On Thu, May 18, 2017 at 01:43:59PM +0200, Michal Hocko wrote:
> On Mon 15-05-17 15:12:18, Kirill A. Shutemov wrote:
> [...]
> > @@ -195,6 +207,16 @@ arch_get_unmapped_area_topdown(struct file *filp,
> > const unsigned long addr0,
> > info.length = len;
> > info.low_limit = PAGE_SIZE;
> > info.high_limit = get_mmap_base(0);
> > +
> > + /*
> > + * If hint address is above DEFAULT_MAP_WINDOW, look for unmapped area
> > + * in the full address space.
> > + *
> > + * !in_compat_syscall() check to avoid high addresses for x32.
> > + */
> > + if (addr > DEFAULT_MAP_WINDOW && !in_compat_syscall())
> > + info.high_limit += TASK_SIZE_MAX - DEFAULT_MAP_WINDOW;
> > +
> > info.align_mask = 0;
> > info.align_offset = pgoff << PAGE_SHIFT;
> > if (filp) {
>
> I have two questions/concerns here. The above assumes that any address above
> 1<<47 will use the _whole_ address space. Is this what we want?
Yes, I believe so.
> What if somebody does mmap(1<<52, ...) because he wants to (ab)use 53+
> bits for some other purpose? Shouldn't we cap the high_limit by the
> given address?
This would screw existing semantics of hint address -- "map here if
free, please".
> Another thing would be that
> /* requesting a specific address */
> if (addr) {
> addr = PAGE_ALIGN(addr);
> vma = find_vma(mm, addr);
> if (TASK_SIZE - len >= addr &&
> (!vma || addr + len <= vma->vm_start))
> return addr;
> }
>
> would fail for mmap(-1UL, ...) which is good because we do want to
> fallback to vm_unmapped_area and have randomized address which is
> ensured by your info.high_limit += ... but that wouldn't work for
> mmap(1<<N, ...) where N>47. So the first such mapping won't be
> randomized while others will be. This is quite unexpected I would say.
> So it should be documented at least or maybe we want to skip the above
> shortcut for addr > DEFAULT_MAP_WINDOW altogether.
Again, you're missing existing semantics of hint address. You may have a
reason to set hint address above 47-bit, besides getting access to full
address space.
--
Kirill A. Shutemov
