diff options
author | Oleg Nesterov <oleg@redhat.com> | 2010-11-30 20:55:34 +0100 |
---|---|---|
committer | Willy Tarreau <w@1wt.eu> | 2011-04-30 16:53:36 +0200 |
commit | bb4659e0d68667f5eac7ef551a95762564e484a3 (patch) | |
tree | d30262be5da55e8c3a34a8415193abdfe73294ac /include | |
parent | 8ece1d67112e1a03cf7cad1c5a5771bdaa8966e8 (diff) |
exec: make argv/envp memory visible to oom-killer
commit 3c77f845722158206a7209c45ccddc264d19319c upstream.
Brad Spengler published a local memory-allocation DoS that
evades the OOM-killer (though not the virtual memory RLIMIT):
http://www.grsecurity.net/~spender/64bit_dos.c
execve()->copy_strings() can allocate a lot of memory, but
this is not visible to oom-killer, nobody can see the nascent
bprm->mm and take it into account.
With this patch get_arg_page() increments current's MM_ANONPAGES
counter every time we allocate the new page for argv/envp. When
do_execve() succeds or fails, we change this counter back.
Technically this is not 100% correct, we can't know if the new
page is swapped out and turn MM_ANONPAGES into MM_SWAPENTS, but
I don't think this really matters and everything becomes correct
once exec changes ->mm or fails.
Reported-by: Brad Spengler <spender@grsecurity.net>
Reviewed-and-discussed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Moritz Muehlenhoff <jmm@debian.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Diffstat (limited to 'include')
-rw-r--r-- | include/linux/binfmts.h | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/include/linux/binfmts.h b/include/linux/binfmts.h index 12413a1e2784..a2e0013900ef 100644 --- a/include/linux/binfmts.h +++ b/include/linux/binfmts.h @@ -28,6 +28,7 @@ struct linux_binprm{ char buf[BINPRM_BUF_SIZE]; #ifdef CONFIG_MMU struct vm_area_struct *vma; + unsigned long vma_pages; #else # define MAX_ARG_PAGES 32 struct page *page[MAX_ARG_PAGES]; |