Re: locking question: do_mmap(), do_munmap()

[kanoj@google.engr.sgi.com (Kanoj Sarcar)]

> 
> What about something like a rw-semaphore which protects the vma list:
> vma-list modifiers [ie merge_segments(), insert_vm_struct() and
> do_munmap()] grab it exclusive, swapper grabs it "shared, starve
> exclusive".
> All other vma-list readers are protected by mm->mmap_sem.
> 
> This should not dead-lock, and no changes are required in
> vm_ops->swapout().
>

I have tried to follow most of the logic and solutions proposed
on this thread. This is the best solution, imo. In fact, I had
already coded something on these lines against a 2.2.10 kernel,
which I still have around. I will try to port this against a 
2.3.19 kernel over the next couple of days and post it for
everyone to review.

Thanks.

Kanoj
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://humbolt.geo.uu.nl/Linux-MM/