linux-kernel - [PATCH RFC 4/8] KVM: x86/pmu: Add Intel CPUID-hinted Topdown Slots event

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Mon, 12 Dec 2022 20:58:40 +0800
From:   Like Xu <like.xu.linux@...il.com>
To:     Peter Zijlstra <peterz@...radead.org>,
        Sean Christopherson <seanjc@...gle.com>
Cc:     Paolo Bonzini <pbonzini@...hat.com>, linux-kernel@...r.kernel.org,
        kvm@...r.kernel.org
Subject: [PATCH RFC 4/8] KVM: x86/pmu: Add Intel CPUID-hinted Topdown Slots event

From: Like Xu <likexu@...cent.com>

This event counts the total number of available slots for an unhalted
logical processor. Software can use this event as the denominator for the
top-level metrics of the Top-down Microarchitecture Analysis method.

Although the MSR_PERF_METRICS MSR required for topdown events is not
currently available in the guest, relying only on the data provided by
the slots event is sufficient for pmu users to perceive differences in
cpu pipeline machine-width across micro-architectures.

The standalone slots event, like the instruction event, can be counted
with gp counter or fixed counter 3 (if any). As the last CPUID-hinted
Architectural Performance Events, its availability is also controlled by
CPUID.AH.EBX. On the Linux, perf user may encode "-e cpu/event=0xa4,
umask=0x01/" or "-e cpu/slots/" to count slots events.

Signed-off-by: Like Xu <likexu@...cent.com>
---
 arch/x86/kvm/vmx/pmu_intel.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/arch/x86/kvm/vmx/pmu_intel.c b/arch/x86/kvm/vmx/pmu_intel.c
index b746381307c7..d86a6ba8c3f9 100644
--- a/arch/x86/kvm/vmx/pmu_intel.c
+++ b/arch/x86/kvm/vmx/pmu_intel.c
@@ -31,10 +31,11 @@
  * 4 - PERF_COUNT_HW_CACHE_MISSES
  * 5 - PERF_COUNT_HW_BRANCH_INSTRUCTIONS
  * 6 - PERF_COUNT_HW_BRANCH_MISSES
+ * 7 - CPUID-hinted Topdown Slots event (available on gp counter)
  *
  * the second part of hw_events is defined by the generic kernel perf:
  *
- * 7 - PERF_COUNT_HW_REF_CPU_CYCLES
+ * 8 - PERF_COUNT_HW_REF_CPU_CYCLES
  */
 static struct kvm_pmu_hw_event intel_arch_events[] = {
 	[0] = { 0x3c, 0x00 },
@@ -44,12 +45,13 @@ static struct kvm_pmu_hw_event intel_arch_events[] = {
 	[4] = { 0x2e, 0x41 },
 	[5] = { 0xc4, 0x00 },
 	[6] = { 0xc5, 0x00 },
+	[7] = { 0xa4, 0x01 },
 	/* The above index must match CPUID 0x0A.EBX bit vector */
-	[7] = { 0x00, 0x03 },
+	[8] = { 0x00, 0x03 },
 };
 
 /* mapping between fixed pmc index and intel_arch_events array */
-static int fixed_pmc_events[] = {1, 0, 7};
+static int fixed_pmc_events[] = {1, 0, 8};
 
 static void reprogram_fixed_counters(struct kvm_pmu *pmu, u64 data)
 {
@@ -109,7 +111,7 @@ static bool intel_hw_event_available(struct kvm_pmc *pmc)
 			continue;
 
 		/* disable event that reported as not present by cpuid */
-		if ((i < 7) && !(pmu->available_event_types & (1 << i)))
+		if (i < 8 && !(pmu->available_event_types & (1 << i)))
 			return false;
 
 		break;
-- 
2.38.2