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 v4 03/13] x86/cpufeatures: Add Slow Memory Bandwidth
 Allocation feature flag
From:   Babu Moger <babu.moger@amd.com>
To:     <corbet@lwn.net>, <reinette.chatre@intel.com>,
        <tglx@linutronix.de>, <mingo@redhat.com>, <bp@alien8.de>
CC:     <fenghua.yu@intel.com>, <dave.hansen@linux.intel.com>,
        <x86@kernel.org>, <hpa@zytor.com>, <paulmck@kernel.org>,
        <akpm@linux-foundation.org>, <quic_neeraju@quicinc.com>,
        <rdunlap@infradead.org>, <damien.lemoal@opensource.wdc.com>,
        <songmuchun@bytedance.com>, <peterz@infradead.org>,
        <jpoimboe@kernel.org>, <pbonzini@redhat.com>, <babu.moger@amd.com>,
        <chang.seok.bae@intel.com>, <pawan.kumar.gupta@linux.intel.com>,
        <jmattson@google.com>, <daniel.sneddon@linux.intel.com>,
        <sandipan.das@amd.com>, <tony.luck@intel.com>,
        <james.morse@arm.com>, <linux-doc@vger.kernel.org>,
        <linux-kernel@vger.kernel.org>, <bagasdotme@gmail.com>,
        <eranian@google.com>
Date:   Wed, 7 Sep 2022 13:00:15 -0500
Message-ID: <166257361558.1043018.12046795385220064687.stgit@bmoger-ubuntu>
In-Reply-To: <166257348081.1043018.11227924488792315932.stgit@bmoger-ubuntu>
References: <166257348081.1043018.11227924488792315932.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: 07 Sep 2022 18:00:23.5539
 (UTC)
X-MS-Exchange-CrossTenant-Network-Message-Id: 07ccbdb9-83a1-44b0-e734-08da90fad631
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: BN8NAM11FT074.eop-nam11.prod.protection.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Anonymous
X-MS-Exchange-CrossTenant-FromEntityHeader: HybridOnPrem
X-MS-Exchange-Transport-CrossTenantHeadersStamped: IA1PR12MB6308
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,T_SCC_BODY_TEXT_LINE
        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

Adds the new AMD feature X86_FEATURE_SMBA. With this feature, the QOS=0A=
enforcement policies can be applied to external slow memory connected=0A=
to the host. QOS enforcement is accomplished by assigning a Class Of=0A=
Service (COS) to a processor and specifying allocations or limits for=0A=
that COS for each resource to be allocated.=0A=
=0A=
This feature is identified by the CPUID Function 8000_0020_EBX_x0.=0A=
=0A=
CPUID Fn8000_0020_EBX_x0 AMD Bandwidth Enforcement Feature Identifiers (ECX=
=3D0)=0A=
Bits    Field Name      Description=0A=
2       L3SBE           L3 external slow memory bandwidth enforcement=0A=
=0A=
Currently, CXL.memory is the only supported "slow" memory device. With the=
=0A=
support of SMBA feature the hardware enables bandwidth allocation on the=0A=
slow memory devices. If there are multiple slow memory devices in the syste=
m,=0A=
then the throttling logic groups all the slow sources together and applies=
=0A=
the limit on them as a whole.=0A=
=0A=
The presence of the SMBA feature(with CXL.memory) is independent of whether=
=0A=
slow memory device is actually present in the system. If there is no slow=
=0A=
memory in the system, then setting a SMBA limit will have no impact on the=
=0A=
performance of the system.=0A=
=0A=
Presence of CXL memory can be identified by numactl command.=0A=
=0A=
$numactl -H=0A=
available: 2 nodes (0-1)=0A=
node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16=0A=
node 0 size: 63678 MB node 0 free: 59542 MB=0A=
node 1 cpus:=0A=
node 1 size: 16122 MB=0A=
node 1 free: 15627 MB=0A=
node distances:=0A=
node   0   1=0A=
   0:  10  50=0A=
   1:  50  10=0A=
=0A=
CPU list for CXL memory will be emply. The cpu-cxl node distance is greater=
=0A=
than cpu-to-cpu distances. Node 1 has the CXL memory in this case. CXL=0A=
memory can also be identified using ACPI SRAT table and memory maps.=0A=
=0A=
Feature description is available in the specification, "AMD64 Technology Pl=
atform Quality=0A=
of Service Extensions, Revision: 1.03 Publication # 56375 Revision: 1.03 Is=
sue Date: February 2022".=0A=
=0A=
Link: https://www.amd.com/en/support/tech-docs/amd64-technology-platform-qu=
ality-service-extensions=0A=
Link: https://bugzilla.kernel.org/show_bug.cgi?id=3D206537=0A=
Signed-off-by: Babu Moger <babu.moger@amd.com>=0A=
---=0A=
 arch/x86/include/asm/cpufeatures.h |    1 +=0A=
 arch/x86/kernel/cpu/scattered.c    |    1 +=0A=
 2 files changed, 2 insertions(+)=0A=
=0A=
diff --git a/arch/x86/include/asm/cpufeatures.h b/arch/x86/include/asm/cpuf=
eatures.h=0A=
index 235dc85c91c3..1815435c9c88 100644=0A=
--- a/arch/x86/include/asm/cpufeatures.h=0A=
+++ b/arch/x86/include/asm/cpufeatures.h=0A=
@@ -304,6 +304,7 @@=0A=
 #define X86_FEATURE_UNRET		(11*32+15) /* "" AMD BTB untrain return */=0A=
 #define X86_FEATURE_USE_IBPB_FW		(11*32+16) /* "" Use IBPB during runtime =
firmware calls */=0A=
 #define X86_FEATURE_RSB_VMEXIT_LITE	(11*32+17) /* "" Fill RSB on VM exit w=
hen EIBRS is enabled */=0A=
+#define X86_FEATURE_SMBA		(11*32+18) /* SLOW Memory Bandwidth Allocation *=
/=0A=
 =0A=
 /* Intel-defined CPU features, CPUID level 0x00000007:1 (EAX), word 12 */=
=0A=
 #define X86_FEATURE_AVX_VNNI		(12*32+ 4) /* AVX VNNI instructions */=0A=
diff --git a/arch/x86/kernel/cpu/scattered.c b/arch/x86/kernel/cpu/scattere=
d.c=0A=
index fd44b54c90d5..885ecf46abb2 100644=0A=
--- a/arch/x86/kernel/cpu/scattered.c=0A=
+++ b/arch/x86/kernel/cpu/scattered.c=0A=
@@ -44,6 +44,7 @@ static const struct cpuid_bit cpuid_bits[] =3D {=0A=
 	{ X86_FEATURE_CPB,		CPUID_EDX,  9, 0x80000007, 0 },=0A=
 	{ X86_FEATURE_PROC_FEEDBACK,    CPUID_EDX, 11, 0x80000007, 0 },=0A=
 	{ X86_FEATURE_MBA,		CPUID_EBX,  6, 0x80000008, 0 },=0A=
+	{ X86_FEATURE_SMBA,             CPUID_EBX,  2, 0x80000020, 0 },=0A=
 	{ X86_FEATURE_PERFMON_V2,	CPUID_EAX,  0, 0x80000022, 0 },=0A=
 	{ 0, 0, 0, 0, 0 }=0A=
 };=0A=
=0A=