From: Vlastimil Babka vbabka@suse.cz
mainline inclusion from mainline-v5.19-rc1 commit a5f1783be29adae15666fd803efd7d2979130869 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I9GSSR
Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i...
--------------------------------
In a later patch we want to add stackdepot support for object owner tracking in slub caches, which is enabled by slub_debug boot parameter. This creates a bootstrap problem as some caches are created early in boot when slab_is_available() is false and thus stack_depot_init() tries to use memblock. But, as reported by Hyeonggon Yoo [1] we are already beyond memblock_free_all(). Ideally memblock allocation should fail, yet it succeeds, but later the system crashes, which is a separately handled issue.
To resolve this boostrap issue in a robust way, this patch adds another way to request stack_depot_early_init(), which happens at a well-defined point of time. In addition to build-time CONFIG_STACKDEPOT_ALWAYS_INIT, code that's e.g. processing boot parameters (which happens early enough) can call a new function stack_depot_want_early_init(), which sets a flag that stack_depot_early_init() will check.
In this patch we also convert page_owner to this approach. While it doesn't have the bootstrap issue as slub, it's also a functionality enabled by a boot param and can thus request stack_depot_early_init() with memblock allocation instead of later initialization with kvmalloc().
As suggested by Mike, make stack_depot_early_init() only attempt memblock allocation and stack_depot_init() only attempt kvmalloc(). Also change the latter to kvcalloc(). In both cases we can lose the explicit array zeroing, which the allocations do already.
As suggested by Marco, provide empty implementations of the init functions for !CONFIG_STACKDEPOT builds to simplify the callers.
[1] https://lore.kernel.org/all/YhnUcqyeMgCrWZbd@ip-172-31-19-208.ap-northeast-1...
Reported-by: Hyeonggon Yoo 42.hyeyoo@gmail.com Suggested-by: Mike Rapoport rppt@linux.ibm.com Suggested-by: Marco Elver elver@google.com Signed-off-by: Vlastimil Babka vbabka@suse.cz Reviewed-by: Marco Elver elver@google.com Reviewed-and-tested-by: Hyeonggon Yoo 42.hyeyoo@gmail.com Reviewed-by: Mike Rapoport rppt@linux.ibm.com Acked-by: David Rientjes rientjes@google.com
Conflicts: include/linux/stackdepot.h lib/stackdepot.c mm/page_owner.c [context conflicts] Signed-off-by: Jinjiang Tu tujinjiang@huawei.com --- include/linux/stackdepot.h | 25 +++++++++++--- lib/stackdepot.c | 67 +++++++++++++++++++++++++------------- mm/page_owner.c | 5 +-- 3 files changed, 69 insertions(+), 28 deletions(-)
diff --git a/include/linux/stackdepot.h b/include/linux/stackdepot.h index 30dfc8549b5f..563e92754035 100644 --- a/include/linux/stackdepot.h +++ b/include/linux/stackdepot.h @@ -14,18 +14,35 @@ typedef u32 depot_stack_handle_t;
/* - * Every user of stack depot has to call this during its own init when it's - * decided that it will be calling stack_depot_save() later. + * Every user of stack depot has to call stack_depot_init() during its own init + * when it's decided that it will be calling stack_depot_save() later. This is + * recommended for e.g. modules initialized later in the boot process, when + * slab_is_available() is true. * * The alternative is to select STACKDEPOT_ALWAYS_INIT to have stack depot * enabled as part of mm_init(), for subsystems where it's known at compile time * that stack depot will be used. + * + * Another alternative is to call stack_depot_want_early_init(), when the + * decision to use stack depot is taken e.g. when evaluating kernel boot + * parameters, which precedes the enablement point in mm_init(). + * + * stack_depot_init() and stack_depot_want_early_init() can be called regardless + * of CONFIG_STACKDEPOT and are no-op when disabled. The actual save/fetch/print + * functions should only be called from code that makes sure CONFIG_STACKDEPOT + * is enabled. */ +#ifdef CONFIG_STACKDEPOT int stack_depot_init(void); +void __init stack_depot_want_early_init(void);
-#ifdef CONFIG_STACKDEPOT_ALWAYS_INIT -static inline int stack_depot_early_init(void) { return stack_depot_init(); } +/* This is supposed to be called only from mm_init() */ +int __init stack_depot_early_init(void); #else +static inline int stack_depot_init(void) { return 0; } + +static inline void stack_depot_want_early_init(void) { } + static inline int stack_depot_early_init(void) { return 0; } #endif
diff --git a/lib/stackdepot.c b/lib/stackdepot.c index ede2a433055c..8aed500d5ee8 100644 --- a/lib/stackdepot.c +++ b/lib/stackdepot.c @@ -66,6 +66,9 @@ struct stack_record { unsigned long entries[1]; /* Variable-sized array of entries. */ };
+static bool __stack_depot_want_early_init __initdata = IS_ENABLED(CONFIG_STACKDEPOT_ALWAYS_INIT); +static bool __stack_depot_early_init_passed __initdata; + static void *stack_slabs[STACK_ALLOC_MAX_SLABS];
static int depot_index; @@ -164,38 +167,58 @@ static int __init is_stack_depot_disabled(char *str) } early_param("stack_depot_disable", is_stack_depot_disabled);
-/* - * __ref because of memblock_alloc(), which will not be actually called after - * the __init code is gone, because at that point slab_is_available() is true - */ -__ref int stack_depot_init(void) +void __init stack_depot_want_early_init(void) +{ + /* Too late to request early init now */ + WARN_ON(__stack_depot_early_init_passed); + + __stack_depot_want_early_init = true; +} + +int __init stack_depot_early_init(void) +{ + size_t size; + + /* This is supposed to be called only once, from mm_init() */ + if (WARN_ON(__stack_depot_early_init_passed)) + return 0; + + __stack_depot_early_init_passed = true; + + if (!__stack_depot_want_early_init || stack_depot_disable) + return 0; + + size = (STACK_HASH_SIZE * sizeof(struct stack_record *)); + pr_info("Stack Depot early init allocating hash table with memblock_alloc, %zu bytes\n", + size); + stack_table = memblock_alloc(size, SMP_CACHE_BYTES); + + if (!stack_table) { + pr_err("Stack Depot hash table allocation failed, disabling\n"); + stack_depot_disable = true; + return -ENOMEM; + } + + return 0; +} + +int stack_depot_init(void) { static DEFINE_MUTEX(stack_depot_init_mutex); + int ret = 0;
mutex_lock(&stack_depot_init_mutex); if (!stack_depot_disable && !stack_table) { - size_t size = (STACK_HASH_SIZE * sizeof(struct stack_record *)); - int i; - - if (slab_is_available()) { - pr_info("Stack Depot allocating hash table with kvmalloc\n"); - stack_table = kvmalloc(size, GFP_KERNEL); - } else { - pr_info("Stack Depot allocating hash table with memblock_alloc\n"); - stack_table = memblock_alloc(size, SMP_CACHE_BYTES); - } - if (stack_table) { - for (i = 0; i < STACK_HASH_SIZE; i++) - stack_table[i] = NULL; - } else { + pr_info("Stack Depot allocating hash table with kvcalloc\n"); + stack_table = kvcalloc(STACK_HASH_SIZE, sizeof(struct stack_record *), GFP_KERNEL); + if (!stack_table) { pr_err("Stack Depot hash table allocation failed, disabling\n"); stack_depot_disable = true; - mutex_unlock(&stack_depot_init_mutex); - return -ENOMEM; + ret = -ENOMEM; } } mutex_unlock(&stack_depot_init_mutex); - return 0; + return ret; } EXPORT_SYMBOL_GPL(stack_depot_init);
diff --git a/mm/page_owner.c b/mm/page_owner.c index 61ba829e895c..a80917074147 100644 --- a/mm/page_owner.c +++ b/mm/page_owner.c @@ -48,6 +48,9 @@ static int __init early_page_owner_param(char *buf) if (strcmp(buf, "on") == 0) page_owner_enabled = true;
+ if (page_owner_enabled) + stack_depot_want_early_init(); + return 0; } early_param("page_owner", early_page_owner_param); @@ -86,8 +89,6 @@ static void init_page_owner(void) if (!page_owner_enabled) return;
- stack_depot_init(); - register_dummy_stack(); register_failure_stack(); register_early_stack();