On Wed, Jul 22, 2015 at 10:32:20AM -0400, Eric B Munson wrote:

On Wed, 22 Jul 2015, Kirill A. Shutemov wrote:

quoted

On Tue, Jul 21, 2015 at 03:59:40PM -0400, Eric B Munson wrote:

quoted

The cost of faulting in all memory to be locked can be very high when
working with large mappings.  If only portions of the mapping will be
used this can incur a high penalty for locking.

Now that we have the new VMA flag for the locked but not present state,
expose it as an mmap option like MAP_LOCKED -> VM_LOCKED.

What is advantage over mmap() + mlock(MLOCK_ONFAULT)?

There isn't one, it was added to maintain parity with the
mlock(MLOCK_LOCK) -> mmap(MAP_LOCKED) set.  I think not having will lead
to confusion because we have MAP_LOCKED so why don't we support
LOCKONFAULT from mmap as well.

I don't think it's ia good idea to spend bits in flags unless we have a
reason for that.

BTW, you have typo on sparc: s/0x8000/0x80000/.


-- 
 Kirill A. Shutemov