Re: try_to_swap_out() return value problem?

From: Linus Torvalds (torvalds@transmeta.com)
Date: Thu Jan 04 2001 - 11:24:35 EST


On Thu, 4 Jan 2001, Rik van Riel wrote:

> On Wed, 3 Jan 2001, Linus Torvalds wrote:
> > On Thu, 4 Jan 2001, Marcelo Tosatti wrote:
>
> > I agree that the return value of swap_out() is fairly meaningless. It's
> > been fairly meaningless for a long time now, and it's entirely possible
> > that the "while (swap_out())" loop should be just something like
> >
> > /* Scan the VM space, try to clean up the page tables a bit */
> > for (i = 0 ; i <= nr_threads >> priority; i++)
> > swap_out(gfp_mask);
>
> The problem with this is that it means that page aging of
> the mapped active pages is no longer balanced against the
> aging of the unmapped active pages.

Is there any reason to not just remove the lines

        if (!onlist)
                /* The page is still mapped, so it can't be freeable... */
                age_page_down_ageonly(page);

        /*
         * If the page is in active use by us, or if the page
         * is in active use by others, don't unmap it or
         * (worse) start unneeded IO.
         */
        if (page->age > 0)
                goto out_failed;

from vmscan?

They look like a complete hack, although an understandable one from the
time when the virtual memory scan used to actually do IO as well.

These days, if you think of the VM scanning as just a "shrink the page
tables" operation, the down-aging doesn't make much sense. That will
happen once the page has been moved to the lists, no?

                Linus

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



This archive was generated by hypermail 2b29 : Sun Jan 07 2001 - 21:00:18 EST