Re: [PATCH] oom killer (Core)

[William Lee Irwin III]

On Fri, Dec 10, 2004 at 09:43:36AM -0800, William Lee Irwin III wrote:
>> Well, the only way I see this happening is the process exiting followed
>> by use_mm() on init_mm for unobvious reasons (perhaps reasons not in
>> the tree).

On Fri, Dec 10, 2004 at 06:55:04PM +0100, Andrea Arcangeli wrote:
> I don't see the problem with use_mm. use_mm has either the mm set to
> ctx->mm or to NULL, and ctx->mm is set to the mm of the process calling
> io_setup.
> The only thing using init_mm is the idle task/swapper as far as I can
> tell, kernel threads and exiting tasks have a NULL mm.

Yet those don't appear in the tasklist, so some task in the tasklist
has to get ->mm set to &init_mm. The notion above was that the init_mm
check was to handle some out-of-tree attempt to do aio from kernel threads.


-- wli
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/