[PATCH v5] stackdepot: Make max number of pools boot-time configurable

[Matt Fleming <matt@readmodwrite.com>]

From: Matt Fleming <mfleming@cloudflare.com>

We're hitting the WARN in depot_init_pool() about reaching the stack
depot limit because we have long stacks that don't dedup very well.

Introduce a new start-up parameter to allow users to set the number of
maximum stack depot pools.

Acked-by: Marco Elver <elver@google.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Matt Fleming <mfleming@cloudflare.com>
---

Changes in v5:
 - Add Acked-by tags
 - Document magic number as STACK_MIN_POOLS
 - Clamp out of bounds stack_depot_max_pools values instead of using default
 - Don't blindly follow clang-format

Changes in v4:
 - Simplify initial stack_max_pools value courtesy of akpm

Changes in v3:
 - Switch from build-time to boot-time parameter

Changes in v2:
 - Replace BUILD_BUG_ON with static_assert()
 - Hide STACKDEPOT_MAX_POOLS behind EXPERT

 .../admin-guide/kernel-parameters.txt         |  5 ++
 lib/stackdepot.c                              | 64 ++++++++++++++++---
 2 files changed, 60 insertions(+), 9 deletions(-)

diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
index 07e22ba5bfe3..273f6bbd56c2 100644
--- a/Documentation/admin-guide/kernel-parameters.txt
+++ b/Documentation/admin-guide/kernel-parameters.txt
@@ -6988,6 +6988,11 @@
 			consumed by the stack hash table. By default this is set
 			to false.
 
+	stack_depot_max_pools= [KNL,EARLY]
+			Specify the maximum number of pools to use for storing
+			stack traces. Pools are allocated on-demand up to this
+			limit. Default value is 8191 pools.
+
 	stacktrace	[FTRACE]
 			Enabled the stack tracer on boot up.
 
diff --git a/lib/stackdepot.c b/lib/stackdepot.c
index 73d7b50924ef..a9dcfcdc9932 100644
--- a/lib/stackdepot.c
+++ b/lib/stackdepot.c
@@ -36,11 +36,13 @@
 #include <linux/memblock.h>
 #include <linux/kasan-enabled.h>
 
-#define DEPOT_POOLS_CAP 8192
-/* The pool_index is offset by 1 so the first record does not have a 0 handle. */
-#define DEPOT_MAX_POOLS \
-	(((1LL << (DEPOT_POOL_INDEX_BITS)) - 1 < DEPOT_POOLS_CAP) ? \
-	 (1LL << (DEPOT_POOL_INDEX_BITS)) - 1 : DEPOT_POOLS_CAP)
+#define STACK_MIN_POOLS 1024
+
+/*
+ * The pool_index is offset by 1 so the first record does not have a 0 handle.
+ */
+static unsigned int stack_max_pools __ro_after_init =
+	MIN((1LL << DEPOT_POOL_INDEX_BITS) - 1, 8192);
 
 static bool stack_depot_disabled;
 static bool __stack_depot_early_init_requested __initdata = IS_ENABLED(CONFIG_STACKDEPOT_ALWAYS_INIT);
@@ -62,7 +64,7 @@ static unsigned int stack_bucket_number_order;
 static unsigned int stack_hash_mask;
 
 /* Array of memory regions that store stack records. */
-static void *stack_pools[DEPOT_MAX_POOLS];
+static void **stack_pools;
 /* Newly allocated pool that is not yet added to stack_pools. */
 static void *new_pool;
 /* Number of pools in stack_pools. */
@@ -101,6 +103,32 @@ static int __init disable_stack_depot(char *str)
 }
 early_param("stack_depot_disable", disable_stack_depot);
 
+static int __init parse_max_pools(char *str)
+{
+	const long long limit = (1LL << (DEPOT_POOL_INDEX_BITS)) - 1;
+	unsigned int max_pools;
+	int rv;
+
+	rv = kstrtouint(str, 0, &max_pools);
+	if (rv)
+		return rv;
+
+	if (max_pools < STACK_MIN_POOLS) {
+		pr_err("stack_depot_max_pools value %u too low, clamping to %d\n",
+		       max_pools, STACK_MIN_POOLS);
+		max_pools = STACK_MIN_POOLS;
+	} else if (max_pools > limit) {
+		pr_err("stack_depot_max_pools value %u too high, clamping to %lld\n",
+		       max_pools, limit);
+		max_pools = limit;
+	}
+
+	stack_max_pools = max_pools;
+
+	return 0;
+}
+early_param("stack_depot_max_pools", parse_max_pools);
+
 void __init stack_depot_request_early_init(void)
 {
 	/* Too late to request early init now. */
@@ -182,6 +210,15 @@ int __init stack_depot_early_init(void)
 	}
 	init_stack_table(entries);
 
+	pr_info("allocating space for %u stack pools via memblock\n", stack_max_pools);
+	stack_pools = memblock_alloc(stack_max_pools * sizeof(void *), PAGE_SIZE);
+	if (!stack_pools) {
+		pr_err("stack pools allocation failed, disabling\n");
+		memblock_free(stack_table, entries * sizeof(struct list_head));
+		stack_depot_disabled = true;
+		return -ENOMEM;
+	}
+
 	return 0;
 }
 
@@ -231,6 +268,15 @@ int stack_depot_init(void)
 	stack_hash_mask = entries - 1;
 	init_stack_table(entries);
 
+	pr_info("allocating space for %u stack pools via kvcalloc\n", stack_max_pools);
+	stack_pools = kvcalloc(stack_max_pools, sizeof(void *), GFP_KERNEL);
+	if (!stack_pools) {
+		pr_err("stack pools allocation failed, disabling\n");
+		kvfree(stack_table);
+		stack_depot_disabled = true;
+		ret = -ENOMEM;
+	}
+
 out_unlock:
 	mutex_unlock(&stack_depot_init_mutex);
 
@@ -245,9 +291,9 @@ static bool depot_init_pool(void **prealloc)
 {
 	lockdep_assert_held(&pool_lock);
 
-	if (unlikely(pools_num >= DEPOT_MAX_POOLS)) {
+	if (unlikely(pools_num >= stack_max_pools)) {
 		/* Bail out if we reached the pool limit. */
-		WARN_ON_ONCE(pools_num > DEPOT_MAX_POOLS); /* should never happen */
+		WARN_ON_ONCE(pools_num > stack_max_pools); /* should never happen */
 		WARN_ON_ONCE(!new_pool); /* to avoid unnecessary pre-allocation */
 		WARN_ONCE(1, "Stack depot reached limit capacity");
 		return false;
@@ -273,7 +319,7 @@ static bool depot_init_pool(void **prealloc)
 	 * NULL; do not reset to NULL if we have reached the maximum number of
 	 * pools.
 	 */
-	if (pools_num < DEPOT_MAX_POOLS)
+	if (pools_num < stack_max_pools)
 		WRITE_ONCE(new_pool, NULL);
 	else
 		WRITE_ONCE(new_pool, STACK_DEPOT_POISON);
-- 
2.34.1