Against: 2.6.33-mmotm-100302-1838
Patch: in-kernel profiling -- support memoryless nodes.
Another example of using numa_mem_id() to support memoryless
nodes efficiently. I stumbled across this when trying to profile
the kernel in the memoryless nodes configuration. A quick look
at other usages of numa_node_id() and cpu_to_node() for explicit
local allocations indicates that there are several other places
that could be problematic for systems with memoryless nodes that
can also be addressed with this simple substitution:
In-kernel profiling requires that we be able to allocate "local"
memory for each cpu. Use "cpu_to_mem()" instead of "cpu_to_node()"
to support memoryless nodes.
Depends on the "numa_mem_id()" patch.
Signed-off-by: Lee Schermerhorn <lee.schermerhorn@xxxxxx>
kernel/profile.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
Index: linux-2.6.33-mmotm-100302-1838/kernel/profile.c
===================================================================
--- linux-2.6.33-mmotm-100302-1838.orig/kernel/profile.c
+++ linux-2.6.33-mmotm-100302-1838/kernel/profile.c
@@ -363,7 +363,7 @@ static int __cpuinit profile_cpu_callbac
switch (action) {
case CPU_UP_PREPARE:
case CPU_UP_PREPARE_FROZEN:
- node = cpu_to_node(cpu);
+ node = cpu_to_mem(cpu);
per_cpu(cpu_profile_flip, cpu) = 0;
if (!per_cpu(cpu_profile_hits, cpu)[1]) {
page = alloc_pages_exact_node(node,
@@ -565,7 +565,7 @@ static int create_hash_tables(void)
int cpu;
for_each_online_cpu(cpu) {
- int node = cpu_to_node(cpu);
+ int node = cpu_to_mem(cpu);
struct page *page;
page = alloc_pages_exact_node(node,
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxxx For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>