Skip to contentSkip to conversation

perf/x86/intel: Only check GP counters for PEBS constraints validation

PATCH RESEND Applied Dapeng Mi 4 participants 16 messages Feb 28 – Mar 16

Subsystems (3)
lore
Share & tools
lore

Dapeng Mi

author
#0 lore raw

PATCH RESEND 1/2 perf/x86/intel: Only check GP counters for PEBS constraints validation

It's good enough to only check GP counters for PEBS constraints validation since constraints overlap can only happen on GP counters.

Besides opportunistically refine the code style and use pr_warn() to replace pr_info() as the message itself is a warning message.

---
 arch/x86/events/intel/core.c | 22 ++++++++++++++--------
 1 file changed, 14 insertions(+), 8 deletions(-)
arch/x86/events/intel/core.c +14 -8
index cf3a4fe06ff2..4768236c054b 100644
@@ -5770,7 +5770,7 @@ static void __intel_pmu_check_dyn_constr(struct event_constraint *constr,
 			}
 
 			if (check_fail) {
-				pr_info("The two events 0x%llx and 0x%llx may not be "
+				pr_warn("The two events 0x%llx and 0x%llx may not be "
 					"fully scheduled under some circumstances as "
 					"%s.\n",
 					c1->code, c2->code, dyn_constr_type_name[type]);
@@ -5783,6 +5783,7 @@ static void intel_pmu_check_dyn_constr(struct pmu *pmu,
 				       struct event_constraint *constr,
 				       u64 cntr_mask)
 {
+	u64 gp_mask = GENMASK_ULL(INTEL_PMC_MAX_GENERIC - 1, 0);
 	enum dyn_constr_type i;
 	u64 mask;
 
@@ -5797,20 +5798,25 @@ static void intel_pmu_check_dyn_constr(struct pmu *pmu,
 				mask = x86_pmu.lbr_counters;
 			break;
 		case DYN_CONSTR_ACR_CNTR:
-			mask = hybrid(pmu, acr_cntr_mask64) & GENMASK_ULL(INTEL_PMC_MAX_GENERIC - 1, 0);
+			mask = hybrid(pmu, acr_cntr_mask64) & gp_mask;
 			break;
 		case DYN_CONSTR_ACR_CAUSE:
-			if (hybrid(pmu, acr_cntr_mask64) == hybrid(pmu, acr_cause_mask64))
+			if (hybrid(pmu, acr_cntr_mask64) ==
+					hybrid(pmu, acr_cause_mask64))
 				continue;
-			mask = hybrid(pmu, acr_cause_mask64) & GENMASK_ULL(INTEL_PMC_MAX_GENERIC - 1, 0);
+			mask = hybrid(pmu, acr_cause_mask64) & gp_mask;
 			break;
 		case DYN_CONSTR_PEBS:
-			if (x86_pmu.arch_pebs)
-				mask = hybrid(pmu, arch_pebs_cap).counters;
+			if (x86_pmu.arch_pebs) {
+				mask = hybrid(pmu, arch_pebs_cap).counters &
+				       gp_mask;
+			}
 			break;
 		case DYN_CONSTR_PDIST:
-			if (x86_pmu.arch_pebs)
-				mask = hybrid(pmu, arch_pebs_cap).pdists;
+			if (x86_pmu.arch_pebs) {
+				mask = hybrid(pmu, arch_pebs_cap).pdists &
+				       gp_mask;
+			}
 			break;
 		default:
 			pr_warn("Unsupported dynamic constraint type %d\n", i);

base-commit: 6de23f81a5e08be8fbf5e8d7e9febc72a5b5f27f

Signature · 1 line
2.34.1
author
#1 lore raw

PATCH RESEND 2/2 perf/x86/intel: Add missing branch counters constraint apply

When running the command:
'perf record -e "{instructions,instructions:p}" -j any,counter sleep 1', a "shift-out-of-bounds" warning is reported on CWF.

[ 5231.981423][   C17] UBSAN: shift-out-of-bounds in /kbuild/src/consumer/arch/x86/events/intel/lbr.c:970:15
[ 5231.981428][   C17] shift exponent 64 is too large for 64-bit type 'long long unsigned int'
[ 5231.981436][   C17] CPU: 17 UID: 0 PID: 211871 Comm: sleep Tainted: G S      W           6.18.0-2025-12-09-intel-next-48166-g6cf574943ba3 #1 PREEMPT(none)
[ 5231.981445][   C17] Tainted: [S]=CPU_OUT_OF_SPEC, [W]=WARN
[ 5231.981447][   C17] Hardware name: Intel Corporation AvenueCity/AvenueCity, BIOS BHSDCRB1.IPC.3544.P98.2508260307 08/26/2025
[ 5231.981449][   C17] Call Trace:
[ 5231.981453][   C17]  <NMI>
[ 5231.981455][   C17]  dump_stack_lvl+0x4b/0x70
[ 5231.981463][   C17]  ubsan_epilogue+0x5/0x2b
[ 5231.981468][   C17]  __ubsan_handle_shift_out_of_bounds.cold+0x61/0xe6
[ 5231.981472][   C17]  ? __entry_text_end+0x158b/0x102259
[ 5231.981475][   C17]  intel_pmu_lbr_counters_reorder.isra.0.cold+0x2a/0xa7
[ 5231.981480][   C17]  ? __task_pid_nr_ns+0x134/0x2a0
[ 5231.981483][   C17]  ? __pfx_intel_pmu_lbr_counters_reorder.isra.0+0x10/0x10
[ 5231.981486][   C17]  ? __pfx_perf_output_sample+0x10/0x10
[ 5231.981489][   C17]  ? arch_perf_update_userpage+0x293/0x310
[ 5231.981491][   C17]  ? __pfx_arch_perf_update_userpage+0x10/0x10
[ 5231.981494][   C17]  ? local_clock_noinstr+0xd/0x100
[ 5231.981498][   C17]  ? calc_timer_values+0x2cb/0x860
[ 5231.981501][   C17]  ? perf_event_update_userpage+0x399/0x5b0
[ 5231.981505][   C17]  ? __pfx_perf_event_update_userpage+0x10/0x10
[ 5231.981508][   C17]  ? local_clock_noinstr+0xd/0x100
[ 5231.981511][   C17]  ? __perf_event_account_interrupt+0x11c/0x540
[ 5231.981514][   C17]  intel_pmu_lbr_save_brstack+0xc0/0x4c0
[ 5231.981518][   C17]  setup_arch_pebs_sample_data+0x114b/0x2400
[ 5231.981522][   C17]  ? __pfx_x86_perf_event_set_period+0x10/0x10
[ 5231.981526][   C17]  intel_pmu_drain_arch_pebs+0x64d/0xcc0
[ 5231.981530][   C17]  ? __pfx_intel_pmu_drain_arch_pebs+0x10/0x10
[ 5231.981534][   C17]  ? unwind_next_frame+0x11c5/0x1df0
[ 5231.981541][   C17]  ? intel_pmu_drain_bts_buffer+0xbf/0x6e0
[ 5231.981545][   C17]  ? __pfx_intel_pmu_drain_bts_buffer+0x10/0x10
[ 5231.981550][   C17]  handle_pmi_common+0x5c5/0xcb0
[ 5231.981553][   C17]  ? __pfx_handle_pmi_common+0x10/0x10
[ 5231.981556][   C17]  ? intel_idle+0x64/0xb0
[ 5231.981560][   C17]  ? intel_bts_interrupt+0xe5/0x4c0
[ 5231.981562][   C17]  ? __pfx_intel_bts_interrupt+0x10/0x10
[ 5231.981565][   C17]  ? intel_pmu_lbr_filter+0x27f/0x910
[ 5231.981568][   C17]  intel_pmu_handle_irq+0x2ed/0x600
[ 5231.981571][   C17]  perf_event_nmi_handler+0x219/0x280
[ 5231.981575][   C17]  ? __pfx_perf_event_nmi_handler+0x10/0x10
[ 5231.981579][   C17]  ? unwind_next_frame+0x11c5/0x1df0
[ 5231.981582][   C17]  nmi_handle.part.0+0x11b/0x3a0
[ 5231.981585][   C17]  ? unwind_next_frame+0x11c5/0x1df0
[ 5231.981588][   C17]  default_do_nmi+0x6b/0x180
[ 5231.981591][   C17]  fred_exc_nmi+0x3e/0x80
[ 5231.981594][   C17]  asm_fred_entrypoint_kernel+0x41/0x60
[ 5231.981596][   C17] RIP: 0010:unwind_next_frame+0x11c5/0x1df0

......

The warning occurs because the second "instructions:p" event, which involves branch counters sampling, is incorrectly programmed to fixed counter 0 instead of the general-purpose (GP) counters 0-3 that support branch counters sampling. Currently only GP counters 0~3 support branch counters sampling on CWF, any event involving branch counters sampling should be programed on GP counters 0~3. Since the counter index of fixed counter 0 is 32, it leads to the "src" value in below code is right shifted 64 bits and trigger the "shift-out-of-bounds" warning.

cnt = (src >> (order[j] * LBR_INFO_BR_CNTR_BITS)) & LBR_INFO_BR_CNTR_MASK;

The root cause is the loss of the branch counters constraint for the last event in the branch counters sampling event group. This results in the second "instructions:p" event being programmed on fixed counter 0 incorrectly instead of the appropriate GP counters 0~3.

To address this, we apply the missing branch counters constraint for the last event in the group. Additionally, we introduce a new function, intel_set_branch_counter_constr(), to apply the branch counters constraint and avoid code duplication.

Fixes 33744916196b ("perf/x86/intel: Support branch counters logging")
---
 arch/x86/events/intel/core.c | 30 ++++++++++++++++++++----------
 1 file changed, 20 insertions(+), 10 deletions(-)
arch/x86/events/intel/core.c +20 -10
index 4768236c054b..4b042d71104f 100644
@@ -4628,6 +4628,19 @@ static inline void intel_pmu_set_acr_caused_constr(struct perf_event *event,
 		event->hw.dyn_constraint &= hybrid(event->pmu, acr_cause_mask64);
 }
 
+static inline int intel_set_branch_counter_constr(struct perf_event *event,
+						  int *num)
+{
+	if (branch_sample_call_stack(event))
+		return -EINVAL;
+	if (branch_sample_counters(event)) {
+		(*num)++;
+		event->hw.dyn_constraint &= x86_pmu.lbr_counters;
+	}
+
+	return 0;
+}
+
 static int intel_pmu_hw_config(struct perf_event *event)
 {
 	int ret = x86_pmu_hw_config(event);
@@ -4698,21 +4711,18 @@ static int intel_pmu_hw_config(struct perf_event *event)
 		 * group, which requires the extra space to store the counters.
 		 */
 		leader = event->group_leader;
-		if (branch_sample_call_stack(leader))
+		if (intel_set_branch_counter_constr(leader, &num))
 			return -EINVAL;
-		if (branch_sample_counters(leader)) {
-			num++;
-			leader->hw.dyn_constraint &= x86_pmu.lbr_counters;
-		}
 		leader->hw.flags |= PERF_X86_EVENT_BRANCH_COUNTERS;
 
 		for_each_sibling_event(sibling, leader) {
-			if (branch_sample_call_stack(sibling))
+			if (intel_set_branch_counter_constr(sibling, &num))
+				return -EINVAL;
+		}
+
+		if (event != leader) {
+			if (intel_set_branch_counter_constr(event, &num))
 				return -EINVAL;
-			if (branch_sample_counters(sibling)) {
-				num++;
-				sibling->hw.dyn_constraint &= x86_pmu.lbr_counters;
-			}
 		}
 
 		if (num > fls(x86_pmu.lbr_counters))
Signature · 1 line
2.34.1
Reviewed-by on v1 2/2
#2 lore raw

Re: [RESEND Patch 2/2] perf/x86/intel: Add missing branch counters constraint apply

Quoted 74 lines · Dapeng Mi

When running the command:
'perf record -e "{instructions,instructions:p}" -j any,counter sleep 1', a "shift-out-of-bounds" warning is reported on CWF.

[ 5231.981423][   C17] UBSAN: shift-out-of-bounds in /kbuild/src/consumer/arch/x86/events/intel/lbr.c:970:15
[ 5231.981428][   C17] shift exponent 64 is too large for 64-bit type 'long long unsigned int'
[ 5231.981436][   C17] CPU: 17 UID: 0 PID: 211871 Comm: sleep Tainted: G S      W           6.18.0-2025-12-09-intel-next-48166-g6cf574943ba3 #1 PREEMPT(none)
[ 5231.981445][   C17] Tainted: [S]=CPU_OUT_OF_SPEC, [W]=WARN
[ 5231.981447][   C17] Hardware name: Intel Corporation AvenueCity/AvenueCity, BIOS BHSDCRB1.IPC.3544.P98.2508260307 08/26/2025
[ 5231.981449][   C17] Call Trace:
[ 5231.981453][   C17]  <NMI>
[ 5231.981455][   C17]  dump_stack_lvl+0x4b/0x70
[ 5231.981463][   C17]  ubsan_epilogue+0x5/0x2b
[ 5231.981468][   C17]  __ubsan_handle_shift_out_of_bounds.cold+0x61/0xe6
[ 5231.981472][   C17]  ? __entry_text_end+0x158b/0x102259
[ 5231.981475][   C17]  intel_pmu_lbr_counters_reorder.isra.0.cold+0x2a/0xa7
[ 5231.981480][   C17]  ? __task_pid_nr_ns+0x134/0x2a0
[ 5231.981483][   C17]  ? __pfx_intel_pmu_lbr_counters_reorder.isra.0+0x10/0x10
[ 5231.981486][   C17]  ? __pfx_perf_output_sample+0x10/0x10
[ 5231.981489][   C17]  ? arch_perf_update_userpage+0x293/0x310
[ 5231.981491][   C17]  ? __pfx_arch_perf_update_userpage+0x10/0x10
[ 5231.981494][   C17]  ? local_clock_noinstr+0xd/0x100
[ 5231.981498][   C17]  ? calc_timer_values+0x2cb/0x860
[ 5231.981501][   C17]  ? perf_event_update_userpage+0x399/0x5b0
[ 5231.981505][   C17]  ? __pfx_perf_event_update_userpage+0x10/0x10
[ 5231.981508][   C17]  ? local_clock_noinstr+0xd/0x100
[ 5231.981511][   C17]  ? __perf_event_account_interrupt+0x11c/0x540
[ 5231.981514][   C17]  intel_pmu_lbr_save_brstack+0xc0/0x4c0
[ 5231.981518][   C17]  setup_arch_pebs_sample_data+0x114b/0x2400
[ 5231.981522][   C17]  ? __pfx_x86_perf_event_set_period+0x10/0x10
[ 5231.981526][   C17]  intel_pmu_drain_arch_pebs+0x64d/0xcc0
[ 5231.981530][   C17]  ? __pfx_intel_pmu_drain_arch_pebs+0x10/0x10
[ 5231.981534][   C17]  ? unwind_next_frame+0x11c5/0x1df0
[ 5231.981541][   C17]  ? intel_pmu_drain_bts_buffer+0xbf/0x6e0
[ 5231.981545][   C17]  ? __pfx_intel_pmu_drain_bts_buffer+0x10/0x10
[ 5231.981550][   C17]  handle_pmi_common+0x5c5/0xcb0
[ 5231.981553][   C17]  ? __pfx_handle_pmi_common+0x10/0x10
[ 5231.981556][   C17]  ? intel_idle+0x64/0xb0
[ 5231.981560][   C17]  ? intel_bts_interrupt+0xe5/0x4c0
[ 5231.981562][   C17]  ? __pfx_intel_bts_interrupt+0x10/0x10
[ 5231.981565][   C17]  ? intel_pmu_lbr_filter+0x27f/0x910
[ 5231.981568][   C17]  intel_pmu_handle_irq+0x2ed/0x600
[ 5231.981571][   C17]  perf_event_nmi_handler+0x219/0x280
[ 5231.981575][   C17]  ? __pfx_perf_event_nmi_handler+0x10/0x10
[ 5231.981579][   C17]  ? unwind_next_frame+0x11c5/0x1df0
[ 5231.981582][   C17]  nmi_handle.part.0+0x11b/0x3a0
[ 5231.981585][   C17]  ? unwind_next_frame+0x11c5/0x1df0
[ 5231.981588][   C17]  default_do_nmi+0x6b/0x180
[ 5231.981591][   C17]  fred_exc_nmi+0x3e/0x80
[ 5231.981594][   C17]  asm_fred_entrypoint_kernel+0x41/0x60
[ 5231.981596][   C17] RIP: 0010:unwind_next_frame+0x11c5/0x1df0

......

The warning occurs because the second "instructions:p" event, which
involves branch counters sampling, is incorrectly programmed to fixed
counter 0 instead of the general-purpose (GP) counters 0-3 that support
branch counters sampling. Currently only GP counters 0~3 support branch
counters sampling on CWF, any event involving branch counters sampling
should be programed on GP counters 0~3.
Since the counter index of fixed> counter 0 is 32, it leads to the "src" value in below code is right shifted 64 bits and trigger the "shift-out-of-bounds" warning.

cnt = (src >> (order[j] * LBR_INFO_BR_CNTR_BITS)) & LBR_INFO_BR_CNTR_MASK;

The root cause is the loss of the branch counters constraint for the last event in the branch counters sampling event group. This results in the second "instructions:p" event being programmed on fixed counter 0 incorrectly instead of the appropriate GP counters 0~3.

To address this, we apply the missing branch counters constraint for the last event in the group. Additionally, we introduce a new function, intel_set_branch_counter_constr(), to apply the branch counters constraint and avoid code duplication.

Quoted 5 lines
Fixes 33744916196b ("perf/x86/intel: Support branch counters logging")

---

Quoted 56 lines
 arch/x86/events/intel/core.c | 30 ++++++++++++++++++++----------
 1 file changed, 20 insertions(+), 10 deletions(-)
arch/x86/events/intel/core.c +20 -10
index 4768236c054b..4b042d71104f 100644
@@ -4628,6 +4628,19 @@ static inline void intel_pmu_set_acr_caused_constr(struct perf_event *event,
 		event->hw.dyn_constraint &= hybrid(event->pmu, acr_cause_mask64);
 }
 
+static inline int intel_set_branch_counter_constr(struct perf_event *event,
+						  int *num)
+{
+	if (branch_sample_call_stack(event))
+		return -EINVAL;
+	if (branch_sample_counters(event)) {
+		(*num)++;
+		event->hw.dyn_constraint &= x86_pmu.lbr_counters;
+	}
+
+	return 0;
+}
+
 static int intel_pmu_hw_config(struct perf_event *event)
 {
 	int ret = x86_pmu_hw_config(event);
@@ -4698,21 +4711,18 @@ static int intel_pmu_hw_config(struct perf_event *event)
 		 * group, which requires the extra space to store the counters.
 		 */
 		leader = event->group_leader;
-		if (branch_sample_call_stack(leader))
+		if (intel_set_branch_counter_constr(leader, &num))
 			return -EINVAL;
-		if (branch_sample_counters(leader)) {
-			num++;
-			leader->hw.dyn_constraint &= x86_pmu.lbr_counters;
-		}
 		leader->hw.flags |= PERF_X86_EVENT_BRANCH_COUNTERS;
 
 		for_each_sibling_event(sibling, leader) {
-			if (branch_sample_call_stack(sibling))
+			if (intel_set_branch_counter_constr(sibling, &num))
+				return -EINVAL;
+		}
+
+		if (event != leader) {
+			if (intel_set_branch_counter_constr(event, &num))
 				return -EINVAL;
-			if (branch_sample_counters(sibling)) {
-				num++;
-				sibling->hw.dyn_constraint &= x86_pmu.lbr_counters;
-			}
 		}
 
 		if (num > fls(x86_pmu.lbr_counters))
maintainer
#3 lore raw

Re: [RESEND Patch 2/2] perf/x86/intel: Add missing branch counters constraint apply

Quoted 47 lines · Zide
Quoted 44 lines · Dapeng Mi

When running the command:
'perf record -e "{instructions,instructions:p}" -j any,counter sleep 1', a "shift-out-of-bounds" warning is reported on CWF.

[ 5231.981423][   C17] UBSAN: shift-out-of-bounds in /kbuild/src/consumer/arch/x86/events/intel/lbr.c:970:15
[ 5231.981428][   C17] shift exponent 64 is too large for 64-bit type 'long long unsigned int'
[ 5231.981436][   C17] CPU: 17 UID: 0 PID: 211871 Comm: sleep Tainted: G S      W           6.18.0-2025-12-09-intel-next-48166-g6cf574943ba3 #1 PREEMPT(none)
[ 5231.981445][   C17] Tainted: [S]=CPU_OUT_OF_SPEC, [W]=WARN
[ 5231.981447][   C17] Hardware name: Intel Corporation AvenueCity/AvenueCity, BIOS BHSDCRB1.IPC.3544.P98.2508260307 08/26/2025
[ 5231.981449][   C17] Call Trace:
[ 5231.981453][   C17]  <NMI>
[ 5231.981455][   C17]  dump_stack_lvl+0x4b/0x70
[ 5231.981463][   C17]  ubsan_epilogue+0x5/0x2b
[ 5231.981468][   C17]  __ubsan_handle_shift_out_of_bounds.cold+0x61/0xe6
[ 5231.981472][   C17]  ? __entry_text_end+0x158b/0x102259
[ 5231.981475][   C17]  intel_pmu_lbr_counters_reorder.isra.0.cold+0x2a/0xa7
[ 5231.981480][   C17]  ? __task_pid_nr_ns+0x134/0x2a0
[ 5231.981483][   C17]  ? __pfx_intel_pmu_lbr_counters_reorder.isra.0+0x10/0x10
[ 5231.981486][   C17]  ? __pfx_perf_output_sample+0x10/0x10
[ 5231.981489][   C17]  ? arch_perf_update_userpage+0x293/0x310
[ 5231.981491][   C17]  ? __pfx_arch_perf_update_userpage+0x10/0x10
[ 5231.981494][   C17]  ? local_clock_noinstr+0xd/0x100
[ 5231.981498][   C17]  ? calc_timer_values+0x2cb/0x860
[ 5231.981501][   C17]  ? perf_event_update_userpage+0x399/0x5b0
[ 5231.981505][   C17]  ? __pfx_perf_event_update_userpage+0x10/0x10
[ 5231.981508][   C17]  ? local_clock_noinstr+0xd/0x100
[ 5231.981511][   C17]  ? __perf_event_account_interrupt+0x11c/0x540
[ 5231.981514][   C17]  intel_pmu_lbr_save_brstack+0xc0/0x4c0
[ 5231.981518][   C17]  setup_arch_pebs_sample_data+0x114b/0x2400
[ 5231.981522][   C17]  ? __pfx_x86_perf_event_set_period+0x10/0x10
[ 5231.981526][   C17]  intel_pmu_drain_arch_pebs+0x64d/0xcc0
[ 5231.981530][   C17]  ? __pfx_intel_pmu_drain_arch_pebs+0x10/0x10
[ 5231.981534][   C17]  ? unwind_next_frame+0x11c5/0x1df0
[ 5231.981541][   C17]  ? intel_pmu_drain_bts_buffer+0xbf/0x6e0
[ 5231.981545][   C17]  ? __pfx_intel_pmu_drain_bts_buffer+0x10/0x10
[ 5231.981550][   C17]  handle_pmi_common+0x5c5/0xcb0
[ 5231.981553][   C17]  ? __pfx_handle_pmi_common+0x10/0x10
[ 5231.981556][   C17]  ? intel_idle+0x64/0xb0
[ 5231.981560][   C17]  ? intel_bts_interrupt+0xe5/0x4c0
[ 5231.981562][   C17]  ? __pfx_intel_bts_interrupt+0x10/0x10
[ 5231.981565][   C17]  ? intel_pmu_lbr_filter+0x27f/0x910
[ 5231.981568][   C17]  intel_pmu_handle_irq+0x2ed/0x600
[ 5231.981571][   C17]  perf_event_nmi_handler+0x219/0x280
[ 5231.981575][   C17]  ? __pfx_perf_event_nmi_handler+0x10/0x10
Quoted 8 lines
Quoted 8 lines
[ 5231.981579][   C17]  ? unwind_next_frame+0x11c5/0x1df0
[ 5231.981582][   C17]  nmi_handle.part.0+0x11b/0x3a0
[ 5231.981585][   C17]  ? unwind_next_frame+0x11c5/0x1df0
[ 5231.981588][   C17]  default_do_nmi+0x6b/0x180
[ 5231.981591][   C17]  fred_exc_nmi+0x3e/0x80
[ 5231.981594][   C17]  asm_fred_entrypoint_kernel+0x41/0x60
[ 5231.981596][   C17] RIP: 0010:unwind_next_frame+0x11c5/0x1df0

......

That trace could be reduced to:

  UBSAN: shift-out-of-bounds in /kbuild/src/consumer/arch/x86/events/intel/lbr.c:970:15
  shift exponent 64 is too large for 64-bit type 'long long unsigned int'
  ......
  intel_pmu_lbr_counters_reorder.isra.0.cold+0x2a/0xa7
  intel_pmu_lbr_save_brstack+0xc0/0x4c0
  setup_arch_pebs_sample_data+0x114b/0x2400

Without loosing anything valuable.

Quoted 3 lines
Quoted 3 lines

The warning occurs because the second "instructions:p" event, which involves branch counters sampling, is incorrectly programmed to fixed counter 0 instead of the general-purpose (GP) counters 0-3 that support

So here you have 0-3, the normal 'range' notation, but then you go all funny and use ~ instead:

Quoted 6 lines
Quoted 6 lines

branch counters sampling. Currently only GP counters 0~3 support branch
counters sampling on CWF, any event involving branch counters sampling
should be programed on GP counters 0~3.
Since the counter index of fixed> counter 0 is 32, it leads to the "src" value in below code is right shifted 64 bits and trigger the "shift-out-of-bounds" warning.

Quoted 6 lines
Quoted 6 lines

cnt = (src >> (order[j] * LBR_INFO_BR_CNTR_BITS)) & LBR_INFO_BR_CNTR_MASK;

The root cause is the loss of the branch counters constraint for the last event in the branch counters sampling event group. This results in the second "instructions:p" event being programmed on fixed counter 0 incorrectly instead of the appropriate GP counters 0~3.

s/0~3/0-3/ ?

author
#4 lore raw

Re: [RESEND Patch 2/2] perf/x86/intel: Add missing branch counters constraint apply

Quoted 57 lines · Peter Zijlstra
Quoted 46 lines · Zide
Quoted 44 lines · Dapeng Mi

When running the command:
'perf record -e "{instructions,instructions:p}" -j any,counter sleep 1', a "shift-out-of-bounds" warning is reported on CWF.

[ 5231.981423][   C17] UBSAN: shift-out-of-bounds in /kbuild/src/consumer/arch/x86/events/intel/lbr.c:970:15
[ 5231.981428][   C17] shift exponent 64 is too large for 64-bit type 'long long unsigned int'
[ 5231.981436][   C17] CPU: 17 UID: 0 PID: 211871 Comm: sleep Tainted: G S      W           6.18.0-2025-12-09-intel-next-48166-g6cf574943ba3 #1 PREEMPT(none)
[ 5231.981445][   C17] Tainted: [S]=CPU_OUT_OF_SPEC, [W]=WARN
[ 5231.981447][   C17] Hardware name: Intel Corporation AvenueCity/AvenueCity, BIOS BHSDCRB1.IPC.3544.P98.2508260307 08/26/2025
[ 5231.981449][   C17] Call Trace:
[ 5231.981453][   C17]  <NMI>
[ 5231.981455][   C17]  dump_stack_lvl+0x4b/0x70
[ 5231.981463][   C17]  ubsan_epilogue+0x5/0x2b
[ 5231.981468][   C17]  __ubsan_handle_shift_out_of_bounds.cold+0x61/0xe6
[ 5231.981472][   C17]  ? __entry_text_end+0x158b/0x102259
[ 5231.981475][   C17]  intel_pmu_lbr_counters_reorder.isra.0.cold+0x2a/0xa7
[ 5231.981480][   C17]  ? __task_pid_nr_ns+0x134/0x2a0
[ 5231.981483][   C17]  ? __pfx_intel_pmu_lbr_counters_reorder.isra.0+0x10/0x10
[ 5231.981486][   C17]  ? __pfx_perf_output_sample+0x10/0x10
[ 5231.981489][   C17]  ? arch_perf_update_userpage+0x293/0x310
[ 5231.981491][   C17]  ? __pfx_arch_perf_update_userpage+0x10/0x10
[ 5231.981494][   C17]  ? local_clock_noinstr+0xd/0x100
[ 5231.981498][   C17]  ? calc_timer_values+0x2cb/0x860
[ 5231.981501][   C17]  ? perf_event_update_userpage+0x399/0x5b0
[ 5231.981505][   C17]  ? __pfx_perf_event_update_userpage+0x10/0x10
[ 5231.981508][   C17]  ? local_clock_noinstr+0xd/0x100
[ 5231.981511][   C17]  ? __perf_event_account_interrupt+0x11c/0x540
[ 5231.981514][   C17]  intel_pmu_lbr_save_brstack+0xc0/0x4c0
[ 5231.981518][   C17]  setup_arch_pebs_sample_data+0x114b/0x2400
[ 5231.981522][   C17]  ? __pfx_x86_perf_event_set_period+0x10/0x10
[ 5231.981526][   C17]  intel_pmu_drain_arch_pebs+0x64d/0xcc0
[ 5231.981530][   C17]  ? __pfx_intel_pmu_drain_arch_pebs+0x10/0x10
[ 5231.981534][   C17]  ? unwind_next_frame+0x11c5/0x1df0
[ 5231.981541][   C17]  ? intel_pmu_drain_bts_buffer+0xbf/0x6e0
[ 5231.981545][   C17]  ? __pfx_intel_pmu_drain_bts_buffer+0x10/0x10
[ 5231.981550][   C17]  handle_pmi_common+0x5c5/0xcb0
[ 5231.981553][   C17]  ? __pfx_handle_pmi_common+0x10/0x10
[ 5231.981556][   C17]  ? intel_idle+0x64/0xb0
[ 5231.981560][   C17]  ? intel_bts_interrupt+0xe5/0x4c0
[ 5231.981562][   C17]  ? __pfx_intel_bts_interrupt+0x10/0x10
[ 5231.981565][   C17]  ? intel_pmu_lbr_filter+0x27f/0x910
[ 5231.981568][   C17]  intel_pmu_handle_irq+0x2ed/0x600
[ 5231.981571][   C17]  perf_event_nmi_handler+0x219/0x280
[ 5231.981575][   C17]  ? __pfx_perf_event_nmi_handler+0x10/0x10
Quoted 8 lines
Quoted 8 lines
[ 5231.981579][   C17]  ? unwind_next_frame+0x11c5/0x1df0
[ 5231.981582][   C17]  nmi_handle.part.0+0x11b/0x3a0
[ 5231.981585][   C17]  ? unwind_next_frame+0x11c5/0x1df0
[ 5231.981588][   C17]  default_do_nmi+0x6b/0x180
[ 5231.981591][   C17]  fred_exc_nmi+0x3e/0x80
[ 5231.981594][   C17]  asm_fred_entrypoint_kernel+0x41/0x60
[ 5231.981596][   C17] RIP: 0010:unwind_next_frame+0x11c5/0x1df0

......

That trace could be reduced to:

Quoted 8 lines
  UBSAN: shift-out-of-bounds in /kbuild/src/consumer/arch/x86/events/intel/lbr.c:970:15
  shift exponent 64 is too large for 64-bit type 'long long unsigned int'
  ......
  intel_pmu_lbr_counters_reorder.isra.0.cold+0x2a/0xa7
  intel_pmu_lbr_save_brstack+0xc0/0x4c0
  setup_arch_pebs_sample_data+0x114b/0x2400

Without loosing anything valuable.

Sure.

Quoted 7 lines
Quoted 5 lines
Quoted 5 lines

The warning occurs because the second "instructions:p" event, which involves branch counters sampling, is incorrectly programmed to fixed counter 0 instead of the general-purpose (GP) counters 0-3 that support

So here you have 0-3, the normal 'range' notation, but then you go all funny and use ~ instead:

😂

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines

branch counters sampling. Currently only GP counters 0~3 support branch
counters sampling on CWF, any event involving branch counters sampling
should be programed on GP counters 0~3.
Since the counter index of fixed> counter 0 is 32, it leads to the "src" value in below code is right shifted 64 bits and trigger the "shift-out-of-bounds" warning.

Quoted 7 lines
Quoted 6 lines
Quoted 6 lines

cnt = (src >> (order[j] * LBR_INFO_BR_CNTR_BITS)) & LBR_INFO_BR_CNTR_MASK;

The root cause is the loss of the branch counters constraint for the last event in the branch counters sampling event group. This results in the second "instructions:p" event being programmed on fixed counter 0 incorrectly instead of the appropriate GP counters 0~3.

s/0~3/0-3/ ?

Sure.

Quoted 1 line
maintainer
#5 lore raw

Re: [RESEND Patch 2/2] perf/x86/intel: Add missing branch counters constraint apply

In reply to Dapeng Mi
Quoted 27 lines · Dapeng Mi
arch/x86/events/intel/core.c +13 -0
index 4768236c054b..4b042d71104f 100644
@@ -4628,6 +4628,19 @@ static inline void intel_pmu_set_acr_caused_constr(struct perf_event *event,
 		event->hw.dyn_constraint &= hybrid(event->pmu, acr_cause_mask64);
 }
 
+static inline int intel_set_branch_counter_constr(struct perf_event *event,
+						  int *num)
+{
+	if (branch_sample_call_stack(event))
+		return -EINVAL;
+	if (branch_sample_counters(event)) {
+		(*num)++;
+		event->hw.dyn_constraint &= x86_pmu.lbr_counters;
+	}
+
+	return 0;
+}
+
 static int intel_pmu_hw_config(struct perf_event *event)
 {
 	int ret = x86_pmu_hw_config(event);
@@ -4698,21 +4711,18 @@ static int intel_pmu_hw_config(struct perf_event *event)
 		 * group, which requires the extra space to store the counters.
 		 */

arch/x86/events/intel/core.c static int intel_pmu_hw_config(struct perf_event *event) · line 4722

Quoted 10 lines
arch/x86/events/intel/core.c +5 -0
 		leader = event->group_leader;
+		if (intel_set_branch_counter_constr(leader, &num))
 			return -EINVAL;
 		leader->hw.flags |= PERF_X86_EVENT_BRANCH_COUNTERS;
 
 		for_each_sibling_event(sibling, leader) {
+			if (intel_set_branch_counter_constr(sibling, &num))
+				return -EINVAL;
+		}
+

Do the new bit is this, right?

arch/x86/events/intel/core.c static int intel_pmu_hw_config(struct perf_event *event) · line 4726

Quoted 4 lines
arch/x86/events/intel/core.c +2 -0
+		if (event != leader) {
+			if (intel_set_branch_counter_constr(event, &num))
 				return -EINVAL;
 		}

The point being that for_each_sibling_event() will not have iterated the event because its not on the list yet?

That wasn't really clear from the changelog and I think that deserves a comment as well.

Let me go fix that.

author
#6 lore raw

Re: [RESEND Patch 2/2] perf/x86/intel: Add missing branch counters constraint apply

Quoted 33 lines · Peter Zijlstra
Quoted 32 lines · Dapeng Mi
arch/x86/events/intel/core.c +14 -0
index 4768236c054b..4b042d71104f 100644
@@ -4628,6 +4628,19 @@ static inline void intel_pmu_set_acr_caused_constr(struct perf_event *event,
 		event->hw.dyn_constraint &= hybrid(event->pmu, acr_cause_mask64);
 }
 
+static inline int intel_set_branch_counter_constr(struct perf_event *event,
+						  int *num)
+{
+	if (branch_sample_call_stack(event))
+		return -EINVAL;
+	if (branch_sample_counters(event)) {
+		(*num)++;
+		event->hw.dyn_constraint &= x86_pmu.lbr_counters;
+	}
+
+	return 0;
+}
+
 static int intel_pmu_hw_config(struct perf_event *event)
 {
 	int ret = x86_pmu_hw_config(event);
@@ -4698,21 +4711,18 @@ static int intel_pmu_hw_config(struct perf_event *event)
 		 * group, which requires the extra space to store the counters.
 		 */
 		leader = event->group_leader;
+		if (intel_set_branch_counter_constr(leader, &num))
 			return -EINVAL;
 		leader->hw.flags |= PERF_X86_EVENT_BRANCH_COUNTERS;
 
Quoted 6 lines

arch/x86/events/intel/core.c static int intel_pmu_hw_config(struct perf_event *event) · line 4722

Quoted 5 lines
arch/x86/events/intel/core.c +4 -0
 		for_each_sibling_event(sibling, leader) {
+			if (intel_set_branch_counter_constr(sibling, &num))
+				return -EINVAL;
+		}
+

Do the new bit is this, right?

Actually not, the key change is the below one. The last event in the group is not applied the branch counter constraint.

Assume we have a event group {cycles,instructions,branches}. When the 3rd event "branches" is created and the function intel_pmu_hw_config() is called for the "branches" event to check the config.  The event leader is "cycles" and the sibling event has only the "instructions" event at that time since the 3rd event "branches" is in creation and still not added into the sibling_list. So for_each_sibling_event() can't really iterate the "branches" event.

Quoted 7 lines

arch/x86/events/intel/core.c static int intel_pmu_hw_config(struct perf_event *event) · line 4726

Quoted 5 lines
arch/x86/events/intel/core.c +2 -0
+		if (event != leader) {
+			if (intel_set_branch_counter_constr(event, &num))
 				return -EINVAL;
 		}

The point being that for_each_sibling_event() will not have iterated the event because its not on the list yet?

Yes.

Quoted 3 lines

That wasn't really clear from the changelog and I think that deserves a comment as well.

Sure. I would add comment and enhance the changelog to make it clearer. Thanks.

Quoted 2 lines

Let me go fix that.

maintainer
#7 lore raw

Re: [RESEND Patch 2/2] perf/x86/intel: Add missing branch counters constraint apply

Quoted 62 lines · Dapeng
Quoted 33 lines · Peter Zijlstra
Quoted 32 lines · Dapeng Mi
arch/x86/events/intel/core.c +14 -0
index 4768236c054b..4b042d71104f 100644
@@ -4628,6 +4628,19 @@ static inline void intel_pmu_set_acr_caused_constr(struct perf_event *event,
 		event->hw.dyn_constraint &= hybrid(event->pmu, acr_cause_mask64);
 }
 
+static inline int intel_set_branch_counter_constr(struct perf_event *event,
+						  int *num)
+{
+	if (branch_sample_call_stack(event))
+		return -EINVAL;
+	if (branch_sample_counters(event)) {
+		(*num)++;
+		event->hw.dyn_constraint &= x86_pmu.lbr_counters;
+	}
+
+	return 0;
+}
+
 static int intel_pmu_hw_config(struct perf_event *event)
 {
 	int ret = x86_pmu_hw_config(event);
@@ -4698,21 +4711,18 @@ static int intel_pmu_hw_config(struct perf_event *event)
 		 * group, which requires the extra space to store the counters.
 		 */
 		leader = event->group_leader;
+		if (intel_set_branch_counter_constr(leader, &num))
 			return -EINVAL;
 		leader->hw.flags |= PERF_X86_EVENT_BRANCH_COUNTERS;
 
Quoted 6 lines

arch/x86/events/intel/core.c static int intel_pmu_hw_config(struct perf_event *event) · line 4722

Quoted 5 lines
arch/x86/events/intel/core.c +4 -0
 		for_each_sibling_event(sibling, leader) {
+			if (intel_set_branch_counter_constr(sibling, &num))
+				return -EINVAL;
+		}
+

Do the new bit is this, right?

Actually not, the key change is the below one. The last event in the group is not applied the branch counter constraint.

Assume we have a event group {cycles,instructions,branches}. When the 3rd event "branches" is created and the function intel_pmu_hw_config() is called for the "branches" event to check the config.  The event leader is "cycles" and the sibling event has only the "instructions" event at that time since the 3rd event "branches" is in creation and still not added into the sibling_list. So for_each_sibling_event() can't really iterate the "branches" event.

Quoted 7 lines

arch/x86/events/intel/core.c static int intel_pmu_hw_config(struct perf_event *event) · line 4726

Quoted 5 lines
arch/x86/events/intel/core.c +2 -0
+		if (event != leader) {
+			if (intel_set_branch_counter_constr(event, &num))
 				return -EINVAL;
 		}

The point being that for_each_sibling_event() will not have iterated the event because its not on the list yet?

Quoted 9 lines

Yes.

Quoted 3 lines

That wasn't really clear from the changelog and I think that deserves a comment as well.

Sure. I would add comment and enhance the changelog to make it clearer. Thanks.

I already fixed everything up. Should be in queue/perf/urgent.

author
#8 lore raw

Re: [RESEND Patch 2/2] perf/x86/intel: Add missing branch counters constraint apply

Quoted 59 lines · Peter Zijlstra
Quoted 58 lines · Dapeng
Quoted 33 lines · Peter Zijlstra
Quoted 32 lines · Dapeng Mi
arch/x86/events/intel/core.c +14 -0
index 4768236c054b..4b042d71104f 100644
@@ -4628,6 +4628,19 @@ static inline void intel_pmu_set_acr_caused_constr(struct perf_event *event,
 		event->hw.dyn_constraint &= hybrid(event->pmu, acr_cause_mask64);
 }
 
+static inline int intel_set_branch_counter_constr(struct perf_event *event,
+						  int *num)
+{
+	if (branch_sample_call_stack(event))
+		return -EINVAL;
+	if (branch_sample_counters(event)) {
+		(*num)++;
+		event->hw.dyn_constraint &= x86_pmu.lbr_counters;
+	}
+
+	return 0;
+}
+
 static int intel_pmu_hw_config(struct perf_event *event)
 {
 	int ret = x86_pmu_hw_config(event);
@@ -4698,21 +4711,18 @@ static int intel_pmu_hw_config(struct perf_event *event)
 		 * group, which requires the extra space to store the counters.
 		 */
 		leader = event->group_leader;
+		if (intel_set_branch_counter_constr(leader, &num))
 			return -EINVAL;
 		leader->hw.flags |= PERF_X86_EVENT_BRANCH_COUNTERS;
 
Quoted 6 lines

arch/x86/events/intel/core.c static int intel_pmu_hw_config(struct perf_event *event) · line 4722

Quoted 5 lines
arch/x86/events/intel/core.c +4 -0
 		for_each_sibling_event(sibling, leader) {
+			if (intel_set_branch_counter_constr(sibling, &num))
+				return -EINVAL;
+		}
+

Do the new bit is this, right?

Actually not, the key change is the below one. The last event in the group is not applied the branch counter constraint.

Assume we have a event group {cycles,instructions,branches}. When the 3rd event "branches" is created and the function intel_pmu_hw_config() is called for the "branches" event to check the config.  The event leader is "cycles" and the sibling event has only the "instructions" event at that time since the 3rd event "branches" is in creation and still not added into the sibling_list. So for_each_sibling_event() can't really iterate the "branches" event.

Quoted 6 lines

arch/x86/events/intel/core.c static int intel_pmu_hw_config(struct perf_event *event) · line 4726

Quoted 4 lines
arch/x86/events/intel/core.c +2 -0
+		if (event != leader) {
+			if (intel_set_branch_counter_constr(event, &num))
 				return -EINVAL;
 		}

The point being that for_each_sibling_event() will not have iterated the event because its not on the list yet?

Quoted 8 lines
Quoted 7 lines

Yes.

Quoted 2 lines

That wasn't really clear from the changelog and I think that deserves a comment as well.

Sure. I would add comment and enhance the changelog to make it clearer. Thanks.

I already fixed everything up. Should be in queue/perf/urgent.

Thanks.

Peter, As Ian points out, the patch "perf/x86: Update cap_user_rdpmc base on rdpmc user disable state" has a bug
(https://lore.kernel.org/all/CAP-5=fWr2L6miiFZ6Km3HYEdmqp3T0NBL=WY3buKdKztW+HvmA@mail.gmail.com/).
I would posted a patch to fix the issue. Thanks.

Quoted 1 line
merge bot Applied
#10 lore raw

[tip: perf/urgent] perf/x86/intel: Add missing branch counters constraint apply

In reply to Dapeng Mi
The following commit has been merged into the perf/urgent branch of tip:

The following commit has been merged into the perf/urgent branch of tip:

Commit-ID:     1d07bbd7ea36ea0b8dfa8068dbe67eb3a32d9590
Gitweb:        https://git.kernel.org/tip/1d07bbd7ea36ea0b8dfa8068dbe67eb3a32d9590
Author:        Dapeng Mi <[email protected]>
AuthorDate:    Sat, 28 Feb 2026 13:33:20 +08:00
Committer:     Peter Zijlstra <[email protected]>
CommitterDate: Thu, 12 Mar 2026 11:29:16 +01:00

perf/x86/intel: Add missing branch counters constraint apply

When running the command:
'perf record -e "{instructions,instructions:p}" -j any,counter sleep 1', a "shift-out-of-bounds" warning is reported on CWF.

  UBSAN: shift-out-of-bounds in /kbuild/src/consumer/arch/x86/events/intel/lbr.c:970:15
  shift exponent 64 is too large for 64-bit type 'long long unsigned int'
  ......
  intel_pmu_lbr_counters_reorder.isra.0.cold+0x2a/0xa7
  intel_pmu_lbr_save_brstack+0xc0/0x4c0
  setup_arch_pebs_sample_data+0x114b/0x2400

The warning occurs because the second "instructions:p" event, which involves branch counters sampling, is incorrectly programmed to fixed counter 0 instead of the general-purpose (GP) counters 0-3 that support branch counters sampling. Currently only GP counters 0-3 support branch counters sampling on CWF, any event involving branch counters sampling should be programed on GP counters 0-3. Since the counter index of fixed counter 0 is 32, it leads to the "src" value in below code is right shifted 64 bits and trigger the "shift-out-of-bounds" warning.

cnt = (src >> (order[j] * LBR_INFO_BR_CNTR_BITS)) & LBR_INFO_BR_CNTR_MASK;

The root cause is the loss of the branch counters constraint for the new event in the branch counters sampling event group. Since it isn't yet part of the sibling list. This results in the second "instructions:p" event being programmed on fixed counter 0 incorrectly instead of the appropriate GP counters 0-3.

To address this, we apply the missing branch counters constraint for the last event in the group. Additionally, we introduce a new function, intel_set_branch_counter_constr(), to apply the branch counters constraint and avoid code duplication.

Fixes 33744916196b ("perf/x86/intel: Support branch counters logging")
---
 arch/x86/events/intel/core.c | 31 +++++++++++++++++++++----------
 1 file changed, 21 insertions(+), 10 deletions(-)
arch/x86/events/intel/core.c +21 -10
index cf3a4fe..36c6821 100644
@@ -4628,6 +4628,19 @@ static inline void intel_pmu_set_acr_caused_constr(struct perf_event *event,
 		event->hw.dyn_constraint &= hybrid(event->pmu, acr_cause_mask64);
 }
 
+static inline int intel_set_branch_counter_constr(struct perf_event *event,
+						  int *num)
+{
+	if (branch_sample_call_stack(event))
+		return -EINVAL;
+	if (branch_sample_counters(event)) {
+		(*num)++;
+		event->hw.dyn_constraint &= x86_pmu.lbr_counters;
+	}
+
+	return 0;
+}
+
 static int intel_pmu_hw_config(struct perf_event *event)
 {
 	int ret = x86_pmu_hw_config(event);
@@ -4698,21 +4711,19 @@ static int intel_pmu_hw_config(struct perf_event *event)
 		 * group, which requires the extra space to store the counters.
 		 */
 		leader = event->group_leader;
-		if (branch_sample_call_stack(leader))
+		if (intel_set_branch_counter_constr(leader, &num))
 			return -EINVAL;
-		if (branch_sample_counters(leader)) {
-			num++;
-			leader->hw.dyn_constraint &= x86_pmu.lbr_counters;
-		}
 		leader->hw.flags |= PERF_X86_EVENT_BRANCH_COUNTERS;
 
 		for_each_sibling_event(sibling, leader) {
-			if (branch_sample_call_stack(sibling))
+			if (intel_set_branch_counter_constr(sibling, &num))
+				return -EINVAL;
+		}
+
+		/* event isn't installed as a sibling yet. */
+		if (event != leader) {
+			if (intel_set_branch_counter_constr(event, &num))
 				return -EINVAL;
-			if (branch_sample_counters(sibling)) {
-				num++;
-				sibling->hw.dyn_constraint &= x86_pmu.lbr_counters;
-			}
 		}
 
 		if (num > fls(x86_pmu.lbr_counters))
Reviewed-by on v1 1/2
#11 lore raw
In reply to Dapeng Mi
Quoted 8 lines · Dapeng Mi

It's good enough to only check GP counters for PEBS constraints validation since constraints overlap can only happen on GP counters.

Besides opportunistically refine the code style and use pr_warn() to replace pr_info() as the message itself is a warning message.

---

LGTM.

Quoted 59 lines
 arch/x86/events/intel/core.c | 22 ++++++++++++++--------
 1 file changed, 14 insertions(+), 8 deletions(-)
arch/x86/events/intel/core.c +14 -8
index cf3a4fe06ff2..4768236c054b 100644
@@ -5770,7 +5770,7 @@ static void __intel_pmu_check_dyn_constr(struct event_constraint *constr,
 			}
 
 			if (check_fail) {
-				pr_info("The two events 0x%llx and 0x%llx may not be "
+				pr_warn("The two events 0x%llx and 0x%llx may not be "
 					"fully scheduled under some circumstances as "
 					"%s.\n",
 					c1->code, c2->code, dyn_constr_type_name[type]);
@@ -5783,6 +5783,7 @@ static void intel_pmu_check_dyn_constr(struct pmu *pmu,
 				       struct event_constraint *constr,
 				       u64 cntr_mask)
 {
+	u64 gp_mask = GENMASK_ULL(INTEL_PMC_MAX_GENERIC - 1, 0);
 	enum dyn_constr_type i;
 	u64 mask;
 
@@ -5797,20 +5798,25 @@ static void intel_pmu_check_dyn_constr(struct pmu *pmu,
 				mask = x86_pmu.lbr_counters;
 			break;
 		case DYN_CONSTR_ACR_CNTR:
-			mask = hybrid(pmu, acr_cntr_mask64) & GENMASK_ULL(INTEL_PMC_MAX_GENERIC - 1, 0);
+			mask = hybrid(pmu, acr_cntr_mask64) & gp_mask;
 			break;
 		case DYN_CONSTR_ACR_CAUSE:
-			if (hybrid(pmu, acr_cntr_mask64) == hybrid(pmu, acr_cause_mask64))
+			if (hybrid(pmu, acr_cntr_mask64) ==
+					hybrid(pmu, acr_cause_mask64))
 				continue;
-			mask = hybrid(pmu, acr_cause_mask64) & GENMASK_ULL(INTEL_PMC_MAX_GENERIC - 1, 0);
+			mask = hybrid(pmu, acr_cause_mask64) & gp_mask;
 			break;
 		case DYN_CONSTR_PEBS:
-			if (x86_pmu.arch_pebs)
-				mask = hybrid(pmu, arch_pebs_cap).counters;
+			if (x86_pmu.arch_pebs) {
+				mask = hybrid(pmu, arch_pebs_cap).counters &
+				       gp_mask;
+			}
 			break;
 		case DYN_CONSTR_PDIST:
-			if (x86_pmu.arch_pebs)
-				mask = hybrid(pmu, arch_pebs_cap).pdists;
+			if (x86_pmu.arch_pebs) {
+				mask = hybrid(pmu, arch_pebs_cap).pdists &
+				       gp_mask;
+			}
 			break;
 		default:
 			pr_warn("Unsupported dynamic constraint type %d\n", i);

base-commit: 6de23f81a5e08be8fbf5e8d7e9febc72a5b5f27f

author
#12 lore raw
In reply to Dapeng Mi

Hi Peter,

Could you please review and merge this change? This change suppose to be low risky. Without this change, the dynamic constraints check would trigger false positive on DMR/NVL. Thanks.

Quoted 67 lines · Dapeng Mi

It's good enough to only check GP counters for PEBS constraints validation since constraints overlap can only happen on GP counters.

Besides opportunistically refine the code style and use pr_warn() to replace pr_info() as the message itself is a warning message.

---
 arch/x86/events/intel/core.c | 22 ++++++++++++++--------
 1 file changed, 14 insertions(+), 8 deletions(-)
arch/x86/events/intel/core.c +14 -8
index cf3a4fe06ff2..4768236c054b 100644
@@ -5770,7 +5770,7 @@ static void __intel_pmu_check_dyn_constr(struct event_constraint *constr,
 			}
 
 			if (check_fail) {
-				pr_info("The two events 0x%llx and 0x%llx may not be "
+				pr_warn("The two events 0x%llx and 0x%llx may not be "
 					"fully scheduled under some circumstances as "
 					"%s.\n",
 					c1->code, c2->code, dyn_constr_type_name[type]);
@@ -5783,6 +5783,7 @@ static void intel_pmu_check_dyn_constr(struct pmu *pmu,
 				       struct event_constraint *constr,
 				       u64 cntr_mask)
 {
+	u64 gp_mask = GENMASK_ULL(INTEL_PMC_MAX_GENERIC - 1, 0);
 	enum dyn_constr_type i;
 	u64 mask;
 
@@ -5797,20 +5798,25 @@ static void intel_pmu_check_dyn_constr(struct pmu *pmu,
 				mask = x86_pmu.lbr_counters;
 			break;
 		case DYN_CONSTR_ACR_CNTR:
-			mask = hybrid(pmu, acr_cntr_mask64) & GENMASK_ULL(INTEL_PMC_MAX_GENERIC - 1, 0);
+			mask = hybrid(pmu, acr_cntr_mask64) & gp_mask;
 			break;
 		case DYN_CONSTR_ACR_CAUSE:
-			if (hybrid(pmu, acr_cntr_mask64) == hybrid(pmu, acr_cause_mask64))
+			if (hybrid(pmu, acr_cntr_mask64) ==
+					hybrid(pmu, acr_cause_mask64))
 				continue;
-			mask = hybrid(pmu, acr_cause_mask64) & GENMASK_ULL(INTEL_PMC_MAX_GENERIC - 1, 0);
+			mask = hybrid(pmu, acr_cause_mask64) & gp_mask;
 			break;
 		case DYN_CONSTR_PEBS:
-			if (x86_pmu.arch_pebs)
-				mask = hybrid(pmu, arch_pebs_cap).counters;
+			if (x86_pmu.arch_pebs) {
+				mask = hybrid(pmu, arch_pebs_cap).counters &
+				       gp_mask;
+			}
 			break;
 		case DYN_CONSTR_PDIST:
-			if (x86_pmu.arch_pebs)
-				mask = hybrid(pmu, arch_pebs_cap).pdists;
+			if (x86_pmu.arch_pebs) {
+				mask = hybrid(pmu, arch_pebs_cap).pdists &
+				       gp_mask;
+			}
 			break;
 		default:
 			pr_warn("Unsupported dynamic constraint type %d\n", i);

base-commit: 6de23f81a5e08be8fbf5e8d7e9febc72a5b5f27f

maintainer
#13 lore raw
Quoted 5 lines · Dapeng

Hi Peter,

Could you please review and merge this change? This change suppose to be low risky. Without this change, the dynamic constraints check would trigger false positive on DMR/NVL. Thanks.

Ah, so I have it in queue/perf/core, it didn't have no Fixes on so I didn't think it urgent.

author
#14 lore raw
Quoted 8 lines · Peter Zijlstra
Quoted 5 lines · Dapeng

Hi Peter,

Could you please review and merge this change? This change suppose to be low risky. Without this change, the dynamic constraints check would trigger false positive on DMR/NVL. Thanks.

Ah, so I have it in queue/perf/core, it didn't have no Fixes on so I didn't think it urgent.

Oh, just saw it. Thanks a lot! 😊

Quoted 1 line
merge bot Applied
#15 lore raw
In reply to Dapeng Mi
The following commit has been merged into the perf/core branch of tip:

The following commit has been merged into the perf/core branch of tip:

Commit-ID:     b191aa32be2c960ca2391b40e862c389230cadef
Gitweb:        https://git.kernel.org/tip/b191aa32be2c960ca2391b40e862c389230cadef
Author:        Dapeng Mi <[email protected]>
AuthorDate:    Sat, 28 Feb 2026 13:33:19 +08:00
Committer:     Peter Zijlstra <[email protected]>
CommitterDate: Thu, 12 Mar 2026 11:29:17 +01:00

perf/x86/intel: Only check GP counters for PEBS constraints validation

It's good enough to only check GP counters for PEBS constraints validation since constraints overlap can only happen on GP counters.

Besides opportunistically refine the code style and use pr_warn() to replace pr_info() as the message itself is a warning message.

---
 arch/x86/events/intel/core.c | 22 ++++++++++++++--------
 1 file changed, 14 insertions(+), 8 deletions(-)
arch/x86/events/intel/core.c +14 -8
index cf3a4fe..4768236 100644
@@ -5770,7 +5770,7 @@ static void __intel_pmu_check_dyn_constr(struct event_constraint *constr,
 			}
 
 			if (check_fail) {
-				pr_info("The two events 0x%llx and 0x%llx may not be "
+				pr_warn("The two events 0x%llx and 0x%llx may not be "
 					"fully scheduled under some circumstances as "
 					"%s.\n",
 					c1->code, c2->code, dyn_constr_type_name[type]);
@@ -5783,6 +5783,7 @@ static void intel_pmu_check_dyn_constr(struct pmu *pmu,
 				       struct event_constraint *constr,
 				       u64 cntr_mask)
 {
+	u64 gp_mask = GENMASK_ULL(INTEL_PMC_MAX_GENERIC - 1, 0);
 	enum dyn_constr_type i;
 	u64 mask;
 
@@ -5797,20 +5798,25 @@ static void intel_pmu_check_dyn_constr(struct pmu *pmu,
 				mask = x86_pmu.lbr_counters;
 			break;
 		case DYN_CONSTR_ACR_CNTR:
-			mask = hybrid(pmu, acr_cntr_mask64) & GENMASK_ULL(INTEL_PMC_MAX_GENERIC - 1, 0);
+			mask = hybrid(pmu, acr_cntr_mask64) & gp_mask;
 			break;
 		case DYN_CONSTR_ACR_CAUSE:
-			if (hybrid(pmu, acr_cntr_mask64) == hybrid(pmu, acr_cause_mask64))
+			if (hybrid(pmu, acr_cntr_mask64) ==
+					hybrid(pmu, acr_cause_mask64))
 				continue;
-			mask = hybrid(pmu, acr_cause_mask64) & GENMASK_ULL(INTEL_PMC_MAX_GENERIC - 1, 0);
+			mask = hybrid(pmu, acr_cause_mask64) & gp_mask;
 			break;
 		case DYN_CONSTR_PEBS:
-			if (x86_pmu.arch_pebs)
-				mask = hybrid(pmu, arch_pebs_cap).counters;
+			if (x86_pmu.arch_pebs) {
+				mask = hybrid(pmu, arch_pebs_cap).counters &
+				       gp_mask;
+			}
 			break;
 		case DYN_CONSTR_PDIST:
-			if (x86_pmu.arch_pebs)
-				mask = hybrid(pmu, arch_pebs_cap).pdists;
+			if (x86_pmu.arch_pebs) {
+				mask = hybrid(pmu, arch_pebs_cap).pdists &
+				       gp_mask;
+			}
 			break;
 		default:
 			pr_warn("Unsupported dynamic constraint type %d\n", i);