[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH 1/4] hvm: NUMA guest: extend memops hypercall



Keir Fraser wrote:
On 4/7/08 16:28, "Andre Przywara" <andre.przywara@xxxxxxx> wrote:

Have you applied the patches correctly? From 02_numa_guest.patch:
@@ -115,7 +113,7 @@
              goto out;

          page = alloc_domheap_pages(
-            d, a->extent_order, a->memflags | MEMF_node(node));
+            d, a->extent_order, a->memflags);
          if ( unlikely(page == NULL) )
          {
              gdprintk(XENLOG_INFO, "Could not allocate order=%d extent:"

The other use of MEMF_node is in exchange_memory, which is not given any
NUMA node info from the caller, so this is correct.

When sent a patch sequence I expect the patches to apply and work
independently (when applied one-by-one in order).
IMHO that is what they do (beside the below issue patch 1 and 2 are more or less refactoring without functional changes), but anyway...

Anyway, your second patch changes the default NUMA allocation policy from
allocate on home node for the domain to allocate on node I'm currently
executing on. That would seem a net loss for PV guests (whose builder will
not be explicitly specifying the numa node for allocations).
Right you are, I have missed the subtle difference between both (the code isn't as clear as your sentence). The below patch should fix this (by catching NUMA_NO_NODE while still knowing struct domain*).
If there are no further issues, I will resend the patches.

Regards,
Andre.

--
Andre Przywara
AMD-Operating System Research Center (OSRC), Dresden, Germany
Tel: +49 351 277-84917
----to satisfy European Law for business letters:
AMD Saxony Limited Liability Company & Co. KG,
Wilschdorfer Landstr. 101, 01109 Dresden, Germany
Register Court Dresden: HRA 4896, General Partner authorized
to represent: AMD Saxony LLC (Wilmington, Delaware, US)
General Manager of AMD Saxony LLC: Dr. Hans-R. Deppe, Thomas McCoy
diff -r a3d712538f1b xen/common/page_alloc.c
--- a/xen/common/page_alloc.c   Fri Jul 04 15:57:01 2008 +0200
+++ b/xen/common/page_alloc.c   Sat Jul 05 01:09:48 2008 +0200
@@ -792,6 +792,8 @@ struct page_info *alloc_domheap_pages(
 
     ASSERT(!in_irq());
 
+    if ( node == NUMA_NO_NODE ) node = domain_to_node (d);
+
     bits = domain_clamp_alloc_bitsize(d, bits ? : (BITS_PER_LONG+PAGE_SHIFT));
     if ( bits <= (PAGE_SHIFT + 1) )
         return NULL;
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.