public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: Dapeng Mi <dapeng1.mi@linux.intel.com>
To: Sean Christopherson <seanjc@google.com>,
	Paolo Bonzini <pbonzini@redhat.com>
Cc: kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
	Jim Mattson <jmattson@google.com>,
	Mingwei Zhang <mizhang@google.com>,
	Zide Chen <zide.chen@intel.com>,
	Das Sandipan <Sandipan.Das@amd.com>,
	Shukla Manali <Manali.Shukla@amd.com>,
	Xiaoyao Li <xiaoyao.li@intel.com>,
	Dapeng Mi <dapeng1.mi@intel.com>,
	dongsheng <dongsheng.x.zhang@intel.com>,
	Dapeng Mi <dapeng1.mi@linux.intel.com>,
	Yi Lai <yi1.lai@intel.com>
Subject: [kvm-unit-tests patch v3 4/8] x86/pmu: Handle instruction overcount issue in overflow test
Date: Wed,  3 Sep 2025 14:45:57 +0800	[thread overview]
Message-ID: <20250903064601.32131-5-dapeng1.mi@linux.intel.com> (raw)
In-Reply-To: <20250903064601.32131-1-dapeng1.mi@linux.intel.com>

From: dongsheng <dongsheng.x.zhang@intel.com>

During the execution of __measure(), VM exits (e.g., due to
WRMSR/EXTERNAL_INTERRUPT) may occur. On systems affected by the
instruction overcount issue, each VM-Exit/VM-Entry can erroneously
increment the instruction count by one, leading to false failures in
overflow tests.

To address this, the patch introduces a range-based validation in place
of precise instruction count checks. Additionally, overflow_preset is
now statically set to 1 - LOOP_INSNS, rather than being dynamically
determined via measure_for_overflow().

These changes ensure consistent and predictable behavior aligned with the
intended loop instruction count, while avoiding modifications to the
subsequent status and status-clear testing logic.

The chosen validation range is empirically derived to maintain test
reliability across hardware variations.

Signed-off-by: dongsheng <dongsheng.x.zhang@intel.com>
Signed-off-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Tested-by: Yi Lai <yi1.lai@intel.com>
---
 x86/pmu.c | 23 +++++++++++++++++++++--
 1 file changed, 21 insertions(+), 2 deletions(-)

diff --git a/x86/pmu.c b/x86/pmu.c
index 44c728a5..c54c0988 100644
--- a/x86/pmu.c
+++ b/x86/pmu.c
@@ -518,6 +518,21 @@ static void check_counters_many(void)
 
 static uint64_t measure_for_overflow(pmu_counter_t *cnt)
 {
+	/*
+	 * During the execution of __measure(), VM exits (e.g., due to
+	 * WRMSR/EXTERNAL_INTERRUPT) may occur. On systems affected by the
+	 * instruction overcount issue, each VM-Exit/VM-Entry can erroneously
+	 * increment the instruction count by one, leading to false failures
+	 * in overflow tests.
+	 *
+	 * To mitigate this, if the overcount issue is detected, we hardcode
+	 * the overflow preset to (1 - LOOP_INSNS) instead of calculating it
+	 * dynamically. This ensures that an overflow will reliably occur,
+	 * regardless of any overcounting caused by VM exits.
+	 */
+	if (intel_inst_overcount_flags & INST_RETIRED_OVERCOUNT)
+		return 1 - LOOP_INSNS;
+
 	__measure(cnt, 0);
 	/*
 	 * To generate overflow, i.e. roll over to '0', the initial count just
@@ -574,8 +589,12 @@ static void check_counter_overflow(void)
 			cnt.config &= ~EVNTSEL_INT;
 		idx = event_to_global_idx(&cnt);
 		__measure(&cnt, cnt.count);
-		if (pmu.is_intel)
-			report(cnt.count == 1, "cntr-%d", i);
+		if (pmu.is_intel) {
+			if (intel_inst_overcount_flags & INST_RETIRED_OVERCOUNT)
+				report(cnt.count < 14, "cntr-%d", i);
+			else
+				report(cnt.count == 1, "cntr-%d", i);
+		}
 		else
 			report(cnt.count == 0xffffffffffff || cnt.count < 7, "cntr-%d", i);
 
-- 
2.34.1


  parent reply	other threads:[~2025-09-03  6:47 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-09-03  6:45 [kvm-unit-tests patch v3 0/8] Fix pmu test errors on GNR/SRF/CWF Dapeng Mi
2025-09-03  6:45 ` [kvm-unit-tests patch v3 1/8] x86/pmu: Add helper to detect Intel overcount issues Dapeng Mi
2025-11-20 22:27   ` Sean Christopherson
2025-11-21  1:18     ` Mi, Dapeng
2025-09-03  6:45 ` [kvm-unit-tests patch v3 2/8] x86/pmu: Relax precise count validation for Intel overcounted platforms Dapeng Mi
2025-09-03  6:45 ` [kvm-unit-tests patch v3 3/8] x86/pmu: Fix incorrect masking of fixed counters Dapeng Mi
2025-11-20 22:28   ` Sean Christopherson
2025-11-21  1:25     ` Mi, Dapeng
2025-09-03  6:45 ` Dapeng Mi [this message]
2025-09-03  6:45 ` [kvm-unit-tests patch v3 5/8] x86/pmu: Relax precise count check for emulated instructions tests Dapeng Mi
2025-11-20 22:29   ` Sean Christopherson
2025-11-21  0:53     ` Mi, Dapeng
2025-09-03  6:45 ` [kvm-unit-tests patch v3 6/8] x86/pmu: Expand "llc references" upper limit for broader compatibility Dapeng Mi
2025-09-03  6:46 ` [kvm-unit-tests patch v3 7/8] x86: pmu_pebs: Remove abundant data_cfg_match calculation Dapeng Mi
2025-09-03  6:46 ` [kvm-unit-tests patch v3 8/8] x86: pmu_pebs: Support to validate timed PEBS record on GNR/SRF Dapeng Mi
2025-11-20 22:30 ` [kvm-unit-tests patch v3 0/8] Fix pmu test errors on GNR/SRF/CWF Sean Christopherson

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20250903064601.32131-5-dapeng1.mi@linux.intel.com \
    --to=dapeng1.mi@linux.intel.com \
    --cc=Manali.Shukla@amd.com \
    --cc=Sandipan.Das@amd.com \
    --cc=dapeng1.mi@intel.com \
    --cc=dongsheng.x.zhang@intel.com \
    --cc=jmattson@google.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mizhang@google.com \
    --cc=pbonzini@redhat.com \
    --cc=seanjc@google.com \
    --cc=xiaoyao.li@intel.com \
    --cc=yi1.lai@intel.com \
    --cc=zide.chen@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox