Splitting the page migration patches out of the memory hotplug patch

Splitting the page migration patches out of the memory hotplug patch

[Ray Bryant]

(Addling linux-mm, Marcello and Hirakazu....what I'm up
to is trying to create a roll-up patch that contains just
the memory migration code from mhp3, since I need page
migration for some work I am doing.)

Dave,

Well, the only other big change I made (other than dropping
P32-memsection_migrate.patch) to get it to compile and boot
for Altix were as follows:

(I'll work on testing it after the holiday.)

P29-add-memory-migration-to-Kconfig-ia64.patch

         Add memory migration to the config menu
         for ia64.

P30-remove-page_under_capture.patch

         removed page_under_capture() from the end
         of shrink_cache() in mm/vmscan.c.  This
         is not defined in the P series of patches.
         This particular call was introduced by patch
         P01-steal_page_from_lru.patch

(I'm not sure how to number these patches to fit in with
your scheme, so just made some stuff up.)

(Oh yeah, this is in top of 2.6.10-rc2-mm4.)

Now it would be nice if we could figure out a way to keep
these patchsets distinct (i. e. so work on page migration
and hotplug can continue without me redoing this every
week or two.)

One way to do that would be to fix it so that the page
migration patches are first in the hotplug patch, or to
separate the two out and assume that hotplug patch goes
on top of the page migration patch.  How would you like
to go about this? (I'll take a whack at moving them to
the top of the mhp3 series file and see how much trouble
I get into....)

PS:  It doesn't look like Marcello and Hirakazu's
migration patch is part of your P* series.  Is
that correct?

-- 
Best Regards,
Ray
-----------------------------------------------
                   Ray Bryant
512-453-9679 (work)         512-507-7807 (cell)
raybry@sgi.com             raybry@austin.rr.com
The box said: "Requires Windows 98 or better",
            so I installed Linux.
-----------------------------------------------
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"aart@kvack.org"> aart@kvack.org </a>

Re: Splitting the page migration patches out of the memory hotplug patch

[Marcelo Tosatti]

On Fri, Dec 24, 2004 at 10:34:31AM -0600, Ray Bryant wrote:
> (Addling linux-mm, Marcello and Hirakazu....what I'm up
> to is trying to create a roll-up patch that contains just
> the memory migration code from mhp3, since I need page
> migration for some work I am doing.)
> 
> Dave,
> 
> Well, the only other big change I made (other than dropping
> P32-memsection_migrate.patch) to get it to compile and boot
> for Altix were as follows:
> 
> (I'll work on testing it after the holiday.)
> 
> P29-add-memory-migration-to-Kconfig-ia64.patch
> 
>         Add memory migration to the config menu
>         for ia64.
> 
> P30-remove-page_under_capture.patch
> 
>         removed page_under_capture() from the end
>         of shrink_cache() in mm/vmscan.c.  This
>         is not defined in the P series of patches.
>         This particular call was introduced by patch
>         P01-steal_page_from_lru.patch
> 
> (I'm not sure how to number these patches to fit in with
> your scheme, so just made some stuff up.)
> 
> (Oh yeah, this is in top of 2.6.10-rc2-mm4.)
> 
> Now it would be nice if we could figure out a way to keep
> these patchsets distinct (i. e. so work on page migration
> and hotplug can continue without me redoing this every
> week or two.)
> 
> One way to do that would be to fix it so that the page
> migration patches are first in the hotplug patch, or to
> separate the two out and assume that hotplug patch goes
> on top of the page migration patch.  How would you like
> to go about this? (I'll take a whack at moving them to
> the top of the mhp3 series file and see how much trouble
> I get into....)

I think thats up to Dave who maintains the patchset.

> PS:  It doesn't look like Marcello and Hirakazu's
> migration patch is part of your P* series.  Is
> that correct?

Yes, the plan is to merge the migration cache to the memory 
hotplug as soon as its well tested - we still lack transformation of 
migration pages into swap pages, which involves redoing all pte's
of a given migration page. 

Happy holiday!
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"aart@kvack.org"> aart@kvack.org </a>