Re[2]: your mail on mmap() to the kernel list

[Peter Zaitsev]

Hello Andrea,

>>
>> OTOH, I doubt it would help real-world workloads where the
>> application maps and unmaps areas of different sizes and
>> actually does something with the memory instead of just
>> mapping and unmapping it ;)))

AA> exactly, while that would be simple to implement and very lightweight at
AA> runtime, that's not enough to mathematically drop the complexity of the
AA> get_unmapped_area algorithm. It would optimize only the case where
AA> there's no fragmentation of the mapped virtual address space.

And also will optimize all mappings of 4K and (which are at least 70%
in mu case) :)

AA> For finding the best fit in the heap with O(log(N)) complexity (rather
AA> than the current O(N) complexity of the linked list) one tree indexed by
AA> the size of each hole would be necessary.

This of course would be the best way.

-- 
Best regards,
 Peter                            mailto:[e-mail]