import tuned-2.22.1-4.el8_10

.gitignore

@@ -1 +1 @@
-SOURCES/tuned-2.21.0.tar.gz
+SOURCES/tuned-2.22.1.tar.gz

.tuned.metadata

@@ -1 +1 @@
-586db5263ea7dabcc57e24aceb249d397d7d0239 SOURCES/tuned-2.21.0.tar.gz
+60e206fe73ea537e64141f92b331f65879766f97 SOURCES/tuned-2.22.1.tar.gz

SOURCES/tuned-2.22.0-rhel-8-profiles.patch

@@ -1,13 +1,14 @@
 diff --git a/profiles/latency-performance/tuned.conf b/profiles/latency-performance/tuned.conf
-index d200b5c..877229f 100644
+index 1dec690..e592138 100644
 --- a/profiles/latency-performance/tuned.conf
 +++ b/profiles/latency-performance/tuned.conf
-@@ -32,3 +32,16 @@ vm.dirty_background_ratio=3
+@@ -35,3 +35,17 @@ vm.dirty_background_ratio=3
  # 100 tells the kernel to aggressively swap processes out of physical memory
  # and move them to swap cache
  vm.swappiness=10
 +
 +[scheduler]
++runtime=0
 +# ktune sysctl settings for rhel6 servers, maximizing i/o throughput
 +#
 +# Minimal preemption granularity for CPU-bound tasks:
@@ -20,26 +21,28 @@ index d200b5c..877229f 100644
 +# (system default is 500000, i.e. 0.5 ms)
 +sched_migration_cost_ns = 5000000
 diff --git a/profiles/sap-hana/tuned.conf b/profiles/sap-hana/tuned.conf
-index aeecf53..8dcee57 100644
+index aeecf53..1b15ea3 100644
 --- a/profiles/sap-hana/tuned.conf
 +++ b/profiles/sap-hana/tuned.conf
-@@ -20,3 +20,7 @@ kernel.numa_balancing = 0
+@@ -20,3 +20,8 @@ kernel.numa_balancing = 0
  vm.dirty_ratio = 40
  vm.dirty_background_ratio = 10
  vm.swappiness = 10
 +
 +[scheduler]
++runtime=0
 +sched_min_granularity_ns = 3000000
 +sched_wakeup_granularity_ns = 4000000
 diff --git a/profiles/throughput-performance/tuned.conf b/profiles/throughput-performance/tuned.conf
-index 62a7322..7c55fb2 100644
+index e4e832f..3d9c42f 100644
 --- a/profiles/throughput-performance/tuned.conf
 +++ b/profiles/throughput-performance/tuned.conf
-@@ -63,9 +63,31 @@ vm.swappiness=10
+@@ -67,9 +67,33 @@ vm.swappiness=10
  # on older kernels
  net.core.somaxconn=>2048
  
 +[scheduler]
++runtime=0
 +# ktune sysctl settings for rhel6 servers, maximizing i/o throughput
 +#
 +# Minimal preemption granularity for CPU-bound tasks:
@@ -66,17 +69,19 @@ index 62a7322..7c55fb2 100644
 +type=scheduler
 +uname_regex=x86_64
 +cpuinfo_regex=${amd_cpuinfo_regex}
++runtime=0
 +sched_migration_cost_ns=5000000
 diff --git a/profiles/virtual-host/tuned.conf b/profiles/virtual-host/tuned.conf
-index 5301d9f..74a5fb0 100644
+index 5301d9f..24d0fb4 100644
 --- a/profiles/virtual-host/tuned.conf
 +++ b/profiles/virtual-host/tuned.conf
-@@ -14,3 +14,9 @@ vm.dirty_background_ratio = 5
+@@ -14,3 +14,10 @@ vm.dirty_background_ratio = 5
  [cpu]
  # Setting C3 state sleep mode/power savings
  force_latency=cstate.id_no_zero:3|70
 +
 +[scheduler]
++runtime=0
 +# The total time the scheduler will consider a migrated process
 +# "cache hot" and thus less likely to be re-migrated
 +# (system default is 500000, i.e. 0.5 ms)

SOURCES/tuned-2.22.1-profile-epyc-eda.patch

@@ -0,0 +1,54 @@
+From 7557cf975282326cdbfe55b7b803d8075ff37cba Mon Sep 17 00:00:00 2001
+From: =?UTF-8?q?Jaroslav=20=C5=A0karvada?= <jskarvad@redhat.com>
+Date: Tue, 12 Mar 2024 20:25:43 +0100
+Subject: [PATCH] epyc-eda: added new profile for EDA compute workloads on AMD
+ EPYC CPUs
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+Signed-off-by: Jaroslav Škarvada <jskarvad@redhat.com>
+---
+ man/tuned-profiles.7         |  4 ++++
+ profiles/epyc-eda/tuned.conf | 14 ++++++++++++++
+ 2 files changed, 18 insertions(+)
+ create mode 100644 profiles/epyc-eda/tuned.conf
+
+diff --git a/man/tuned-profiles.7 b/man/tuned-profiles.7
+index 10cad7b..600e8bb 100644
+--- a/man/tuned-profiles.7
++++ b/man/tuned-profiles.7
+@@ -141,6 +141,10 @@ profiles (e.g. throughput\-performance profile), example:
+ Profile optimized for AWS EC2 instances. It is based on the
+ throughput\-performance profile.
+ 
++.TP
++.BI "epyc-eda"
++Profile optimized for EDA compute workloads on AMD EPYC CPUs.
++
+ .SH "FILES"
+ .nf
+ .I /etc/tuned/*
+diff --git a/profiles/epyc-eda/tuned.conf b/profiles/epyc-eda/tuned.conf
+new file mode 100644
+index 0000000..482d404
+--- /dev/null
++++ b/profiles/epyc-eda/tuned.conf
+@@ -0,0 +1,14 @@
++#
++# tuned configuration
++#
++
++[main]
++summary=Optimize for EDA compute workloads on AMD EPYC CPUs
++description=Configures virtual memory, CPU governors, and network settings for EDA compute workloads.
++include=throughput-performance
++
++# AMD
++[scheduler.amd]
++type=scheduler
++#Allow processes to rapidly move between cores to avoid idle time and maximize CPU usage
++sched_migration_cost_ns=10000
+-- 
+2.44.0
+

SPECS/tuned.spec

@@ -34,8 +34,8 @@
 
 Summary: A dynamic adaptive system tuning daemon
 Name: tuned
-Version: 2.21.0
+Version: 2.22.1
-Release: 1%{?prerel1}%{?dist}
+Release: 4%{?prerel1}%{?dist}
 License: GPLv2+
 Source0: https://github.com/redhat-performance/%{name}/archive/v%{version}%{?prerel2}/%{name}-%{version}%{?prerel2}.tar.gz
 # RHEL-8 specific recommend.conf:
@@ -94,9 +94,11 @@ Requires: python3-syspurpose
 %endif
 %endif
 # Revert upstream profiles changes which have not been approved for RHEL-8 (yet)
-Patch0: tuned-2.20.0-rhel-8-profiles.patch
+Patch0: tuned-2.22.0-rhel-8-profiles.patch
 # Revert no balancing cores to use SD_LOAD_BALANCE (see rhbz#1874596 for details)
 Patch1: tuned-2.21.0-sd-load-balance.patch
+# epyc-eda TuneD profile only for RHEL-8 (see RHEL-27528 for details)
+Patch2: tuned-2.22.1-profile-epyc-eda.patch
 
 %description
 The tuned package contains a daemon that tunes system settings dynamically.
@@ -255,6 +257,17 @@ Requires: %{name} = %{version}
 %description profiles-openshift
 Additional TuneD profile(s) optimized for OpenShift.
 
+%package ppd
+Summary: PPD compatibility daemon
+Requires: %{name} = %{version}
+# The compatibility daemon is swappable for power-profiles-daemon
+Provides: ppd-service
+Conflicts: ppd-service
+
+%description ppd
+An API translation daemon that allows applications to easily transition
+to TuneD from power-profiles-daemon (PPD).
+
 %prep
 %autosetup -p1 -n %{name}-%{version}%{?prerel2}
 
@@ -271,9 +284,7 @@ make html %{make_python_arg}
 
 %install
 make install DESTDIR=%{buildroot} DOCDIR=%{docdir} %{make_python_arg}
-%if 0%{?rhel}
+make install-ppd DESTDIR=%{buildroot} DOCDIR=%{docdir} %{make_python_arg}
-sed -i 's/\(dynamic_tuning[ \t]*=[ \t]*\).*/\10/' %{buildroot}%{_sysconfdir}/tuned/tuned-main.conf
-%endif
 
 %if ! 0%{?rhel}
 # manual
@@ -551,7 +562,58 @@ fi
 %{_prefix}/lib/tuned/openshift-node
 %{_mandir}/man7/tuned-profiles-openshift.7*
 
+%files ppd
+%{_sbindir}/tuned-ppd
+%{_unitdir}/tuned-ppd.service
+%{_datadir}/dbus-1/system-services/net.hadess.PowerProfiles.service
+%{_datadir}/dbus-1/system.d/net.hadess.PowerProfiles.conf
+%{_datadir}/polkit-1/actions/net.hadess.PowerProfiles.policy
+%config(noreplace) %{_sysconfdir}/tuned/ppd.conf
+
 %changelog
+* Wed Mar 13 2024 Jaroslav Škarvada <jskarvad@redhat.com> - 2.22.1-4
+- release bump due to broken c8s
+  related: RHEL-27528
+
+* Wed Mar 13 2024 Jaroslav Škarvada <jskarvad@redhat.com> - 2.22.1-3
+- release bump
+  related: RHEL-27528
+
+* Tue Mar 12 2024 Jaroslav Škarvada <jskarvad@redhat.com> - 2.22.1-2
+- profiles: added epyc-eda profile
+  resolves: RHEL-27528
+
+* Thu Feb 22 2024 Jaroslav Škarvada <jskarvad@redhat.com> - 2.22.1-1
+- new release
+  - rebased tuned to latest upstream
+    related: RHEL-17119
+  - renamed intel_uncore plugin to uncore
+  - network-throughput: increased net.ipv4.tcp_rmem default value
+
+* Fri Feb 16 2024 Jaroslav Škarvada <jskarvad@redhat.com> - 2.22.0-1
+- new release
+  - rebased tuned to latest upstream
+    related: RHEL-17119
+  - lower CPU usage by using runtime=0 in profiles using scheduler plugin
+    resolves: RHEL-6869
+
+* Fri Feb  9 2024 Jaroslav Škarvada <jskarvad@redhat.com> - 2.22.0-0.1.rc1
+- new release
+  - rebased tuned to latest upstream
+    resolves: RHEL-17119
+  - print all arguments of failing commands in error messages
+  - plugin_sysctl: added support for sysctl names with slash
+  - tuned-adm: added support for moving devices between plugin instances
+  - api: added methods for retrieval of plugin instances and devices
+  - plugin_cpu: amd-pstate mentioned instead of just intel_pstate
+  - hotplug: do not report ENOENT errors on device remove
+  - plugin_sysctl: expand variables when reporting overrides
+  - plugin_acpi: new plugin which handles ACPI platform_profile
+  - plugin_bootloader: skip calling rpm-ostree kargs in no-op case
+  - plugin_cpu: support cstate settings of pm_qos_resume_latency_us
+  - scheduler: add option for ignoring IRQs affinity
+  - plugin_intel_uncore: new plugin for uncore setting
+
 * Tue Aug 29 2023 Jaroslav Škarvada <jskarvad@redhat.com> - 2.21.0-1
 - new release
   - api: fixed stop method not to require any parameter