On Sun, 10 Oct 1999, Manfred Spraul wrote:
> AFAIK the problem is OOM:
> * a process accesses a not-present, ie page fault:
> ...
> handle_mm_fault(): this process own mm->mmap_sem.
> ->handle_pte_fault().
> -> (eg.) do_wp_page().
> -> get_free_page().
> now get_free_page() notices that there is no free memory.
> --> wakeup kswapd.
>
> * the swapper runs, and it tries to swap out data from that process.
> mm->mmap_sem is already acquired --> lock-up.
Nasty... But adding a big lock around all traversals of the mm->mmap will
hurt like hell. Let's see... The problem is in swap_out_mm(), right? And
it looks for the first suitable page. OK, it looks like we can get a
deadlock only if we call try_to_free_pages() with ->mmap_sem grabbed and
__GFP_WAIT in flags. Umhm... Called only from __get_free_pages() which
sets PF_MEMALLOC... OK, I'll try to look at it.
-
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/