kvm.vger.kernel.org archive mirror
 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>,
	Xiong Zhang <xiong.y.zhang@intel.com>,
	Zhenyu Wang <zhenyuw@linux.intel.com>,
	Like Xu <like.xu.linux@gmail.com>,
	Jinrong Liang <cloudliang@tencent.com>,
	Yongwei Ma <yongwei.ma@intel.com>,
	Dapeng Mi <dapeng1.mi@intel.com>,
	Dapeng Mi <dapeng1.mi@linux.intel.com>
Subject: [kvm-unit-tests patch v6 07/18] x86: pmu: Fix potential out of bound access for fixed events
Date: Sat, 14 Sep 2024 10:17:17 +0000	[thread overview]
Message-ID: <20240914101728.33148-8-dapeng1.mi@linux.intel.com> (raw)
In-Reply-To: <20240914101728.33148-1-dapeng1.mi@linux.intel.com>

Current PMU code doesn't check whether PMU fixed counter number is
larger than pre-defined fixed events. If so, it would cause memory
access out of range.

So limit validated fixed counters number to
MIN(pmu.nr_fixed_counters, ARRAY_SIZE(fixed_events)) and print message
to warn that KUT/pmu tests need to be updated if fixed counters number
exceeds defined fixed events number.

Signed-off-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
---
 x86/pmu.c | 27 +++++++++++++++++++++------
 1 file changed, 21 insertions(+), 6 deletions(-)

diff --git a/x86/pmu.c b/x86/pmu.c
index 53bb7ec0..cc940a61 100644
--- a/x86/pmu.c
+++ b/x86/pmu.c
@@ -54,6 +54,7 @@ char *buf;
 
 static struct pmu_event *gp_events;
 static unsigned int gp_events_size;
+static unsigned int fixed_counters_num;
 
 static inline void loop(void)
 {
@@ -113,8 +114,12 @@ static struct pmu_event* get_counter_event(pmu_counter_t *cnt)
 		for (i = 0; i < gp_events_size; i++)
 			if (gp_events[i].unit_sel == (cnt->config & 0xffff))
 				return &gp_events[i];
-	} else
-		return &fixed_events[cnt->ctr - MSR_CORE_PERF_FIXED_CTR0];
+	} else {
+		unsigned int idx = cnt->ctr - MSR_CORE_PERF_FIXED_CTR0;
+
+		if (idx < ARRAY_SIZE(fixed_events))
+			return &fixed_events[idx];
+	}
 
 	return (void*)0;
 }
@@ -204,8 +209,12 @@ static noinline void __measure(pmu_counter_t *evt, uint64_t count)
 
 static bool verify_event(uint64_t count, struct pmu_event *e)
 {
-	bool pass = count >= e->min && count <= e->max;
+	bool pass;
 
+	if (!e)
+		return false;
+
+	pass = count >= e->min && count <= e->max;
 	if (!pass)
 		printf("FAIL: %d <= %"PRId64" <= %d\n", e->min, count, e->max);
 
@@ -250,7 +259,7 @@ static void check_fixed_counters(void)
 	};
 	int i;
 
-	for (i = 0; i < pmu.nr_fixed_counters; i++) {
+	for (i = 0; i < fixed_counters_num; i++) {
 		cnt.ctr = fixed_events[i].unit_sel;
 		measure_one(&cnt);
 		report(verify_event(cnt.count, &fixed_events[i]), "fixed-%d", i);
@@ -271,7 +280,7 @@ static void check_counters_many(void)
 			gp_events[i % gp_events_size].unit_sel;
 		n++;
 	}
-	for (i = 0; i < pmu.nr_fixed_counters; i++) {
+	for (i = 0; i < fixed_counters_num; i++) {
 		cnt[n].ctr = fixed_events[i].unit_sel;
 		cnt[n].config = EVNTSEL_OS | EVNTSEL_USR;
 		n++;
@@ -419,7 +428,7 @@ static void check_rdpmc(void)
 		else
 			report(cnt.count == (u32)val, "fast-%d", i);
 	}
-	for (i = 0; i < pmu.nr_fixed_counters; i++) {
+	for (i = 0; i < fixed_counters_num; i++) {
 		uint64_t x = val & ((1ull << pmu.fixed_counter_width) - 1);
 		pmu_counter_t cnt = {
 			.ctr = MSR_CORE_PERF_FIXED_CTR0 + i,
@@ -744,6 +753,12 @@ int main(int ac, char **av)
 	printf("Fixed counters:      %d\n", pmu.nr_fixed_counters);
 	printf("Fixed counter width: %d\n", pmu.fixed_counter_width);
 
+	fixed_counters_num = MIN(pmu.nr_fixed_counters, ARRAY_SIZE(fixed_events));
+	if (pmu.nr_fixed_counters > ARRAY_SIZE(fixed_events))
+		report_info("Fixed counters number %d > defined fixed events %ld. "
+			    "Please update test case.", pmu.nr_fixed_counters,
+			    ARRAY_SIZE(fixed_events));
+
 	apic_write(APIC_LVTPC, PMI_VECTOR);
 
 	check_counters();
-- 
2.40.1


  parent reply	other threads:[~2024-09-14  7:01 UTC|newest]

Thread overview: 33+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-09-14 10:17 [kvm-unit-tests patch v6 00/18] pmu test bugs fix and improvements Dapeng Mi
2024-09-14 10:17 ` [kvm-unit-tests patch v6 01/18] x86: pmu: Remove duplicate code in pmu_init() Dapeng Mi
2024-09-14 10:17 ` [kvm-unit-tests patch v6 02/18] x86: pmu: Remove blank line and redundant space Dapeng Mi
2024-09-14 10:17 ` [kvm-unit-tests patch v6 03/18] x86: pmu: Refine fixed_events[] names Dapeng Mi
2024-09-14 10:17 ` [kvm-unit-tests patch v6 04/18] x86: pmu: Fix the issue that pmu_counter_t.config crosses cache line Dapeng Mi
2025-02-14 21:05   ` Sean Christopherson
2025-02-18  9:07     ` Mi, Dapeng
2024-09-14 10:17 ` [kvm-unit-tests patch v6 05/18] x86: pmu: Enlarge cnt[] length to 48 in check_counters_many() Dapeng Mi
2025-02-14 21:06   ` Sean Christopherson
2025-02-18  9:24     ` Mi, Dapeng
2025-02-18 15:56       ` Sean Christopherson
2024-09-14 10:17 ` [kvm-unit-tests patch v6 06/18] x86: pmu: Print measured event count if test fails Dapeng Mi
2024-09-14 10:17 ` Dapeng Mi [this message]
2025-02-14 21:07   ` [kvm-unit-tests patch v6 07/18] x86: pmu: Fix potential out of bound access for fixed events Sean Christopherson
2025-02-18  9:34     ` Mi, Dapeng
2025-02-18 15:04       ` Sean Christopherson
2024-09-14 10:17 ` [kvm-unit-tests patch v6 08/18] x86: pmu: Fix cycles event validation failure Dapeng Mi
2025-02-14 21:07   ` Sean Christopherson
2025-02-18  9:36     ` Mi, Dapeng
2024-09-14 10:17 ` [kvm-unit-tests patch v6 09/18] x86: pmu: Use macro to replace hard-coded branches event index Dapeng Mi
2024-09-14 10:17 ` [kvm-unit-tests patch v6 10/18] x86: pmu: Use macro to replace hard-coded ref-cycles " Dapeng Mi
2024-09-14 10:17 ` [kvm-unit-tests patch v6 11/18] x86: pmu: Use macro to replace hard-coded instructions " Dapeng Mi
2024-09-14 10:17 ` [kvm-unit-tests patch v6 12/18] x86: pmu: Enable and disable PMCs in loop() asm blob Dapeng Mi
2024-09-14 10:17 ` [kvm-unit-tests patch v6 13/18] x86: pmu: Improve instruction and branches events verification Dapeng Mi
2025-02-14 21:08   ` Sean Christopherson
2025-02-18  9:40     ` Mi, Dapeng
2024-09-14 10:17 ` [kvm-unit-tests patch v6 14/18] x86: pmu: Improve LLC misses event verification Dapeng Mi
2024-09-14 10:17 ` [kvm-unit-tests patch v6 15/18] x86: pmu: Adjust lower boundary of llc-misses event to 0 for legacy CPUs Dapeng Mi
2024-09-14 10:17 ` [kvm-unit-tests patch v6 16/18] x86: pmu: Add IBPB indirect jump asm blob Dapeng Mi
2024-09-14 10:17 ` [kvm-unit-tests patch v6 17/18] x86: pmu: Adjust lower boundary of branch-misses event Dapeng Mi
2025-02-14 21:09   ` Sean Christopherson
2025-02-18  9:42     ` Mi, Dapeng
2024-09-14 10:17 ` [kvm-unit-tests patch v6 18/18] x86: pmu: Optimize emulated instruction validation Dapeng Mi

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=20240914101728.33148-8-dapeng1.mi@linux.intel.com \
    --to=dapeng1.mi@linux.intel.com \
    --cc=cloudliang@tencent.com \
    --cc=dapeng1.mi@intel.com \
    --cc=jmattson@google.com \
    --cc=kvm@vger.kernel.org \
    --cc=like.xu.linux@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mizhang@google.com \
    --cc=pbonzini@redhat.com \
    --cc=seanjc@google.com \
    --cc=xiong.y.zhang@intel.com \
    --cc=yongwei.ma@intel.com \
    --cc=zhenyuw@linux.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;
as well as URLs for NNTP newsgroup(s).