On Tue, Sep 28, 2021 at 02:29:43PM +0200, Daniel Borkmann wrote: > On 9/25/21 7:31 AM, Tatsuhiko Yasumatsu wrote: > > In prealloc_elems_and_freelist(), the multiplication to calculate the > > size passed to bpf_map_area_alloc() could lead to an integer overflow. > > As a result, out-of-bounds write could occur in pcpu_freelist_populate() > > as reported by KASAN: > > > > [...] > > [ 16.968613] BUG: KASAN: slab-out-of-bounds in pcpu_freelist_populate+0xd9/0x100 > > [ 16.969408] Write of size 8 at addr ffff888104fc6ea0 by task crash/78 > > [ 16.970038] > > [ 16.970195] CPU: 0 PID: 78 Comm: crash Not tainted 5.15.0-rc2+ #1 > > [ 16.970878] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1ubuntu1.1 04/01/2014 > > [ 16.972026] Call Trace: > > [ 16.972306] dump_stack_lvl+0x34/0x44 > > [ 16.972687] print_address_description.constprop.0+0x21/0x140 > > [ 16.973297] ? pcpu_freelist_populate+0xd9/0x100 > > [ 16.973777] ? pcpu_freelist_populate+0xd9/0x100 > > [ 16.974257] kasan_report.cold+0x7f/0x11b > > [ 16.974681] ? pcpu_freelist_populate+0xd9/0x100 > > [ 16.975190] pcpu_freelist_populate+0xd9/0x100 > > [ 16.975669] stack_map_alloc+0x209/0x2a0 > > [ 16.976106] __sys_bpf+0xd83/0x2ce0 > > [...] > > > > The possibility of this overflow was originally discussed in [0], but > > was overlooked. > > > > Fix the integer overflow by casting one operand to u64. > > > > [0] https://lore.kernel.org/bpf/728b238e-a481-eb50-98e9-b0f430ab01e7@xxxxxxxxx/ > > > > Fixes: 557c0c6e7df8 ("bpf: convert stackmap to pre-allocation") > > Signed-off-by: Tatsuhiko Yasumatsu <th.yasumatsu@xxxxxxxxx> > > --- > > kernel/bpf/stackmap.c | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/kernel/bpf/stackmap.c b/kernel/bpf/stackmap.c > > index 09a3fd97d329..8941dc83a769 100644 > > --- a/kernel/bpf/stackmap.c > > +++ b/kernel/bpf/stackmap.c > > @@ -66,7 +66,7 @@ static int prealloc_elems_and_freelist(struct bpf_stack_map *smap) > > u32 elem_size = sizeof(struct stack_map_bucket) + smap->map.value_size; > > Thanks a lot for the fix, Tatsuhiko! Could we just change the above elem_size to u64 instead? Thank you for your review, Daniel! Yes, I think it's possible to just change elem_size to u64. We just have to be careful to cast one operand (smap->map.value_size) to u64, so that the integer overflow won't happen in 32-bit architectures. This is necessary because in 32-bit architectures, the result of sizeof() is a 32-bit integer. I will update the patch. > > > int err; > > - smap->elems = bpf_map_area_alloc(elem_size * smap->map.max_entries, > > + smap->elems = bpf_map_area_alloc((u64)elem_size * smap->map.max_entries, > > smap->map.numa_node); > > if (!smap->elems) > > return -ENOMEM; > > > > Best, > Daniel Best regards, Tatsuhiko Yasumatsu
