Re: [Lhms-devel] Re: 2.6.14-mm2

[Kamezawa Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>]

Dave Hansen wrote:

> 
> Can you explain in a little bit more detail why this matters, and
> exactly how it fixes your problem.  I'm not sure it's correct.
> 
Ah, okay.

It's just because free_area[] is not initaialized at all if this is not called.
It is list.next and list.prev has bad value.
Then, the first free_page(page) will cause panic.

> Also, if you're doing hot-adds of _new_ zones at runtime, you need to do
> something fancy with the zonelist locking that I never got around to
> because nobody needs it yet.  See something along these lines:
> 
When node 0's higmem size is 0 at boot time, I have to add new page into empty zone.
This happens because my machine has only 700M mem.
I use mem=500M and hot add extra 200M memory for testing.

"Nobody needs " is sane in real world. But it's useful to my tiny test enveironment.

Could you spin out initializing free_area[] from
init_currently_empty_zone(zone, zone_start_pfn, size);
Then I'll be happy.

-- Kame
> http://www.sr71.net/patches/2.6.14/2.6.14-rc2-git8-mhp1/broken-out/E2-for-debugging-handle-add-to-empty-zone.patch
> 
> -- Dave
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 
>