Re: Linux kernel ELF core dump privilege elevation

Re: Linux kernel ELF core dump privilege elevation

[Greg KH]

On Wed, May 11, 2005 at 01:08:56PM +0200, Paul Starzetz wrote:
> Hi,
> 
> since it became clear from the discussion in January about the uselib() 
> vulnerability, that the Linux community prefers full, non-embargoed 
> disclosure of kernel bugs, I release full details right now. However to 
> follows at least some of the responsable disclosure rules, no exploit code will be 
> released. Instead, only a proof-of-concept code is released to demonstrate 
> the vulnerability.

<snip>

And here's a patch for 2.6 that is completly untested.  I'll work on
testing it today and if it works, we will release a new 2.6.11.y release
with this fix in it.

thanks,

greg k-h


Subject: possibly fix Linux kernel ELF core dump privilege elevation

As noted by Paul Starzetz

references CAN-something-I-need-to-go-look-up...

Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

---
 fs/binfmt_elf.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

--- gregkh-2.6.orig/fs/binfmt_elf.c	2005-05-11 00:03:45.000000000 -0700
+++ gregkh-2.6/fs/binfmt_elf.c	2005-05-11 00:09:17.000000000 -0700
@@ -251,7 +251,7 @@
 	}
 
 	/* Populate argv and envp */
-	p = current->mm->arg_start;
+	p = current->mm->arg_end = current->mm->arg_start;
 	while (argc-- > 0) {
 		size_t len;
 		__put_user((elf_addr_t)p, argv++);
@@ -1301,7 +1301,7 @@
 static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p,
 		       struct mm_struct *mm)
 {
-	int i, len;
+	unsigned int i, len;
 	
 	/* first copy the parameters from user space */
 	memset(psinfo, 0, sizeof(struct elf_prpsinfo));

Re: Linux kernel ELF core dump privilege elevation

[Greg KH]

On Wed, May 11, 2005 at 11:12:11AM -0700, Greg KH wrote:
> On Wed, May 11, 2005 at 01:08:56PM +0200, Paul Starzetz wrote:
> > Hi,
> > 
> > since it became clear from the discussion in January about the uselib() 
> > vulnerability, that the Linux community prefers full, non-embargoed 
> > disclosure of kernel bugs, I release full details right now. However to 
> > follows at least some of the responsable disclosure rules, no exploit code will be 
> > released. Instead, only a proof-of-concept code is released to demonstrate 
> > the vulnerability.
> 
> <snip>
> 
> And here's a patch for 2.6 that is completly untested.  I'll work on
> testing it today and if it works, we will release a new 2.6.11.y release
> with this fix in it.
> 
> thanks,
> 
> greg k-h
> 
> 
> Subject: possibly fix Linux kernel ELF core dump privilege elevation
> 
> As noted by Paul Starzetz
> 
> references CAN-something-I-need-to-go-look-up...

CAN-2005-1263  is the correct one.  Sorry for being lazy and not looking
it up right away.

thanks,

greg k-h

Re: Linux kernel ELF core dump privilege elevation

[Paul Starzetz]

On Wed, 11 May 2005, Greg KH wrote:

that seems ok.

> --- gregkh-2.6.orig/fs/binfmt_elf.c	2005-05-11 00:03:45.000000000 -0700
> +++ gregkh-2.6/fs/binfmt_elf.c	2005-05-11 00:09:17.000000000 -0700
> @@ -251,7 +251,7 @@
>  	}
>  
>  	/* Populate argv and envp */
> -	p = current->mm->arg_start;
> +	p = current->mm->arg_end = current->mm->arg_start;
>  	while (argc-- > 0) {
>  		size_t len;
>  		__put_user((elf_addr_t)p, argv++);
> @@ -1301,7 +1301,7 @@
>  static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p,
>  		       struct mm_struct *mm)
>  {
> -	int i, len;
> +	unsigned int i, len;
>  	
>  	/* first copy the parameters from user space */
>  	memset(psinfo, 0, sizeof(struct elf_prpsinfo));
> 

-- 
Paul Starzetz
iSEC Security Research
http://isec.pl/