Re: Should mmap MAP_LOCKED fail if mm_poppulate fails?

[Michal Hocko]

Hi,
it seems that the initial email got lost (or ignored). I would like to
revive it again. I've cooked up a potential fix to this issue which will
follow as a reply to this email.

The first patch is dumb and straightforward. It should be safe as is and
also good without the follow up 2 patches which try to handle potential
allocation failures in the do_munmap path more gracefully. As we still
do not fail small allocations even the first patch could be simplified
a bit and the retry loop replaced by a BUG_ON right away. But I felt this
would better be done robust.

An obvious alternative would be patching the man pages to mention the
subtle difference between mlock and MAP_LOCKED semantic. I have checked
debian code search and it shown some applications relying on MAP_LOCKED
but I have no idea whether they really require the mlock all-or-nothing
fault in semantic.

Any thoughts, ideas?

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/