2008-02-15 14:52:31

by Larry Woodman

Subject: Problem with /proc/sys/vm/lowmem_reserve_ratio ???

balance_pgdat() calls zone_watermark_ok() three times, the first call
passes a zero(0) in as the 4th argument. This 4th argument is the
classzone_idx which is used as the index into the
zone->lowmem_reserve[] array. Since setup_per_zone_lowmem_reserve()
always sets the zone->lowmem_reserve[0] = 0(because there is nothing
below the DMA zone), zone_watermark_ok() will not consider the
lowmem_reserve pages when zero is passed as the 4th arg. Shouldnt this
4th argument be either "i" or "nr_zones - 1" ???

-------------------------------------------------------------------------
--- linux-2.6.24.noarch/mm/vmscan.c.orig 2008-02-13
11:14:55.000000000 -0500
+++ linux-2.6.24.noarch/mm/vmscan.c 2008-02-13 11:15:02.000000000
-0500
@@ -1375,7 +1375,7 @@ loop_again:
continue;

if (!zone_watermark_ok(zone, order, zone->pages_high,

- 0, 0)) {
+ i, 0)) {
end_zone = i;
break;
-------------------------------------------------------------------------
--- linux-2.6.24.noarch/mm/vmscan.c.orig 2008-02-13
11:14:55.000000000 -0500
+++ linux-2.6.24.noarch/mm/vmscan.c 2008-02-13 11:16:35.000000000
-0500
@@ -1375,7 +1375,7 @@ loop_again:
continue;

if (!zone_watermark_ok(zone, order, zone->pages_high,

- 0, 0)) {
+ nr_zones - 1, 0)) {
end_zone = i;
break;
}
-------------------------------------------------------------------------

2008-02-18 14:33:18

by Jeff Moyer

Subject: Re: Problem with /proc/sys/vm/lowmem_reserve_ratio ???

Larry Woodman <[email protected]> writes:

> balance_pgdat() calls zone_watermark_ok() three times, the first call
> passes a zero(0) in as the 4th argument. This 4th argument is the
> classzone_idx which is used as the index into the
> zone->lowmem_reserve[] array. Since setup_per_zone_lowmem_reserve()
> always sets the zone->lowmem_reserve[0] = 0(because there is nothing
> below the DMA zone), zone_watermark_ok() will not consider the
> lowmem_reserve pages when zero is passed as the 4th arg. Shouldnt this
> 4th argument be either "i" or "nr_zones - 1" ???

That certainly looks like the intent of this code. I'd say patch 1,
passing i in as the 4th arg. Rik, what do you think?

Cheers,

Jeff

>
> -------------------------------------------------------------------------
> --- linux-2.6.24.noarch/mm/vmscan.c.orig 2008-02-13
> 11:14:55.000000000 -0500
> +++ linux-2.6.24.noarch/mm/vmscan.c 2008-02-13 11:15:02.000000000
> -0500
> @@ -1375,7 +1375,7 @@ loop_again:
> continue;
>
> if (!zone_watermark_ok(zone, order, zone->pages_high,
>
> - 0, 0)) {
> + i, 0)) {
> end_zone = i;
> break;
> -------------------------------------------------------------------------
> --- linux-2.6.24.noarch/mm/vmscan.c.orig 2008-02-13
> 11:14:55.000000000 -0500
> +++ linux-2.6.24.noarch/mm/vmscan.c 2008-02-13 11:16:35.000000000
> -0500
> @@ -1375,7 +1375,7 @@ loop_again:
> continue;
>
> if (!zone_watermark_ok(zone, order, zone->pages_high,
>
> - 0, 0)) {
> + nr_zones - 1, 0)) {
> end_zone = i;
> break;
> }
> -------------------------------------------------------------------------

2008-02-18 15:49:17

by Rik van Riel

Subject: Re: Problem with /proc/sys/vm/lowmem_reserve_ratio ???

On Mon, 18 Feb 2008 09:33:05 -0500
Jeff Moyer <[email protected]> wrote:

> Larry Woodman <[email protected]> writes:
>
> > balance_pgdat() calls zone_watermark_ok() three times, the first call
> > passes a zero(0) in as the 4th argument. This 4th argument is the
> > classzone_idx which is used as the index into the
> > zone->lowmem_reserve[] array. Since setup_per_zone_lowmem_reserve()
> > always sets the zone->lowmem_reserve[0] = 0(because there is nothing
> > below the DMA zone), zone_watermark_ok() will not consider the
> > lowmem_reserve pages when zero is passed as the 4th arg. Shouldnt this
> > 4th argument be either "i" or "nr_zones - 1" ???
>
> That certainly looks like the intent of this code. I'd say patch 1,
> passing i in as the 4th arg. Rik, what do you think?

Yes, that is probably the best option.

--
All rights reversed.