Merge tag 'perf-core-2026-04-13' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull performance events updates from Ingo Molnar:
 "Core updates:

   - Try to allocate task_ctx_data quickly, to optimize O(N^2) algorithm
     on large systems with O(100k) threads (Namhyung Kim)

  AMD PMU driver IBS support updates and fixes, by Ravi Bangoria:
   - Fix interrupt accounting for discarded samples
   - Fix a Zen5-specific quirk
   - Fix PhyAddrVal handling
   - Fix NMI-safety with perf_allow_kernel()
   - Fix a race between event add and NMIs

  Intel PMU driver updates:
   - Only check GP counters for PEBS constraints validation (Dapeng Mi)

  MSR driver:
   - Turn SMI_COUNT and PPERF on by default, instead of a long list of
     CPU models to enable them on (Kan Liang)

  ... and misc cleanups and fixes by Aldf Conte, Anshuman Khandual,
  Namhyung Kim, Ravi Bangoria and Yen-Hsiang Hsu"

* tag 'perf-core-2026-04-13' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf/events: Replace READ_ONCE() with standard pgtable accessors
  perf/x86/msr: Make SMI and PPERF on by default
  perf/x86/intel/p4: Fix unused variable warning in p4_pmu_init()
  perf/x86/intel: Only check GP counters for PEBS constraints validation
  perf/x86/amd/ibs: Fix comment typo in ibs_op_data
  perf/amd/ibs: Advertise remote socket capability
  perf/amd/ibs: Enable streaming store filter
  perf/amd/ibs: Enable RIP bit63 hardware filtering
  perf/amd/ibs: Enable fetch latency filtering
  perf/amd/ibs: Support IBS_{FETCH|OP}_CTL2[Dis] to eliminate RMW race
  perf/amd/ibs: Add new MSRs and CPUID bits definitions
  perf/amd/ibs: Define macro for ldlat mask and shift
  perf/amd/ibs: Avoid race between event add and NMI
  perf/amd/ibs: Avoid calling perf_allow_kernel() from the IBS NMI handler
  perf/amd/ibs: Preserve PhyAddrVal bit when clearing PhyAddr MSR
  perf/amd/ibs: Limit ldlat->l3missonly dependency to Zen5
  perf/amd/ibs: Account interrupt for discarded samples
  perf/core: Simplify __detach_global_ctx_data()
  perf/core: Try to allocate task_ctx_data quickly
  perf/core: Pass GFP flags to attach_task_ctx_data()

1 files changed, 21 insertions, 22 deletions

diff --git a/kernel/events/core.c b/kernel/events/core.c
index 89b40e439717..9e66fbee37f0 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -5368,15 +5368,15 @@ static void unaccount_freq_event(void)
 
 
 static struct perf_ctx_data *
-alloc_perf_ctx_data(struct kmem_cache *ctx_cache, bool global)
+alloc_perf_ctx_data(struct kmem_cache *ctx_cache, bool global, gfp_t gfp_flags)
 {
 	struct perf_ctx_data *cd;
 
-	cd = kzalloc_obj(*cd);
+	cd = kzalloc_obj(*cd, gfp_flags);
 	if (!cd)
 		return NULL;
 
-	cd->data = kmem_cache_zalloc(ctx_cache, GFP_KERNEL);
+	cd->data = kmem_cache_zalloc(ctx_cache, gfp_flags);
 	if (!cd->data) {
 		kfree(cd);
 		return NULL;
@@ -5410,11 +5410,11 @@ static inline void perf_free_ctx_data_rcu(struct perf_ctx_data *cd)
 
 static int
 attach_task_ctx_data(struct task_struct *task, struct kmem_cache *ctx_cache,
-		     bool global)
+		     bool global, gfp_t gfp_flags)
 {
 	struct perf_ctx_data *cd, *old = NULL;
 
-	cd = alloc_perf_ctx_data(ctx_cache, global);
+	cd = alloc_perf_ctx_data(ctx_cache, global, gfp_flags);
 	if (!cd)
 		return -ENOMEM;
 
@@ -5487,6 +5487,12 @@ again:
 					cd = NULL;
 			}
 			if (!cd) {
+				/*
+				 * Try to allocate context quickly before
+				 * traversing the whole thread list again.
+				 */
+				if (!attach_task_ctx_data(p, ctx_cache, true, GFP_NOWAIT))
+					continue;
 				get_task_struct(p);
 				goto alloc;
 			}
@@ -5497,7 +5503,7 @@ again:
 
 	return 0;
 alloc:
-	ret = attach_task_ctx_data(p, ctx_cache, true);
+	ret = attach_task_ctx_data(p, ctx_cache, true, GFP_KERNEL);
 	put_task_struct(p);
 	if (ret) {
 		__detach_global_ctx_data();
@@ -5517,7 +5523,7 @@ attach_perf_ctx_data(struct perf_event *event)
 		return -ENOMEM;
 
 	if (task)
-		return attach_task_ctx_data(task, ctx_cache, false);
+		return attach_task_ctx_data(task, ctx_cache, false, GFP_KERNEL);
 
 	ret = attach_global_ctx_data(ctx_cache);
 	if (ret)
@@ -5552,22 +5558,15 @@ static void __detach_global_ctx_data(void)
 	struct task_struct *g, *p;
 	struct perf_ctx_data *cd;
 
-again:
 	scoped_guard (rcu) {
 		for_each_process_thread(g, p) {
 			cd = rcu_dereference(p->perf_ctx_data);
-			if (!cd || !cd->global)
-				continue;
-			cd->global = 0;
-			get_task_struct(p);
-			goto detach;
+			if (cd && cd->global) {
+				cd->global = 0;
+				detach_task_ctx_data(p);
+			}
 		}
 	}
-	return;
-detach:
-	detach_task_ctx_data(p);
-	put_task_struct(p);
-	goto again;
 }
 
 static void detach_global_ctx_data(void)
@@ -8420,7 +8419,7 @@ static u64 perf_get_pgtable_size(struct mm_struct *mm, unsigned long addr)
 	pte_t *ptep, pte;
 
 	pgdp = pgd_offset(mm, addr);
-	pgd = READ_ONCE(*pgdp);
+	pgd = pgdp_get(pgdp);
 	if (pgd_none(pgd))
 		return 0;
 
@@ -8428,7 +8427,7 @@ static u64 perf_get_pgtable_size(struct mm_struct *mm, unsigned long addr)
 		return pgd_leaf_size(pgd);
 
 	p4dp = p4d_offset_lockless(pgdp, pgd, addr);
-	p4d = READ_ONCE(*p4dp);
+	p4d = p4dp_get(p4dp);
 	if (!p4d_present(p4d))
 		return 0;
 
@@ -8436,7 +8435,7 @@ static u64 perf_get_pgtable_size(struct mm_struct *mm, unsigned long addr)
 		return p4d_leaf_size(p4d);
 
 	pudp = pud_offset_lockless(p4dp, p4d, addr);
-	pud = READ_ONCE(*pudp);
+	pud = pudp_get(pudp);
 	if (!pud_present(pud))
 		return 0;
 
@@ -9238,7 +9237,7 @@ perf_event_alloc_task_data(struct task_struct *child,
 
 	return;
 attach:
-	attach_task_ctx_data(child, ctx_cache, true);
+	attach_task_ctx_data(child, ctx_cache, true, GFP_KERNEL);
 }
 
 void perf_event_fork(struct task_struct *task)