Re: [v2 3/5] arm64: mte: implement CONFIG_ARM64_MTE_COMP

[Yury Norov]

> > +		bitmap_set_value_unaligned((unsigned long *)buf, largest_idx,
> > +					   bit_pos, 4);
> 
> > +		bitmap_set_value_unaligned((unsigned long *)buf, largest_idx,
> > +					   bit_pos, 6);
> 
> > +		bitmap_set_value_unaligned((unsigned long *)buf, tags[i],
> > +					   bit_pos, 4);
> 
> > +		bitmap_set_value_unaligned((unsigned long *)buf, 0, bit_pos, 4);
> 
> > +		bitmap_set_value_unaligned((unsigned long *)buf, sizes[i],
> > +					   bit_pos, 7);
> 
> > +	largest_idx = bitmap_get_value_unaligned((unsigned long *)buf, bit_pos,
> > +						 l_bits);
> 
> > +		r_tags[i] = bitmap_get_value_unaligned((unsigned long *)buf,
> > +						       bit_pos, 4);
> 
> > +		r_sizes[i] = bitmap_get_value_unaligned((unsigned long *)buf,
> > +							bit_pos, 7);
> 
> These castings is a red flag. bitmap API shouldn't be used like this. Something
> is not okay here.

Big-endian arches are not OK. Out-of-boundary access is not OK when
the buf is not exactly a multiple of words.

> > +void ea0_release_handle(u64 handle)
> > +{
> > +	void *storage = ea0_storage(handle);
> > +	int size = ea0_storage_size(handle);
> > +	struct kmem_cache *c;
> 
> > +	if (!handle || !storage)
> > +		return;
> 
> You use handle before this check. Haven't you run static analysers?

This approach is called 'defensive programming' as I learned from
previous iteration. Another interesting thing is that the only caller
of the function in patch #5 explicitly checks the handle for NULL, so
we're surely double-defensed here.

        +void _mte_free_saved_tags(void *storage)
        +{
        +       unsigned long handle = xa_to_value(storage);
        +       int size;
        +
        +       if (!handle)
        +               return;
        +       size = ea0_storage_size(handle);
        +       ea0_release_handle(handle);
        +}

_mte_free_saved_tags() calculates size, but doesn't use it in any form,
just to calculate it again in callee...

Thanks,
Yury