OCPBUGS-61317: account for ppc64le used with TuneD/NTO

assets/performanceprofile/tuned/openshift-node-performance

@@ -11,8 +11,14 @@ summary=Openshift node optimized for deterministic performance at the cost of in
 # This has two possible results:
 #     openshift-node,cpu-partitioning
 #     openshift-node,cpu-partitioning,openshift-node-performance-rt-<PerformanceProfile name>
+#> arm64,amd64 doesn't support realtime
+{{if not .ppc64le}}
 include=openshift-node,cpu-partitioning${f:regex_search_ternary:${f:exec:uname:-r}:rt:,openshift-node-performance-rt-{{.PerformanceProfileName}}:};
     openshift-node-performance-${f:lscpu_check:Vendor ID\:\s*GenuineIntel:intel:Vendor ID\:\s*AuthenticAMD:amd:Architecture\:\s*aarch64:arm}-${f:lscpu_check:Architecture\:\s*x86_64:x86:Architecture\:\s*aarch64:aarch64}-{{.PerformanceProfileName}}
+{{else}}
+include=openshift-node,cpu-partitioning${f:regex_search_ternary:${f:exec:uname:-r}:rt:,
+    openshift-node-performance-ppc64le-{{.PerformanceProfileName}}
+{{end}}
 
 # Inheritance of base profiles legend:
 # cpu-partitioning -> network-latency -> latency-performance

assets/performanceprofile/tuned/openshift-node-performance-ppc64le

@@ -0,0 +1,165 @@
+[main]
+summary=Openshift node optimized for deterministic performance at the cost of increased power consumption, focused on low latency network performance. Based on Tuned 2.11 and Cluster node tuning (oc 4.5)
+
+# The final result of the include depends on cpu vendor, cpu architecture, and whether the real time kernel is enabled
+# The first line will be evaluated based on the CPU vendor and architecture
+# This has three possible results:
+#   include=openshift-node-performance-amd-x86;
+#   include=openshift-node-performance-arm-aarch64;
+#   include=openshift-node-performance-intel-x86;
+# The second line will be evaluated based on whether the real time kernel is enabled
+# This has one result for power
+#     openshift-node,cpu-partitioning
+include=openshift-node,cpu-partitioning${f:regex_search_ternary:${f:exec:uname:-r}:rt:,openshift-node-performance-rt-{{.PerformanceProfileName}}:};
+
+# Inheritance of base profiles legend:
+# cpu-partitioning -> network-latency -> latency-performance
+# https://github.com/redhat-performance/tuned/blob/master/profiles/latency-performance/tuned.conf
+# https://github.com/redhat-performance/tuned/blob/master/profiles/network-latency/tuned.conf
+# https://github.com/redhat-performance/tuned/blob/master/profiles/cpu-partitioning/tuned.conf
+
+# All values are mapped with a comment where a parent profile contains them.
+# Different values will override the original values in parent profiles.
+
+[variables]
+#> isolated_cores take a list of ranges; e.g. isolated_cores=2,4-7
+{{if .IsolatedCpus}}
+isolated_cores={{.IsolatedCpus}}
+{{end}}
+
+not_isolated_cores_expanded=${f:cpulist_invert:${isolated_cores_expanded}}
+
+{{if .PerPodPowerManagement}}
+[cpu]
+enabled=false
+{{else}}
+[cpu]
+#> latency-performance
+#> (override)
+force_latency=cstate.id:1|3
+governor=performance
+energy_perf_bias=performance
+min_perf_pct=100
+{{end}}
+
+[vm]
+#> network-latency
+transparent_hugepages=never
+
+{{if not .GloballyDisableIrqLoadBalancing}}
+[irqbalance]
+# Disable the plugin entirely, which was enabled by the parent profile `cpu-partitioning`.
+# It can be racy if TuneD restarts for whatever reason.
+#> cpu-partitioning
+enabled=false
+{{end}}
+
+[scheduler]
+runtime=0
+group.ksoftirqd=0:f:11:*:ksoftirqd.*
+group.rcuc=0:f:11:*:rcuc.*
+group.ktimers=0:f:11:*:ktimers.*
+{{if not .GloballyDisableIrqLoadBalancing}}
+default_irq_smp_affinity = ignore
+irq_process=false
+{{end}}
+
+[sysctl]
+{{if .RealTimeHint}}
+#> cpu-partitioning #RealTimeHint
+kernel.hung_task_timeout_secs=600
+#> cpu-partitioning #RealTimeHint
+kernel.nmi_watchdog=0
+#> RealTimeHint
+kernel.sched_rt_runtime_us=-1
+#> cpu-partitioning  #RealTimeHint
+vm.stat_interval=10
+{{end}}
+
+#> network-latency
+net.ipv4.tcp_fastopen=3
+
+# If a workload mostly uses anonymous memory and it hits this limit, the entire
+# working set is buffered for I/O, and any more write buffering would require
+# swapping, so it's time to throttle writes until I/O can catch up.  Workloads
+# that mostly use file mappings may be able to use even higher values.
+#
+# The generator of dirty data starts writeback at this percentage (system default
+# is 20%)
+#> latency-performance
+vm.dirty_ratio=10
+
+# Start background writeback (via writeback threads) at this percentage (system
+# default is 10%)
+#> latency-performance
+vm.dirty_background_ratio=3
+
+# The swappiness parameter controls the tendency of the kernel to move
+# processes out of physical memory and onto the swap disk.
+# 0 tells the kernel to avoid swapping processes out of physical memory
+# for as long as possible
+# 100 tells the kernel to aggressively swap processes out of physical memory
+# and move them to swap cache
+#> latency-performance
+vm.swappiness=10
+
+# also configured via a sysctl.d file
+# placed here for documentation purposes and commented out due
+# to a tuned logging bug complaining about duplicate sysctl:
+#   https://issues.redhat.com/browse/RHEL-18972
+#> rps configuration
+# net.core.rps_default_mask=${not_isolated_cpumask}
+
+
+[selinux]
+#> Custom (atomic host)
+avc_cache_threshold=8192
+
+{{if .NetDevices}}
+{{.NetDevices}}
+{{end}}
+
+[bootloader]
+# !! The names are important for Intel and are referenced in openshift-node-performance-intel-x86
+
+# set empty values to disable RHEL initrd setting in cpu-partitioning
+initrd_remove_dir=
+initrd_dst_img=
+initrd_add_dir=
+
+# overrides cpu-partitioning cmdline
+cmdline_cpu_part=+nohz=on rcu_nocbs=${isolated_cores} tuned.non_isolcpus=${not_isolated_cpumask} systemd.cpu_affinity=${not_isolated_cores_expanded}
+
+# No default value but will be composed conditionally based on platform
+cmdline_iommu=
+
+{{if .StaticIsolation}}
+cmdline_isolation=+isolcpus=domain,managed_irq,${isolated_cores}
+{{else}}
+cmdline_isolation=+isolcpus=managed_irq,${isolated_cores}
+{{end}}
+
+{{if .HighPowerConsumption}}
+# No default value but will be composed conditionally based on platform
+cmdline_power_performance=
+{{end}}
+
+{{if .DefaultHugepagesSize}}
+cmdline_hugepages=+ default_hugepagesz={{.DefaultHugepagesSize}} {{end}} {{if .Hugepages}} {{.Hugepages}} {{end}}
+
+{{if .AdditionalArgs}}
+cmdline_additionalArg=+{{.AdditionalArgs}}
+{{end}}
+
+[rtentsk]
+
+{{ if .HardwareTuning }}
+[sysfs]
+# sets provided frequencies to isolated and reserved cpus
+{{ range .IsolatedCpuList }}
+/sys/devices/system/cpu/cpufreq/policy{{.}}/scaling_max_freq={{$.IsolatedCpuMaxFreq}}
+{{- end -}}
+{{ range .ReservedCpuList }}
+/sys/devices/system/cpu/cpufreq/policy{{.}}/scaling_max_freq={{$.ReservedCpuMaxFreq}}
+{{- end -}}
+{{ end }}

pkg/performanceprofile/controller/performanceprofile/components/consts.go

@@ -29,6 +29,8 @@ const (
 	ProfileNameArmAarch64 = "openshift-node-performance-arm-aarch64"
 	// ProfileNameIntelX86 defines the Intel X86 specific tuning parameters
 	ProfileNameIntelX86 = "openshift-node-performance-intel-x86"
+	// ProfileNamePowerPC defines the PowerPC LE specific tuning parameters
+	ProfileNamePowerPC = "openshift-node-performance-ppc64le"
 )
 
 const (

pkg/performanceprofile/controller/performanceprofile/components/tuned/tuned.go

@@ -240,6 +240,12 @@ func NewNodePerformance(profile *performancev2.PerformanceProfile) (*tunedv1.Tun
 	}
 	IntelX86ProfileName := components.GetComponentName(profile.Name, components.ProfileNameIntelX86)
 
+	PowerPCProfileData, err := getProfileData(filepath.Join("tuned", components.ProfileNamePowerPC), templateArgs)
+	if err != nil {
+		return nil, err
+	}
+	PowerPCProfileName := components.GetComponentName(profile.Name, components.ProfileNamePowerPC)
+
 	profiles := []tunedv1.TunedProfile{
 		{
 			Name: &name,
@@ -261,6 +267,10 @@ func NewNodePerformance(profile *performancev2.PerformanceProfile) (*tunedv1.Tun
 			Name: &IntelX86ProfileName,
 			Data: &IntelX86ProfileData,
 		},
+		{
+			Name: &PowerPCProfileName,
+			Data: &PowerPCProfileData,
+		},
 	}
 
 	priority := uint64(20)