[PATCH v3 3/5] kfence: move saving stack trace of allocations into __kfence_alloc()

[Marco Elver]

Move the saving of the stack trace of allocations into __kfence_alloc(),
so that the stack entries array can be used outside of
kfence_guarded_alloc() and we avoid potentially unwinding the stack
multiple times.

Signed-off-by: Marco Elver <elver@xxxxxxxxxx>
Reviewed-by: Dmitry Vyukov <dvyukov@xxxxxxxxxx>
---
v2:
* New patch.
---
 mm/kfence/core.c | 35 ++++++++++++++++++++++++-----------
 1 file changed, 24 insertions(+), 11 deletions(-)

diff --git a/mm/kfence/core.c b/mm/kfence/core.c
index 249d75b7e5ee..db01814f8ff0 100644
--- a/mm/kfence/core.c
+++ b/mm/kfence/core.c
@@ -187,19 +187,26 @@ static inline unsigned long metadata_to_pageaddr(const struct kfence_metadata *m
  * Update the object's metadata state, including updating the alloc/free stacks
  * depending on the state transition.
  */
-static noinline void metadata_update_state(struct kfence_metadata *meta,
-					   enum kfence_object_state next)
+static noinline void
+metadata_update_state(struct kfence_metadata *meta, enum kfence_object_state next,
+		      unsigned long *stack_entries, size_t num_stack_entries)
 {
 	struct kfence_track *track =
 		next == KFENCE_OBJECT_FREED ? &meta->free_track : &meta->alloc_track;
 
 	lockdep_assert_held(&meta->lock);
 
-	/*
-	 * Skip over 1 (this) functions; noinline ensures we do not accidentally
-	 * skip over the caller by never inlining.
-	 */
-	track->num_stack_entries = stack_trace_save(track->stack_entries, KFENCE_STACK_DEPTH, 1);
+	if (stack_entries) {
+		memcpy(track->stack_entries, stack_entries,
+		       num_stack_entries * sizeof(stack_entries[0]));
+	} else {
+		/*
+		 * Skip over 1 (this) functions; noinline ensures we do not
+		 * accidentally skip over the caller by never inlining.
+		 */
+		num_stack_entries = stack_trace_save(track->stack_entries, KFENCE_STACK_DEPTH, 1);
+	}
+	track->num_stack_entries = num_stack_entries;
 	track->pid = task_pid_nr(current);
 	track->cpu = raw_smp_processor_id();
 	track->ts_nsec = local_clock(); /* Same source as printk timestamps. */
@@ -261,7 +268,8 @@ static __always_inline void for_each_canary(const struct kfence_metadata *meta,
 	}
 }
 
-static void *kfence_guarded_alloc(struct kmem_cache *cache, size_t size, gfp_t gfp)
+static void *kfence_guarded_alloc(struct kmem_cache *cache, size_t size, gfp_t gfp,
+				  unsigned long *stack_entries, size_t num_stack_entries)
 {
 	struct kfence_metadata *meta = NULL;
 	unsigned long flags;
@@ -320,7 +328,7 @@ static void *kfence_guarded_alloc(struct kmem_cache *cache, size_t size, gfp_t g
 	addr = (void *)meta->addr;
 
 	/* Update remaining metadata. */
-	metadata_update_state(meta, KFENCE_OBJECT_ALLOCATED);
+	metadata_update_state(meta, KFENCE_OBJECT_ALLOCATED, stack_entries, num_stack_entries);
 	/* Pairs with READ_ONCE() in kfence_shutdown_cache(). */
 	WRITE_ONCE(meta->cache, cache);
 	meta->size = size;
@@ -400,7 +408,7 @@ static void kfence_guarded_free(void *addr, struct kfence_metadata *meta, bool z
 		memzero_explicit(addr, meta->size);
 
 	/* Mark the object as freed. */
-	metadata_update_state(meta, KFENCE_OBJECT_FREED);
+	metadata_update_state(meta, KFENCE_OBJECT_FREED, NULL, 0);
 
 	raw_spin_unlock_irqrestore(&meta->lock, flags);
 
@@ -742,6 +750,9 @@ void kfence_shutdown_cache(struct kmem_cache *s)
 
 void *__kfence_alloc(struct kmem_cache *s, size_t size, gfp_t flags)
 {
+	unsigned long stack_entries[KFENCE_STACK_DEPTH];
+	size_t num_stack_entries;
+
 	/*
 	 * Perform size check before switching kfence_allocation_gate, so that
 	 * we don't disable KFENCE without making an allocation.
@@ -786,7 +797,9 @@ void *__kfence_alloc(struct kmem_cache *s, size_t size, gfp_t flags)
 	if (!READ_ONCE(kfence_enabled))
 		return NULL;
 
-	return kfence_guarded_alloc(s, size, flags);
+	num_stack_entries = stack_trace_save(stack_entries, KFENCE_STACK_DEPTH, 0);
+
+	return kfence_guarded_alloc(s, size, flags, stack_entries, num_stack_entries);
 }
 
 size_t kfence_ksize(const void *addr)
-- 
2.33.0.464.g1972c5931b-goog