summaryrefslogtreecommitdiff
path: root/arch
diff options
context:
space:
mode:
authorAnton Blanchard <anton@samba.org>2011-07-24 16:33:15 (GMT)
committerBenjamin Herrenschmidt <benh@kernel.crashing.org>2011-09-20 05:53:22 (GMT)
commit7bebcf0925f09224393a8992af706fa39aa10395 (patch)
tree49ba8f108c16447dcd22869e48232d2721cdd57e /arch
parentd4761ad2ef18ec2c9a0037d6649c0afc4a7b907d (diff)
downloadlinux-7bebcf0925f09224393a8992af706fa39aa10395.tar.xz
powerpc/numa: Disable NEWIDLE balancing at node level
On big POWER7 boxes we see large amounts of CPU time in system processes like workqueue and watchdog kernel threads. We currently rebalance the entire machine each time a task goes idle and this is very expensive on large machines. Disable newidle balancing at the node level and rely on the scheduler tick to rebalance across nodes. Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Diffstat (limited to 'arch')
-rw-r--r--arch/powerpc/include/asm/topology.h2
1 files changed, 1 insertions, 1 deletions
diff --git a/arch/powerpc/include/asm/topology.h b/arch/powerpc/include/asm/topology.h
index d1c1d31..13efa4d 100644
--- a/arch/powerpc/include/asm/topology.h
+++ b/arch/powerpc/include/asm/topology.h
@@ -75,7 +75,7 @@ static inline int pcibus_to_node(struct pci_bus *bus)
.forkexec_idx = 0, \
\
.flags = 1*SD_LOAD_BALANCE \
- | 1*SD_BALANCE_NEWIDLE \
+ | 0*SD_BALANCE_NEWIDLE \
| 1*SD_BALANCE_EXEC \
| 1*SD_BALANCE_FORK \
| 0*SD_BALANCE_WAKE \