Pekka Enberg wrote:
>
> Yup, if it's not too much trouble Larry, I'd appreciate if you gave it
> a spin before I apply the sucker. Note to regression trackers, I don't
> think this should go to -stable because it depends on other patches in
> mainline and there's an obvious workaround for the problem (disable
> CONFIG_SLUB_DEBUG_ON). Furthermore, it's very unlikely people will hit
> this on production configurations anyway.
Pekka,
I have been running the patch for ~24 hours with slub_debug=0 as a
boot option. So far, no problems. My workload has not involved as many
network operations as some times, but my memory has not fragmented.
The output of 'cat /proc/buddyinfo' is as follows:
Node 0, zone DMA 8 5 5 4 5 5 5
4 3 0 1
Node 0, zone DMA32 1804 1585 3007 36 0 0 0
0 1 0 0
As might be expected, the high-order fragments are gone, but the O(1)
pieces have not been depleted.
Larry
--
To unsubscribe from this list: send the line "unsubscribe kernel-testers" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
