[Skiboot] [PATCH] SLW : Increase stop4-5 residency by 5x

[Abhishek Goel]

Using ppdcf2d run, we observed that there was a drop of about 50%
throughput with stop4/5. In this benchmark the GPU waits on the cpu to
wakeup and provide the subsequent data block to compute. CPUs serving
GPUs workloads are idle enough for the current residency values for the
governor to pick up stop4/5 states for those CPUs. The higher wakeup
latency of these deep states accumulates over the run and shows up as a
performance drop.

We observe no perfromance drop with increased residency values.

Refer commit:87f33f4990612116306ab42fbd7c163a2f90c89c for previous
context.

Signed-off-by: Abhishek Goel <huntbag at linux.vnet.ibm.com>
Tested-by: Parth Shah <parth at linux.ibm.com>
---
 hw/slw.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/hw/slw.c b/hw/slw.c
index 566a1128..1a50be71 100644
--- a/hw/slw.c
+++ b/hw/slw.c
@@ -516,7 +516,7 @@ static struct cpu_idle_states power9_cpu_idle_states[] = {
 	{
 		.name = "stop4",
 		.latency_ns = 100000,
-		.residency_ns = 10000000,
+		.residency_ns = 50000000,
 		.flags = 0*OPAL_PM_DEC_STOP \
 		       | 0*OPAL_PM_TIMEBASE_STOP  \
 		       | 1*OPAL_PM_LOSE_USER_CONTEXT \
@@ -532,7 +532,7 @@ static struct cpu_idle_states power9_cpu_idle_states[] = {
 	{
 		.name = "stop5",
 		.latency_ns = 200000,
-		.residency_ns = 20000000,
+		.residency_ns = 100000000,
 		.flags = 0*OPAL_PM_DEC_STOP \
 		       | 0*OPAL_PM_TIMEBASE_STOP  \
 		       | 1*OPAL_PM_LOSE_USER_CONTEXT \
-- 
2.17.1