Messages in this thread | | | Date | Tue, 8 Sep 1998 03:21:09 +0100 | From | Jamie Lokier <> | Subject | Re: lock_kernel in page fault code |
| |
I noticed arch/i386/mm/fault.c:do_page_fault does `down(&mm-mmap_sem)' around most of the code, and `lock_kernel()' around a small bit at the end.
Other architectures do the same semaphore but do `lock_kernel()' right from the beginning.
Questions:
1. Is the i386 missing this, or is it better optimised?
2. So should lock_kernel() be moved for the other architectures?
3. Is there any good reason why individual vmas couldn't be locked instead of the whole mmap list, so page faults can be handled on different processors concurrently?
-- Jamie
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/faq.html
| |