xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Ian Jackson <ian.jackson@eu.citrix.com>
Cc: "xen-devel@lists.xensource.com" <xen-devel@lists.xensource.com>,
	"mattjd@gmail.com" <mattjd@gmail.com>,
	"security@xen.org" <security@xen.org>
Subject: Re: [PATCH 22/22] libxc: check blob size before proceeding in xc_dom_check_gzip
Date: Mon, 10 Jun 2013 21:10:30 +0100	[thread overview]
Message-ID: <51B632B6.9040007@citrix.com> (raw)
In-Reply-To: <1370629642-6990-23-git-send-email-ian.jackson@eu.citrix.com>

On 07/06/13 19:27, Ian Jackson wrote:
> From: Matthew Daley <mattjd@gmail.com>
>
> This is part of the fix to a security issue, XSA-55.
>
> Signed-off-by: Matthew Daley <mattjd@gmail.com>
>
> v6: This patch is new in v6 of the series.
> ---
>  tools/libxc/xc_dom_core.c |    4 ++++
>  1 files changed, 4 insertions(+), 0 deletions(-)
>
> diff --git a/tools/libxc/xc_dom_core.c b/tools/libxc/xc_dom_core.c
> index 2a9c5a2..525b364 100644
> --- a/tools/libxc/xc_dom_core.c
> +++ b/tools/libxc/xc_dom_core.c
> @@ -278,6 +278,10 @@ size_t xc_dom_check_gzip(xc_interface *xch, void *blob, size_t ziplen)
>      unsigned char *gzlen;
>      size_t unziplen;
>  
> +    if ( ziplen < 6 )
> +        /* too small */
> +        return 0;
> +
>      if ( strncmp(blob, "\037\213", 2) )
>          /* not gzipped */
>          return 0;

The above change is certainly good and correct.

Reviewed-by: Andrew Cooper <andrew.cooper3@citrix.com>


I do however have concerns about the rest of the function, immediately
after the context.  I spot any other specific security related issues,
but would appreciate comments/thoughts as to whether these should be
part of this patch or part of a subsequent set of bugfix patches.

>     gzlen = blob + ziplen - 4;

Arithmatic on a void pointer (blob), and a possibility to overflow,
although this seems unlikely given that it would have to pass
xc_dom_malloc_filemap() in the first place.  There is a GNU extention
which allows this line to Do The Right Thing, despite violating the C spec.

>     unziplen = gzlen[3] << 24 | gzlen[2] << 16 | gzlen[1] << 8 | gzlen[0];
>     if ( (unziplen < 0) || (unziplen > XC_DOM_DECOMPRESS_MAX) )

unziplen is unsigned, so (unziplen < 0) is necessarily false.

>     {
>          xc_dom_printf
>             (xch,
>              "%s: size (zip %zd, unzip %zd) looks insane, skip gunzip",
>              __FUNCTION__, ziplen, unziplen);

ziplen and unziplen are both unsigned, so the format string should be
%zu instead.

>     return 0;
>     }
>
>     return unziplen + 16;

While XC_DOM_DECOMPRESS_MAX is set to a default of 1GB, if it is
changes, unziplen + 16 can overflow.

> }

~Andrew

  reply	other threads:[~2013-06-10 20:10 UTC|newest]

Thread overview: 68+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-06-07 18:27 [PATCH v6 00/22] XSA55 libelf fixes for unstable Ian Jackson
2013-06-07 18:27 ` [PATCH 01/22] libelf: abolish libelf-relocate.c Ian Jackson
2013-06-07 18:27 ` [PATCH 02/22] libxc: introduce xc_dom_seg_to_ptr_pages Ian Jackson
2013-06-10 12:21   ` Andrew Cooper
2013-06-10 13:40     ` Ian Jackson
2013-06-10 13:49       ` Andrew Cooper
2013-06-10 14:02         ` Ian Jackson
2013-06-10 15:24           ` Andrew Cooper
2013-06-07 18:27 ` [PATCH 03/22] libxc: Fix range checking in xc_dom_pfn_to_ptr etc Ian Jackson
2013-06-10 15:46   ` Andrew Cooper
2013-06-10 15:48     ` Ian Jackson
2013-06-10 15:53       ` Andrew Cooper
2013-06-10 15:54         ` Ian Jackson
2013-06-07 18:27 ` [PATCH 04/22] libelf: add `struct elf_binary*' parameter to elf_load_image Ian Jackson
2013-06-07 18:27 ` [PATCH 05/22] libelf: abolish elf_sval and elf_access_signed Ian Jackson
2013-06-07 18:27 ` [PATCH 06/22] libelf: move include of <asm/guest_access.h> to top of file Ian Jackson
2013-06-07 18:27 ` [PATCH 07/22] libelf/xc_dom_load_elf_symtab: Do not use "syms" uninitialised Ian Jackson
2013-06-07 18:27 ` [PATCH 08/22] libelf: introduce macros for memory access and pointer handling Ian Jackson
2013-06-10 16:57   ` Andrew Cooper
2013-06-10 16:58     ` Ian Jackson
2013-06-07 18:27 ` [PATCH 09/22] tools/xcutils/readnotes: adjust print_l1_mfn_valid_note Ian Jackson
2013-06-07 18:27 ` [PATCH 10/22] libelf: check nul-terminated strings properly Ian Jackson
2013-06-10 23:17   ` Andrew Cooper
2013-06-11 15:17     ` Ian Jackson
2013-06-07 18:27 ` [PATCH 11/22] libelf: check all pointer accesses Ian Jackson
2013-06-10 23:38   ` Andrew Cooper
2013-06-11 15:28     ` Ian Jackson
2013-06-07 18:27 ` [PATCH 12/22] libelf: Check pointer references in elf_is_elfbinary Ian Jackson
2013-06-10 23:04   ` Andrew Cooper
2013-06-11 15:11     ` Ian Jackson
2013-06-11 16:00       ` Andrew Cooper
2013-06-11 16:33         ` Ian Jackson
2013-06-07 18:27 ` [PATCH 13/22] libelf: Make all callers call elf_check_broken Ian Jackson
2013-06-07 18:27 ` [PATCH 14/22] libelf: use C99 bool for booleans Ian Jackson
2013-06-10 22:09   ` Andrew Cooper
2013-06-11  7:17     ` Jan Beulich
2013-06-11 14:50     ` Ian Jackson
2013-06-07 18:27 ` [PATCH 15/22] libelf: use only unsigned integers Ian Jackson
2013-06-10 18:28   ` Andrew Cooper
2013-06-11 13:16     ` Ian Jackson
2013-06-07 18:27 ` [PATCH 16/22] libelf: check loops for running away Ian Jackson
2013-06-10 22:56   ` Andrew Cooper
2013-06-11 15:07     ` Ian Jackson
2013-06-07 18:27 ` [PATCH 17/22] libelf: abolish obsolete macros Ian Jackson
2013-06-07 18:27 ` [PATCH 18/22] libxc: Add range checking to xc_dom_binloader Ian Jackson
2013-06-10 21:37   ` Andrew Cooper
2013-06-11 14:46     ` Ian Jackson
2013-06-07 18:27 ` [PATCH 19/22] libxc: check failure of xc_dom_*_to_ptr, xc_map_foreign_range Ian Jackson
2013-06-10 21:09   ` Andrew Cooper
2013-06-11 14:44     ` Ian Jackson
2013-06-07 18:27 ` [PATCH 20/22] libxc: check return values from malloc Ian Jackson
2013-06-07 18:32   ` Andrew Cooper
2013-06-11 15:53     ` Ian Jackson
2013-06-10 20:38   ` Andrew Cooper
2013-06-11 14:31     ` Ian Jackson
2013-06-07 18:27 ` [PATCH 21/22] libxc: range checks in xc_dom_p2m_host and _guest Ian Jackson
2013-06-07 18:33   ` Andrew Cooper
2013-06-08 12:05     ` Andrew Cooper
2013-06-11 15:58       ` Ian Jackson
2013-06-11 16:19         ` Tim Deegan
2013-06-07 18:27 ` [PATCH 22/22] libxc: check blob size before proceeding in xc_dom_check_gzip Ian Jackson
2013-06-10 20:10   ` Andrew Cooper [this message]
2013-06-11 13:46     ` Ian Jackson
2013-06-10 23:44 ` [PATCH v6 00/22] XSA55 libelf fixes for unstable Andrew Cooper
  -- strict thread matches above, loose matches on Subject: below --
2013-06-07 18:35 [PATCH v6 00/22] XSA55 libelf fixes for Xen 4.2 Ian Jackson
2013-06-07 18:35 ` [PATCH 22/22] libxc: check blob size before proceeding in xc_dom_check_gzip Ian Jackson
2013-06-11 18:20 [PATCH v7 00/22] XSA55 libelf fixes for unstable Ian Jackson
2013-06-11 18:21 ` [PATCH 22/22] libxc: check blob size before proceeding in xc_dom_check_gzip Ian Jackson
2013-06-11 19:08   ` Andrew Cooper
2013-06-11 18:22 [PATCH v7 00/22] XSA55 libelf fixes for Xen 4.2 Ian Jackson
2013-06-11 18:22 ` [PATCH 22/22] libxc: check blob size before proceeding in xc_dom_check_gzip Ian Jackson

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=51B632B6.9040007@citrix.com \
    --to=andrew.cooper3@citrix.com \
    --cc=ian.jackson@eu.citrix.com \
    --cc=mattjd@gmail.com \
    --cc=security@xen.org \
    --cc=xen-devel@lists.xensource.com \
    /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).