a separate kmalloc pool.
> + *
>   */

The "page", or the allocation?  It's only for slab objects, is it not?  More
importantly, the purpose of this needs to be clearly documented; otherwise
people won't know what this is and whether they should/need to use it or not.

- Eric

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

  reply	other threads:[~2017-06-20  4:24 UTC|newest]

Thread overview: 45+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-06-19 23:36 [PATCH 00/23] Hardened usercopy whitelisting Kees Cook
2017-06-19 23:36 ` [PATCH 01/23] usercopy: Prepare for " Kees Cook
2017-06-19 23:36 ` [PATCH 02/23] usercopy: Enforce slab cache usercopy region boundaries Kees Cook
2017-06-19 23:36 ` [PATCH 03/23] vfs: define usercopy region in names_cache slab caches Kees Cook
2017-06-19 23:36 ` [PATCH 04/23] vfs: copy struct mount.mnt_id to userspace using put_user() Kees Cook
2017-06-19 23:36 ` [PATCH 05/23] befs: define usercopy region in befs_inode_cache slab cache Kees Cook
2017-06-19 23:36 ` [PATCH 06/23] cifs: define usercopy region in cifs_request " Kees Cook
2017-06-19 23:36 ` [PATCH 07/23] exofs: define usercopy region in exofs_inode_cache " Kees Cook
2017-06-19 23:36 ` [PATCH 08/23] ext2: define usercopy region in ext2_inode_cache " Kees Cook
2017-06-19 23:36 ` [PATCH 09/23] ext4: define usercopy region in ext4_inode_cache " Kees Cook
2017-06-19 23:36 ` [PATCH 10/23] vxfs: define usercopy region in vxfs_inode " Kees Cook
2017-06-19 23:36 ` [PATCH 11/23] jfs: define usercopy region in jfs_ip " Kees Cook
2017-06-19 23:36 ` [PATCH 12/23] orangefs: define usercopy region in orangefs_inode_cache " Kees Cook
2017-06-19 23:36 ` [PATCH 13/23] ufs: define usercopy region in ufs_inode_cache " Kees Cook
2017-06-19 23:36 ` [PATCH 14/23] fork: define usercopy region in thread_stack, task_struct, mm_struct slab caches Kees Cook
2017-06-19 23:36 ` [PATCH 15/23] net: define usercopy region in struct proto slab cache Kees Cook
2017-06-19 23:36 ` [PATCH 16/23] net: copy struct sctp_sock.autoclose to userspace using put_user() Kees Cook
2017-06-19 23:36 ` [PATCH 17/23] dcache: define usercopy region in dentry_cache slab cache Kees Cook
2017-06-20  4:08   ` [kernel-hardening] " Eric Biggers
2017-06-28 16:44     ` Kees Cook
2017-06-28 16:55       ` Eric Biggers
2017-06-19 23:36 ` [PATCH 18/23] scsi: define usercopy region in scsi_sense_cache " Kees Cook
2017-06-19 23:36 ` [PATCH 19/23] xfs: define usercopy region in xfs_inode " Kees Cook
2017-06-19 23:36 ` [PATCH 20/23] usercopy: convert kmalloc caches to usercopy caches Kees Cook
2017-06-19 23:36 ` [PATCH 21/23] usercopy: Restrict non-usercopy caches to size 0 Kees Cook
2017-06-20  4:04   ` [kernel-hardening] " Eric Biggers
2017-06-28 17:03     ` Kees Cook
2017-06-19 23:36 ` [PATCH 22/23] usercopy: split user-controlled slabs to separate caches Kees Cook
2017-06-20  4:24   ` Eric Biggers [this message]
2017-06-20  4:47   ` [kernel-hardening] " Eric Biggers
2017-06-20 22:27     ` Kees Cook
2017-06-20 20:24   ` Laura Abbott
2017-06-20 22:22     ` Kees Cook
2017-06-27  7:31       ` Michal Hocko
2017-06-27 22:07         ` Kees Cook
2017-06-28  8:54           ` Michal Hocko
2017-06-19 23:36 ` [PATCH 23/23] mm: Allow slab_nomerge to be set at build time Kees Cook
2017-06-20  4:09   ` [kernel-hardening] " Daniel Micay
2017-06-20 22:51     ` Kees Cook
2017-06-20  4:29   ` Eric Biggers
2017-06-20 23:09     ` Kees Cook
2017-06-20 19:41 ` [kernel-hardening] [PATCH 00/23] Hardened usercopy whitelisting Rik van Riel
2017-10-20 22:40 ` Paolo Bonzini
2017-10-20 23:25   ` Paolo Bonzini
2017-10-21  3:04     ` Kees Cook

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20170620042442.GC610@zzz.localdomain \
    --to=ebiggers3@gmail.com \
    --cc=dave@nullcore.net \
    --cc=keescook@chromium.org \
    --cc=kernel-hardening@lists.openwall.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).