Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Received-SPF: Pass (protection.outlook.com: domain of amd.com designates
 165.204.84.17 as permitted sender) receiver=protection.outlook.com;
 client-ip=165.204.84.17; helo=SATLEXMB04.amd.com; pr=C
Subject: [PATCH v2 10/10] Documentation/x86: Update resctrl_ui.rst for new
 features
From:   Babu Moger <babu.moger@amd.com>
To:     <fenghua.yu@intel.com>, <reinette.chatre@intel.com>,
        <tglx@linutronix.de>, <mingo@redhat.com>, <bp@alien8.de>
CC:     <eranian@google.com>, <dave.hansen@linux.intel.com>,
        <x86@kernel.org>, <hpa@zytor.com>, <corbet@lwn.net>,
        <hpa@zytor.com>, <linux-kernel@vger.kernel.org>,
        <linux-doc@vger.kernel.org>
Date:   Mon, 1 Aug 2022 15:56:43 -0500
Message-ID: <165938740360.724959.7059659465268246182.stgit@bmoger-ubuntu>
In-Reply-To: <165938717220.724959.10931629283087443782.stgit@bmoger-ubuntu>
References: <165938717220.724959.10931629283087443782.stgit@bmoger-ubuntu>
User-Agent: StGit/1.1.dev103+g5369f4c
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 01 Aug 2022 20:56:45.7937
 (UTC)
X-MS-Exchange-CrossTenant-Network-Message-Id: 0e352887-c6c6-4210-ccf8-08da7400586a
X-MS-Exchange-CrossTenant-Id: 3dd8961f-e488-4e60-8e11-a82d994e183d
X-MS-Exchange-CrossTenant-OriginalAttributedTenantConnectingIp: TenantId=3dd8961f-e488-4e60-8e11-a82d994e183d;Ip=[165.204.84.17];Helo=[SATLEXMB04.amd.com]
X-MS-Exchange-CrossTenant-AuthSource: BN8NAM11FT059.eop-nam11.prod.protection.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Anonymous
X-MS-Exchange-CrossTenant-FromEntityHeader: HybridOnPrem
X-MS-Exchange-Transport-CrossTenantHeadersStamped: BN9PR12MB5161
X-Spam-Status: No, score=-2.1 required=5.0 tests=BAYES_00,DKIM_SIGNED,
        DKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF,RCVD_IN_DNSWL_NONE,
        RCVD_IN_MSPIKE_H2,SPF_HELO_PASS,SPF_PASS autolearn=ham
        autolearn_force=no version=3.4.6
X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on
        lindbergh.monkeyblade.net
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

Update the documentation for the new features:=0A=
1. Slow Memory Bandwidth allocation.=0A=
   With this feature, the QOS  enforcement policies can be applied=0A=
   to the external slow memory connected to the host. QOS enforcement=0A=
   is accomplished by assigning a Class Of Service (COS) to a processor=0A=
   and specifying allocations or limits for that COS for each resource=0A=
   to be allocated.=0A=
=0A=
2. Bandwidth Monitoring Event Configuration (BMEC).=0A=
   The bandwidth monitoring events mbm_total_bytes and mbm_local_bytes=0A=
   are set to count all the total and local reads/writes respectively.=0A=
   With the introduction of slow memory, the two counters are not=0A=
   enough to count all the different types are memory events. With the=0A=
   feature BMEC, the users have the option to configure mbm_total_bytes=0A=
   and mbm_local_bytes to count the specific type of events.=0A=
=0A=
Added the instructions to configure with examples.=0A=
=0A=
Signed-off-by: Babu Moger <babu.moger@amd.com>=0A=
---=0A=
 Documentation/x86/resctrl.rst |  123 +++++++++++++++++++++++++++++++++++++=
++++=0A=
 1 file changed, 123 insertions(+)=0A=
=0A=
diff --git a/Documentation/x86/resctrl.rst b/Documentation/x86/resctrl.rst=
=0A=
index 71a531061e4e..ba6833171c0a 100644=0A=
--- a/Documentation/x86/resctrl.rst=0A=
+++ b/Documentation/x86/resctrl.rst=0A=
@@ -167,6 +167,12 @@ with the following files:=0A=
 		bytes) at which a previously used LLC_occupancy=0A=
 		counter can be considered for re-use.=0A=
 =0A=
+"mon_configurable":=0A=
+                Provides the information if the events mbm_total and=0A=
+                mbm_local are configurable. See the configuration=0A=
+                details for "mbm_total_config" and "mbm_local_config"=0A=
+                for more information.=0A=
+=0A=
 Finally, in the top level of the "info" directory there is a file=0A=
 named "last_cmd_status". This is reset with every "command" issued=0A=
 via the file system (making new directories or writing to any of the=0A=
@@ -264,6 +270,29 @@ When monitoring is enabled all MON groups will also co=
ntain:=0A=
 	the sum for all tasks in the CTRL_MON group and all tasks in=0A=
 	MON groups. Please see example section for more details on usage.=0A=
 =0A=
+"mbm_total_config":=0A=
+"mbm_local_config":=0A=
+        This contains the current event configuration for the events=0A=
+        mbm_total_bytes and mbm_local_bytes, respectively, when the=0A=
+        Bandwidth Monitoring Event Configuration (BMEC) feature is support=
ed.=0A=
+        These files are organized by L3 domains under the subdirectories=
=0A=
+        "mon_L3_00" and "mon_L3_01". When BMEC is supported, the events=0A=
+        mbm_local_bytes and mbm_total_bytes are configurable.=0A=
+=0A=
+        Following are the types of events supported.=0A=
+        Bits    Description=0A=
+        6       Dirty Victims from the QOS domain to all types of memory=
=0A=
+        5       Reads to slow memory in the non-local NUMA domain=0A=
+        4       Reads to slow memory in the local NUMA domain=0A=
+        3       Non-temporal writes to non-local NUMA domain=0A=
+        2       Non-temporal writes to local NUMA domain=0A=
+        1       Reads to memory in the non-local NUMA domain=0A=
+        0       Reads to memory in the local NUMA domain=0A=
+=0A=
+        By default, the mbm_total_bytes configuration is set to 0x7f to co=
unt=0A=
+        all the event types and the mbm_local_bytes configuration is set t=
o=0A=
+        0x15 to count all the local memory events.=0A=
+=0A=
 Resource allocation rules=0A=
 -------------------------=0A=
 =0A=
@@ -464,6 +493,14 @@ Memory bandwidth domain is L3 cache.=0A=
 =0A=
 	MB:<cache_id0>=3Dbw_MBps0;<cache_id1>=3Dbw_MBps1;...=0A=
 =0A=
+Slow Memory bandwidth Allocation (when supported)=0A=
+------------------------------------------=0A=
+=0A=
+Slow Memory b/w domain is L3 cache.=0A=
+::=0A=
+=0A=
+	SB:<cache_id0>=3Dbandwidth0;<cache_id1>=3Dbandwidth1;...=0A=
+=0A=
 Reading/writing the schemata file=0A=
 ---------------------------------=0A=
 Reading the schemata file will show the state of all resources=0A=
@@ -479,6 +516,44 @@ which you wish to change.  E.g.=0A=
   L3DATA:0=3Dfffff;1=3Dfffff;2=3D3c0;3=3Dfffff=0A=
   L3CODE:0=3Dfffff;1=3Dfffff;2=3Dfffff;3=3Dfffff=0A=
 =0A=
+Reading/writing the schemata file (on AMD systems)=0A=
+---------------------------------------------------------------=0A=
+Reading the schemata file will show the state of all resources=0A=
+on all domains. When writing the memory bandwidth allocation you=0A=
+only need to specify those values in an absolute number expressed=0A=
+in 1/8 GB/s increments. To allocate bandwidth limit of 2GB, you=0A=
+need to specify the value 16 (16 * 1/8 =3D 2).  E.g.=0A=
+::=0A=
+=0A=
+  # cat schemata=0A=
+    MB:0=3D2048;1=3D2048;2=3D2048;3=3D2048=0A=
+    L3:0=3Dffff;1=3Dffff;2=3Dffff;3=3Dffff=0A=
+=0A=
+  # echo "MB:1=3D16" > schemata=0A=
+  # cat schemata=0A=
+    MB:0=3D2048;1=3D  16;2=3D2048;3=3D2048=0A=
+    L3:0=3Dffff;1=3Dffff;2=3Dffff;3=3Dffff=0A=
+=0A=
+Reading/writing the schemata file (on AMD systems) with slow memory=0A=
+---------------------------------------------------------------=0A=
+Reading the schemata file will show the state of all resources=0A=
+on all domains. When writing the memory bandwidth allocation you=0A=
+only need to specify those values in an absolute number expressed=0A=
+in 1/8 GB/s increments. To allocate bandwidth limit of 8GB, you=0A=
+need to specify the value 64 (64 * 1/8 =3D 8).  E.g.=0A=
+::=0A=
+=0A=
+  # cat schemata=0A=
+    SB:0=3D2048;1=3D2048;2=3D2048;3=3D2048=0A=
+    MB:0=3D2048;1=3D2048;2=3D2048;3=3D2048=0A=
+    L3:0=3Dffff;1=3Dffff;2=3Dffff;3=3Dffff=0A=
+=0A=
+  # echo "SB:1=3D64" > schemata=0A=
+  # cat schemata=0A=
+    SB:0=3D2048;1=3D  64;2=3D2048;3=3D2048=0A=
+    MB:0=3D2048;1=3D2048;2=3D2048;3=3D2048=0A=
+    L3:0=3Dffff;1=3Dffff;2=3Dffff;3=3Dffff=0A=
+=0A=
 Cache Pseudo-Locking=0A=
 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=0A=
 CAT enables a user to specify the amount of cache space that an=0A=
@@ -1210,6 +1285,54 @@ View the llc occupancy snapshot::=0A=
   # cat /sys/fs/resctrl/p1/mon_data/mon_L3_00/llc_occupancy=0A=
   11234000=0A=
 =0A=
+Example 5 (Configure and Monitor specific event types)=0A=
+-------------------------------------------------=0A=
+=0A=
+A single socket system which has real time tasks running on cores 0-4=0A=
+and non real time tasks on other CPUs. We want to monitor the memory=0A=
+bandwidth allocation for specific events.=0A=
+::=0A=
+=0A=
+  # mount -t resctrl resctrl /sys/fs/resctrl=0A=
+  # cd /sys/fs/resctrl=0A=
+  # mkdir p1=0A=
+=0A=
+Move the CPUs 0-4 over to p1::=0A=
+=0A=
+  # echo 0xf > p1/cpus=0A=
+=0A=
+View the current mbm_local_bytes::=0A=
+=0A=
+  # cat /sys/fs/resctrl/p1/mon_data/mon_L3_00/mbm_local_bytes=0A=
+  112501=0A=
+=0A=
+Change the mbm_local_bytes to count mon-temporal writes to both local=0A=
+and non-local NUMA domain. Refer to event supported bitmap under=0A=
+mbm_local_config::=0A=
+=0A=
+  # echo 0xc > /sys/fs/resctrl/p1/mon_data/mon_L3_00/mbm_local_config=0A=
+=0A=
+View the updated mbm_local_bytes::=0A=
+=0A=
+  # cat /sys/fs/resctrl/p1/mon_data/mon_L3_00/mbm_local_bytes=0A=
+  12601=0A=
+=0A=
+Similar experiment on mbm_total_bytes. First view the current mbm_total_by=
tes::=0A=
+=0A=
+  # cat /sys/fs/resctrl/p1/mon_data/mon_L3_00/mbm_total_bytes=0A=
+  1532501=0A=
+=0A=
+Change the mbm_total_bytes to count only reads to slow memory on both loca=
l=0A=
+and non-local NUMA domain. Refer to event supported bitmap under=0A=
+mbm_total_config::=0A=
+=0A=
+  # echo 0x30 > /sys/fs/resctrl/p1/mon_data/mon_L3_00/mbm_total_config=0A=
+=0A=
+View the updated mbm_total_bytes::=0A=
+=0A=
+  # cat /sys/fs/resctrl/p1/mon_data/mon_L3_00/mbm_total_bytes=0A=
+  104562=0A=
+=0A=
 Intel RDT Errata=0A=
 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=0A=
 =0A=
=0A=