Re: [GIT PULL] s390 fixes for 6.6-rc7

[Linus Torvalds]

Just re-quoting my suggestion here and adding Andy and Dmitry, who did
the original bitmap_alloc() helper interfaces a few years ago.

Also adding Kees in case he has any hardening suggestions, since this
is about (incorrect) overflow handling.

Kees: see my rant about mindlessly doing overflow handling in the wrong place in

   https://lore.kernel.org/all/CAHk-=wgTUz1bdY6zvsN4ED0arCLE8Sb==1GH8d0sjm5bu7zesQ@xxxxxxxxxxxxxx/

in case you or somebody has a better idea for BITS_TO_LONG handling
than just "you need to check for zero before and after".

             Linus

On Sat, 21 Oct 2023 at 10:56, Linus Torvalds
<torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
>
> If you *do* want to add proper overflow handling, you'd need to either
> fix BITS_TO_LONGS() some way (which is actually non-trivial since it
> needs to be able to stay a constant and only use the argument once),
> or you do something like
>
>         if (!bits)
>                 return ZERO_SIZE_PTR;
>         longs = BITS_TO_LONG(bits);
>         if (!longs)
>                 return NULL;
>         return vzalloc(longs * sizeof(long));
>
> and I'd suggest maybe we should
>
>  (a) do the above checking in our bitmap_alloc() routines
>
>  (b) also change our bitmap_alloc() routines to take 'size_t' instead
> of 'unsigned int' bit counts
>
>  (c) and finally, add that vzalloc() case, but simply using
>
>         kvmalloc_array(n, size, flags | __GFP_ZERO);
>
> instead.