On Tue, Oct 25, 2016 at 07:11:54AM +0200, Miroslav Rovis wrote: > On 161021-11:04-0400, Rich Freeman wrote: > > On Fri, Oct 21, 2016 at 10:49 AM, Mick <michaelkintz...@gmail.com> wrote: > > > https://github.com/dirtycow/dirtycow.github.io/wiki/VulnerabilityDetails > > > > Not yet: > > https://bugs.gentoo.org/show_bug.cgi?id=597624 > > > > We are talking grsecurity-patched (kind of stable[*]) kernel sources, > the =sys-kernel/hardened-sources-4.4.8-r1 package [**]. > > I read most of the discussion, and I could easily patch the gup.c and > mm.h in question, but those files need to be patched before application > of the grsecurity patch, and that is a little more complex work.
Did you tried it? The patch attached comes straight from the git repo, just run: # cd /usr/src/linux # patch -p1 < path/to/patch It'll likely work. > > Has anybody done this, as I have limited time available to practice user > patching (which in its simplest form, I was able to do here: > >=dev-libs/nss-3.24 - Add USE flag to enable SSL key > https://bugs.gentoo.org/show_bug.cgi?id=587116#c2 ), in case it can be > done with user patching, of course. > > Anyone? > > Regards! > --- > [*] kind of stable, because there are, since about 1 yrs ago, only > testing kernel available for the non-paying users ;-( > > [**] I have to use 4.4.8.r1 because recent kernel all crash with libirt > and qemu which I am trying to use: > https://bugs.gentoo.org/show_bug.cgi?id=597554 > -- > Miroslav Rovis > Zagreb, Croatia > http://www.CroatiaFidelis.hr -- Fernando Rodriguez
commit 1294d355881cc5c3421d24fee512f16974addb6c
Author: Linus Torvalds <torva...@linux-foundation.org>
Date: Thu Oct 13 13:07:36 2016 -0700
mm: remove gup_flags FOLL_WRITE games from __get_user_pages()
commit 19be0eaffa3ac7d8eb6784ad9bdbc7d67ed8e619 upstream.
This is an ancient bug that was actually attempted to be fixed once
(badly) by me eleven years ago in commit 4ceb5db9757a ("Fix
get_user_pages() race for write access") but that was then undone due to
problems on s390 by commit f33ea7f404e5 ("fix get_user_pages bug").
In the meantime, the s390 situation has long been fixed, and we can now
fix it by checking the pte_dirty() bit properly (and do it better). The
s390 dirty bit was implemented in abf09bed3cce ("s390/mm: implement
software dirty bits") which made it into v3.9. Earlier kernels will
have to look at the page state itself.
Also, the VM has become more scalable, and what used a purely
theoretical race back then has become easier to trigger.
To fix it, we introduce a new internal FOLL_COW flag to mark the "yes,
we already did a COW" rather than play racy games with FOLL_WRITE that
is very fundamental, and then use the pte dirty flag to validate that
the FOLL_COW flag is still valid.
Reported-and-tested-by: Phil "not Paul" Oester <ker...@linuxace.com>
Acked-by: Hugh Dickins <hu...@google.com>
Reviewed-by: Michal Hocko <mho...@suse.com>
Cc: Andy Lutomirski <l...@kernel.org>
Cc: Kees Cook <keesc...@chromium.org>
Cc: Oleg Nesterov <o...@redhat.com>
Cc: Willy Tarreau <w...@1wt.eu>
Cc: Nick Piggin <npig...@gmail.com>
Cc: Greg Thelen <gthe...@google.com>
Signed-off-by: Linus Torvalds <torva...@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gre...@linuxfoundation.org>
diff --git a/include/linux/mm.h b/include/linux/mm.h
index cfebb74..f0ffa01 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -2112,6 +2112,7 @@ static inline struct page *follow_page(struct vm_area_struct *vma,
#define FOLL_MIGRATION 0x400 /* wait for page to replace migration entry */
#define FOLL_TRIED 0x800 /* a retry, previous pass started an IO */
#define FOLL_MLOCK 0x1000 /* lock present pages */
+#define FOLL_COW 0x4000 /* internal GUP flag */
typedef int (*pte_fn_t)(pte_t *pte, pgtable_t token, unsigned long addr,
void *data);
diff --git a/mm/gup.c b/mm/gup.c
index deafa2c..4b0b7e7 100644
--- a/mm/gup.c
+++ b/mm/gup.c
@@ -58,6 +58,16 @@ static int follow_pfn_pte(struct vm_area_struct *vma, unsigned long address,
return -EEXIST;
}
+/*
+ * FOLL_FORCE can write to even unwritable pte's, but only
+ * after we've gone through a COW cycle and they are dirty.
+ */
+static inline bool can_follow_write_pte(pte_t pte, unsigned int flags)
+{
+ return pte_write(pte) ||
+ ((flags & FOLL_FORCE) && (flags & FOLL_COW) && pte_dirty(pte));
+}
+
static struct page *follow_page_pte(struct vm_area_struct *vma,
unsigned long address, pmd_t *pmd, unsigned int flags)
{
@@ -92,7 +102,7 @@ retry:
}
if ((flags & FOLL_NUMA) && pte_protnone(pte))
goto no_page;
- if ((flags & FOLL_WRITE) && !pte_write(pte)) {
+ if ((flags & FOLL_WRITE) && !can_follow_write_pte(pte, flags)) {
pte_unmap_unlock(ptep, ptl);
return NULL;
}
@@ -352,7 +362,7 @@ static int faultin_page(struct task_struct *tsk, struct vm_area_struct *vma,
* reCOWed by userspace write).
*/
if ((ret & VM_FAULT_WRITE) && !(vma->vm_flags & VM_WRITE))
- *flags &= ~FOLL_WRITE;
+ *flags |= FOLL_COW;
return 0;
}
signature.asc
Description: Digital signature
