On Mon, Oct 30, 2023 at 12:15:19PM +0300, Dan Carpenter wrote:
> On Mon, Oct 30, 2023 at 05:11:33PM +0800, Herbert Xu wrote:
> > On Mon, Oct 30, 2023 at 12:02:59PM +0300, Dan Carpenter wrote:
> > > Static checkers insist that the mpi_alloc() allocation can fail so add
> > > a check to prevent a NULL dereference. Small allocations like this
> > > can't actually fail in current kernels, but adding a check is very
> > > simple and makes the static checkers happy.
> > >
> > > Fixes: 6637e11e4ad2 ("crypto: rsa - allow only odd e and restrict value in FIPS mode")
> > > Signed-off-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
> > > ---
> > > crypto/rsa.c | 2 ++
> > > 1 file changed, 2 insertions(+)
> >
> > Nack. Please fix the static checker instead.
>
> The checker is not wrong. Kernel policy is that every allocation has
> to be checked for failure. In *current* kernels it won't but we have
> discussed changing this so the policy makes sense.
One way to fix the code would be to take gfp parameter and let people
pass a GFP_NOFAIL. Unless there is a GFP_NOFAIL then the policy is that
the allocation can fail.
Or you could take the brtfs v1 approach and follow every kmalloc() with
a:
p = kmalloc();
BUG_on(!p);
regards,
dan carpenter
