[PATCH 4.9 54/96] bpf, array: fix overflow in max_entries and undefined behavior in index_mask

From: Greg Kroah-Hartman
Date: Mon Jan 15 2018 - 07:44:59 EST


4.9-stable review patch. If anyone has any objections, please let me know.

------------------

From: Daniel Borkmann <daniel@xxxxxxxxxxxxx>

commit bbeb6e4323dad9b5e0ee9f60c223dd532e2403b1 upstream.

syzkaller tried to alloc a map with 0xfffffffd entries out of a userns,
and thus unprivileged. With the recently added logic in b2157399cc98
("bpf: prevent out-of-bounds speculation") we round this up to the next
power of two value for max_entries for unprivileged such that we can
apply proper masking into potentially zeroed out map slots.

However, this will generate an index_mask of 0xffffffff, and therefore
a + 1 will let this overflow into new max_entries of 0. This will pass
allocation, etc, and later on map access we still enforce on the original
attr->max_entries value which was 0xfffffffd, therefore triggering GPF
all over the place. Thus bail out on overflow in such case.

Moreover, on 32 bit archs roundup_pow_of_two() can also not be used,
since fls_long(max_entries - 1) can result in 32 and 1UL << 32 in 32 bit
space is undefined. Therefore, do this by hand in a 64 bit variable.

This fixes all the issues triggered by syzkaller's reproducers.

Fixes: b2157399cc98 ("bpf: prevent out-of-bounds speculation")
Reported-by: syzbot+b0efb8e572d01bce1ae0@xxxxxxxxxxxxxxxxxxxxxxxxx
Reported-by: syzbot+6c15e9744f75f2364773@xxxxxxxxxxxxxxxxxxxxxxxxx
Reported-by: syzbot+d2f5524fb46fd3b312ee@xxxxxxxxxxxxxxxxxxxxxxxxx
Reported-by: syzbot+61d23c95395cc90dbc2b@xxxxxxxxxxxxxxxxxxxxxxxxx
Reported-by: syzbot+0d363c942452cca68c01@xxxxxxxxxxxxxxxxxxxxxxxxx
Signed-off-by: Daniel Borkmann <daniel@xxxxxxxxxxxxx>
Signed-off-by: Alexei Starovoitov <ast@xxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>

---
kernel/bpf/arraymap.c | 18 +++++++++++++++---
1 file changed, 15 insertions(+), 3 deletions(-)

--- a/kernel/bpf/arraymap.c
+++ b/kernel/bpf/arraymap.c
@@ -49,7 +49,7 @@ static struct bpf_map *array_map_alloc(u
u32 elem_size, index_mask, max_entries;
bool unpriv = !capable(CAP_SYS_ADMIN);
struct bpf_array *array;
- u64 array_size;
+ u64 array_size, mask64;

/* check sanity of attributes */
if (attr->max_entries == 0 || attr->key_size != 4 ||
@@ -65,13 +65,25 @@ static struct bpf_map *array_map_alloc(u
elem_size = round_up(attr->value_size, 8);

max_entries = attr->max_entries;
- index_mask = roundup_pow_of_two(max_entries) - 1;

- if (unpriv)
+ /* On 32 bit archs roundup_pow_of_two() with max_entries that has
+ * upper most bit set in u32 space is undefined behavior due to
+ * resulting 1U << 32, so do it manually here in u64 space.
+ */
+ mask64 = fls_long(max_entries - 1);
+ mask64 = 1ULL << mask64;
+ mask64 -= 1;
+
+ index_mask = mask64;
+ if (unpriv) {
/* round up array size to nearest power of 2,
* since cpu will speculate within index_mask limits
*/
max_entries = index_mask + 1;
+ /* Check for overflows. */
+ if (max_entries < attr->max_entries)
+ return ERR_PTR(-E2BIG);
+ }

array_size = sizeof(*array);
if (percpu)