On Thu, Oct 22, 2020 at 09:21:35AM -0400, Jeff Layton wrote:
> On Thu, 2020-10-22 at 10:03 +0800, Luo Meng wrote:
> > When the sum of fl->fl_start and l->l_len overflows,
> > UBSAN shows the following warning:
> >
> > UBSAN: Undefined behaviour in fs/locks.c:482:29
> > signed integer overflow: 2 + 9223372036854775806
> > cannot be represented in type 'long long int'
> > Call Trace:
> > __dump_stack lib/dump_stack.c:77 [inline]
> > dump_stack+0xe4/0x14e lib/dump_stack.c:118
> > ubsan_epilogue+0xe/0x81 lib/ubsan.c:161
> > handle_overflow+0x193/0x1e2 lib/ubsan.c:192
> > flock64_to_posix_lock fs/locks.c:482 [inline]
> > flock_to_posix_lock+0x595/0x690 fs/locks.c:515
> > fcntl_setlk+0xf3/0xa90 fs/locks.c:2262
> > do_fcntl+0x456/0xf60 fs/fcntl.c:387
> > __do_sys_fcntl fs/fcntl.c:483 [inline]
> > __se_sys_fcntl fs/fcntl.c:468 [inline]
> > __x64_sys_fcntl+0x12d/0x180 fs/fcntl.c:468
> > do_syscall_64+0xc8/0x5a0 arch/x86/entry/common.c:293
> > entry_SYSCALL_64_after_hwframe+0x49/0xbe
> >
> > Fix it by moving -1 forward.
> >
> > Signed-off-by: Luo Meng <luomeng12@xxxxxxxxxx>
> > ---
> > fs/locks.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/fs/locks.c b/fs/locks.c
> > index 1f84a03601fe..8489787ca97e 100644
> > --- a/fs/locks.c
> > +++ b/fs/locks.c
> > @@ -542,7 +542,7 @@ static int flock64_to_posix_lock(struct file *filp, struct file_lock *fl,
> > if (l->l_len > 0) {
> > if (l->l_len - 1 > OFFSET_MAX - fl->fl_start)
> > return -EOVERFLOW;
> > - fl->fl_end = fl->fl_start + l->l_len - 1;
> > + fl->fl_end = fl->fl_start - 1 + l->l_len;
> >
> > } else if (l->l_len < 0) {
> > if (fl->fl_start + l->l_len < 0)
>
> Wow, ok. Interesting that the order would have such an effect here, but
> it seems legit. I'll plan to merge this for v5.11. Let me know if we
> need to get this in earlier.
It's the kind of pedantic correctness thing that should be merged because
C doesn't exactly define the behaviour. eg a sign-magnitude machine
will behave differently from a twos-complement machine. The fact that
nobody's made a sign-magnitude integer arithmetic machine in the last
60 years does not matter to the C spec.
It's a shame there's no uoff_t since it would be defined.
