On Sun, Oct 23, 2011 at 09:05:05PM -0700, Michal Nazarewicz wrote:

quoted

On Thu, Oct 06, 2011 at 03:54:42PM +0200, Marek Szyprowski wrote:

quoted

This commit introduces alloc_contig_freed_pages() function
which allocates (ie. removes from buddy system) free pages
in range. Caller has to guarantee that all pages in range
are in buddy system.

On Tue, 18 Oct 2011 05:21:09 -0700, Mel Gorman [off-list ref] wrote:

quoted

Straight away, I'm wondering why you didn't use
mm/compaction.c#isolate_freepages()

Does the below look like a step in the right direction?

It basically moves isolate_freepages_block() to page_alloc.c (changing

For the purposes of review, have a separate patch for moving
isolate_freepages_block to another file that does not alter the
function in any way. When the function is updated in a follow-on patch,
it'll be far easier to see what has changed.

page_isolation.c may also be a better fit than page_alloc.c

As it is, the patch for isolate_freepages_block is almost impossible
to read because it's getting munged with existing code that is already
in page_alloc.c . About all I caught from it is that scannedp does
not have a type. It defaults to unsigned int but it's unnecessarily
obscure.

it name to isolate_freepages_range()) and changes it so that depending
on arguments it treats holes (either invalid PFN or non-free page) as
errors so that CMA can use it.

I haven't actually read the function because it's too badly mixed with
page_alloc.c code but this description fits what I'm looking for.

It also accepts a range rather then just assuming a single pageblock
thus the change moves range calculation in compaction.c from
isolate_freepages_block() up to isolate_freepages().

The change also modifies spilt_free_page() so that it does not try to
change pageblock's migrate type if current migrate type is ISOLATE or
CMA.

This is fine. Later, the flags that determine what happens to pageblocks
may be placed in compact_control.

---
 include/linux/mm.h             |    1 -
 include/linux/page-isolation.h |    4 +-
 mm/compaction.c                |   73 +++--------------------
 mm/internal.h                  |    5 ++
 mm/page_alloc.c                |  128 +++++++++++++++++++++++++---------------
 5 files changed, 95 insertions(+), 116 deletions(-)

I confess I didn't read closely because of the mess in page_alloc.c but
the intent seems fine. Hopefully there will be a new version of CMA
posted that will be easier to review.

Thanks

-- 
Mel Gorman
SUSE Labs