From nobody Sat Apr 11 12:31:05 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 1896FECAAD2 for ; Thu, 1 Sep 2022 04:43:54 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232879AbiIAEnw (ORCPT ); Thu, 1 Sep 2022 00:43:52 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:60704 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232724AbiIAEn2 (ORCPT ); Thu, 1 Sep 2022 00:43:28 -0400 Received: from smtp-out1.suse.de (smtp-out1.suse.de [IPv6:2001:67c:2178:6::1c]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 14000E1A92 for ; Wed, 31 Aug 2022 21:43:01 -0700 (PDT) Received: from imap2.suse-dmz.suse.de (imap2.suse-dmz.suse.de [192.168.254.74]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (P-521) server-digest SHA512) (No client certificate requested) by smtp-out1.suse.de (Postfix) with ESMTPS id 7C8CA221E1; Thu, 1 Sep 2022 04:42:56 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=suse.de; s=susede2_rsa; t=1662007376; h=from:from:reply-to:date:date:message-id:message-id:to:to:cc:cc: mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=kOrpVxoLkMuflgUaHdVwwjK9xZD6TEf1SyG3O69U9sM=; b=oqUP1He733QNLVLXx7ZWAPtgoRicZnRMKKIqn8PvRcZFqUMfNU+jubRai5zZXdbSjGZfB2 tKE3euSDQAC6MFj2MJXx5k6xZSq1IAA2fL0ZIZZsqjYwbGO3vlnds8nEelP7CxpLh+NQYZ o3Ilo8vI2yNYqqnLcy6r/u3kaQL7ueA= DKIM-Signature: v=1; a=ed25519-sha256; c=relaxed/relaxed; d=suse.de; s=susede2_ed25519; t=1662007376; h=from:from:reply-to:date:date:message-id:message-id:to:to:cc:cc: mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=kOrpVxoLkMuflgUaHdVwwjK9xZD6TEf1SyG3O69U9sM=; b=JHXCx1XeJ41HxPNZTsatF5O2LGA8NRw2HNBLQSosOcF4hihZSq3/QPnWCCbEBPm7bBx6OP zbzHRN1kgE6rixCw== Received: from imap2.suse-dmz.suse.de (imap2.suse-dmz.suse.de [192.168.254.74]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (P-521) server-digest SHA512) (No client certificate requested) by imap2.suse-dmz.suse.de (Postfix) with ESMTPS id E7CBD139C4; Thu, 1 Sep 2022 04:42:55 +0000 (UTC) Received: from dovecot-director2.suse.de ([192.168.254.65]) by imap2.suse-dmz.suse.de with ESMTPSA id +GvANU84EGPheAAAMHmgww (envelope-from ); Thu, 01 Sep 2022 04:42:55 +0000 From: Oscar Salvador To: Andrew Morton Cc: linux-kernel@vger.kernel.org, linux-mm@kvack.org, Michal Hocko , Vlastimil Babka , Eric Dumazet , Waiman Long , Suren Baghdasaryan , Oscar Salvador Subject: [PATCH 1/3] lib/stackdepot: Add a refcount field in stack_record Date: Thu, 1 Sep 2022 06:42:47 +0200 Message-Id: <20220901044249.4624-2-osalvador@suse.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220901044249.4624-1-osalvador@suse.de> References: <20220901044249.4624-1-osalvador@suse.de> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" We want to filter out page_owner output and print only those stacks that have been repeated beyond a certain threshold. This gives us the chance to get rid of a lot of noise. In order to do that, we need to keep track of how many repeated stacks (for allocation) do we have, so we add a new refcount_t field in the stack_record struct. Note that on __set_page_owner_handle(), page_owner->handle is set, and on __reset_page_owner(), page_owner->free_handle is set. We are interested in page_owner->handle, so when __set_page_owner() gets called, we derive the stack_record struct from page_owner->handle, and we increment its refcount_t field; and when __reset_page_owner() gets called, we derive its stack_record from page_owner->handle() and we decrement its refcount_t field. This is a preparation for patch#2. Signed-off-by: Oscar Salvador --- include/linux/stackdepot.h | 13 ++++++- lib/stackdepot.c | 79 +++++++++++++++++++++++++++++++------- mm/kasan/common.c | 3 +- mm/page_owner.c | 13 +++++-- 4 files changed, 88 insertions(+), 20 deletions(-) diff --git a/include/linux/stackdepot.h b/include/linux/stackdepot.h index bc2797955de9..5ee0cf5be88f 100644 --- a/include/linux/stackdepot.h +++ b/include/linux/stackdepot.h @@ -15,9 +15,16 @@ =20 typedef u32 depot_stack_handle_t; =20 +typedef enum stack_action { + STACK_ACTION_NONE, + STACK_ACTION_INC, +}stack_action_t; + depot_stack_handle_t __stack_depot_save(unsigned long *entries, unsigned int nr_entries, - gfp_t gfp_flags, bool can_alloc); + gfp_t gfp_flags, bool can_alloc, + stack_action_t action); +void stack_depot_dec_count(depot_stack_handle_t handle); =20 /* * Every user of stack depot has to call stack_depot_init() during its own= init @@ -55,6 +62,10 @@ static inline int stack_depot_early_init(void) { return = 0; } =20 depot_stack_handle_t stack_depot_save(unsigned long *entries, unsigned int nr_entries, gfp_t gfp_flags); +depot_stack_handle_t stack_depot_save_action(unsigned long *entries, + unsigned int nr_entries, + gfp_t gfp_flags, + stack_action_t action); =20 unsigned int stack_depot_fetch(depot_stack_handle_t handle, unsigned long **entries); diff --git a/lib/stackdepot.c b/lib/stackdepot.c index 5ca0d086ef4a..aeb59d3557e2 100644 --- a/lib/stackdepot.c +++ b/lib/stackdepot.c @@ -63,6 +63,7 @@ struct stack_record { u32 hash; /* Hash in the hastable */ u32 size; /* Number of frames in the stack */ union handle_parts handle; + refcount_t count; /* Number of the same repeated stacks */ unsigned long entries[]; /* Variable-sized array of entries. */ }; =20 @@ -139,6 +140,7 @@ depot_alloc_stack(unsigned long *entries, int size, u32= hash, void **prealloc) stack->handle.slabindex =3D depot_index; stack->handle.offset =3D depot_offset >> STACK_ALLOC_ALIGN; stack->handle.valid =3D 1; + refcount_set(&stack->count, 1); memcpy(stack->entries, entries, flex_array_size(stack, entries, size)); depot_offset +=3D required_size; =20 @@ -302,6 +304,29 @@ void stack_depot_print(depot_stack_handle_t stack) } EXPORT_SYMBOL_GPL(stack_depot_print); =20 +static struct stack_record *stack_depot_getstack(depot_stack_handle_t hand= le) +{ + union handle_parts parts =3D { .handle =3D handle }; + void *slab; + size_t offset =3D parts.offset << STACK_ALLOC_ALIGN; + struct stack_record *stack; + + if(!handle) + return NULL; + + if (parts.slabindex > depot_index) { + WARN(1, "slab index %d out of bounds (%d) for stack id %08x\n", + parts.slabindex, depot_index, handle); + return NULL; + } + slab =3D stack_slabs[parts.slabindex]; + if (!slab) + return NULL; + + stack =3D slab + offset; + return stack; +} + /** * stack_depot_fetch - Fetch stack entries from a depot * @@ -314,30 +339,42 @@ EXPORT_SYMBOL_GPL(stack_depot_print); unsigned int stack_depot_fetch(depot_stack_handle_t handle, unsigned long **entries) { - union handle_parts parts =3D { .handle =3D handle }; - void *slab; - size_t offset =3D parts.offset << STACK_ALLOC_ALIGN; struct stack_record *stack; =20 *entries =3D NULL; if (!handle) return 0; =20 - if (parts.slabindex > depot_index) { - WARN(1, "slab index %d out of bounds (%d) for stack id %08x\n", - parts.slabindex, depot_index, handle); - return 0; - } - slab =3D stack_slabs[parts.slabindex]; - if (!slab) + stack =3D stack_depot_getstack(handle); + if (!stack) return 0; - stack =3D slab + offset; =20 *entries =3D stack->entries; return stack->size; } EXPORT_SYMBOL_GPL(stack_depot_fetch); =20 +static void stack_depot_inc_count(struct stack_record *stack) +{ + refcount_inc(&stack->count); +} + +void stack_depot_dec_count(depot_stack_handle_t handle) +{ + struct stack_record *stack =3D NULL; + + stack =3D stack_depot_getstack(handle); + if (stack) { + /* + * page_owner creates some stacks via create_dummy_stack(). + * We are not interested in those, so make sure we only decrement + * "valid" stacks. + */ + if (refcount_read(&stack->count) > 1) + refcount_dec(&stack->count); + } +} + /** * __stack_depot_save - Save a stack trace from an array * @@ -363,7 +400,8 @@ EXPORT_SYMBOL_GPL(stack_depot_fetch); */ depot_stack_handle_t __stack_depot_save(unsigned long *entries, unsigned int nr_entries, - gfp_t alloc_flags, bool can_alloc) + gfp_t alloc_flags, bool can_alloc, + stack_action_t action) { struct stack_record *found =3D NULL, **bucket; depot_stack_handle_t retval =3D 0; @@ -449,8 +487,11 @@ depot_stack_handle_t __stack_depot_save(unsigned long = *entries, /* Nobody used this memory, ok to free it. */ free_pages((unsigned long)prealloc, STACK_ALLOC_ORDER); } - if (found) + if (found) { retval =3D found->handle.handle; + if (action =3D=3D STACK_ACTION_INC) + stack_depot_inc_count(found); + } fast_exit: return retval; } @@ -472,6 +513,16 @@ depot_stack_handle_t stack_depot_save(unsigned long *e= ntries, unsigned int nr_entries, gfp_t alloc_flags) { - return __stack_depot_save(entries, nr_entries, alloc_flags, true); + return __stack_depot_save(entries, nr_entries, alloc_flags, true, + STACK_ACTION_NONE); } EXPORT_SYMBOL_GPL(stack_depot_save); + +depot_stack_handle_t stack_depot_save_action(unsigned long *entries, + unsigned int nr_entries, + gfp_t alloc_flags, + stack_action_t action) +{ + return __stack_depot_save(entries, nr_entries, alloc_flags, true, action); +} +EXPORT_SYMBOL_GPL(stack_depot_save_action); diff --git a/mm/kasan/common.c b/mm/kasan/common.c index c40c0e7b3b5f..f434994f3b0d 100644 --- a/mm/kasan/common.c +++ b/mm/kasan/common.c @@ -36,7 +36,8 @@ depot_stack_handle_t kasan_save_stack(gfp_t flags, bool c= an_alloc) unsigned int nr_entries; =20 nr_entries =3D stack_trace_save(entries, ARRAY_SIZE(entries), 0); - return __stack_depot_save(entries, nr_entries, flags, can_alloc); + return __stack_depot_save(entries, nr_entries, flags, can_alloc, + STACK_ACTION_NONE); } =20 void kasan_set_track(struct kasan_track *track, gfp_t flags) diff --git a/mm/page_owner.c b/mm/page_owner.c index e4c6f3f1695b..794f346d7520 100644 --- a/mm/page_owner.c +++ b/mm/page_owner.c @@ -106,7 +106,7 @@ static inline struct page_owner *get_page_owner(struct = page_ext *page_ext) return (void *)page_ext + page_owner_ops.offset; } =20 -static noinline depot_stack_handle_t save_stack(gfp_t flags) +static noinline depot_stack_handle_t save_stack(gfp_t flags, stack_action_= t action) { unsigned long entries[PAGE_OWNER_STACK_DEPTH]; depot_stack_handle_t handle; @@ -125,7 +125,7 @@ static noinline depot_stack_handle_t save_stack(gfp_t f= lags) current->in_page_owner =3D 1; =20 nr_entries =3D stack_trace_save(entries, ARRAY_SIZE(entries), 2); - handle =3D stack_depot_save(entries, nr_entries, flags); + handle =3D stack_depot_save_action(entries, nr_entries, flags, action); if (!handle) handle =3D failure_handle; =20 @@ -138,6 +138,7 @@ void __reset_page_owner(struct page *page, unsigned sho= rt order) int i; struct page_ext *page_ext; depot_stack_handle_t handle; + depot_stack_handle_t alloc_handle; struct page_owner *page_owner; u64 free_ts_nsec =3D local_clock(); =20 @@ -145,7 +146,10 @@ void __reset_page_owner(struct page *page, unsigned sh= ort order) if (unlikely(!page_ext)) return; =20 - handle =3D save_stack(GFP_NOWAIT | __GFP_NOWARN); + page_owner =3D get_page_owner(page_ext); + alloc_handle =3D page_owner->handle; + + handle =3D save_stack(GFP_NOWAIT | __GFP_NOWARN, STACK_ACTION_NONE); for (i =3D 0; i < (1 << order); i++) { __clear_bit(PAGE_EXT_OWNER_ALLOCATED, &page_ext->flags); page_owner =3D get_page_owner(page_ext); @@ -153,6 +157,7 @@ void __reset_page_owner(struct page *page, unsigned sho= rt order) page_owner->free_ts_nsec =3D free_ts_nsec; page_ext =3D page_ext_next(page_ext); } + stack_depot_dec_count(alloc_handle); } =20 static inline void __set_page_owner_handle(struct page_ext *page_ext, @@ -189,7 +194,7 @@ noinline void __set_page_owner(struct page *page, unsig= ned short order, if (unlikely(!page_ext)) return; =20 - handle =3D save_stack(gfp_mask); + handle =3D save_stack(gfp_mask, STACK_ACTION_INC); __set_page_owner_handle(page_ext, handle, order, gfp_mask); } =20 --=20 2.35.3 From nobody Sat Apr 11 12:31:05 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 19C17ECAAD2 for ; Thu, 1 Sep 2022 04:44:10 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232903AbiIAEoI (ORCPT ); Thu, 1 Sep 2022 00:44:08 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:58862 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232744AbiIAEne (ORCPT ); Thu, 1 Sep 2022 00:43:34 -0400 Received: from smtp-out1.suse.de (smtp-out1.suse.de [195.135.220.28]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 80359111AFA for ; Wed, 31 Aug 2022 21:43:10 -0700 (PDT) Received: from imap2.suse-dmz.suse.de (imap2.suse-dmz.suse.de [192.168.254.74]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (P-521) server-digest SHA512) (No client certificate requested) by smtp-out1.suse.de (Postfix) with ESMTPS id 20776221E2; Thu, 1 Sep 2022 04:42:57 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=suse.de; s=susede2_rsa; t=1662007377; h=from:from:reply-to:date:date:message-id:message-id:to:to:cc:cc: mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=I+9op6+s45dBeDDz0hB6cxjRW0r3Pje/FesTmSW2Zek=; b=NAIeqLSDDSE+P9kHg5Syu0+cXWEM5KdAOu1En/FPsEPWos0BsZE1cBPGZR5MRySX53IHfn wTIcA1u8waN9m0Rh3w/lLxoexge/vWQvCv0TcjSze3peDkowN5nJZecsg+6TI/P3K5tBGs 5iCdzLF5xUjKZuBqvi542DvQpNR+D64= DKIM-Signature: v=1; a=ed25519-sha256; c=relaxed/relaxed; d=suse.de; s=susede2_ed25519; t=1662007377; h=from:from:reply-to:date:date:message-id:message-id:to:to:cc:cc: mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=I+9op6+s45dBeDDz0hB6cxjRW0r3Pje/FesTmSW2Zek=; b=0oalvhcOoaYu6EzaGL2RxZ5vSihBQFyAzeJIJ+GbTXuKLQnsKZUIv00y4F5pN+YHdBNWQo Nqe/TXfpazjOnrCg== Received: from imap2.suse-dmz.suse.de (imap2.suse-dmz.suse.de [192.168.254.74]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (P-521) server-digest SHA512) (No client certificate requested) by imap2.suse-dmz.suse.de (Postfix) with ESMTPS id 8C293139C4; Thu, 1 Sep 2022 04:42:56 +0000 (UTC) Received: from dovecot-director2.suse.de ([192.168.254.65]) by imap2.suse-dmz.suse.de with ESMTPSA id IMKAH1A4EGPheAAAMHmgww (envelope-from ); Thu, 01 Sep 2022 04:42:56 +0000 From: Oscar Salvador To: Andrew Morton Cc: linux-kernel@vger.kernel.org, linux-mm@kvack.org, Michal Hocko , Vlastimil Babka , Eric Dumazet , Waiman Long , Suren Baghdasaryan , Oscar Salvador Subject: [PATCH 2/3] mm, page_owner: Add page_owner_stacks file to print out only stacks and their counter Date: Thu, 1 Sep 2022 06:42:48 +0200 Message-Id: <20220901044249.4624-3-osalvador@suse.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220901044249.4624-1-osalvador@suse.de> References: <20220901044249.4624-1-osalvador@suse.de> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" We might be only interested in knowing about stacks <-> count relationship, so instead of having to fiddle with page_owner output and screen through pfns, let us add a new file called 'page_owner_stacks' that does just that. By cating such file, we will get all the stacktraces followed by its counter (allocated - freed times), so we can have a more specific overview. Signed-off-by: Oscar Salvador Reported-by: kernel test robot --- include/linux/stackdepot.h | 2 ++ lib/stackdepot.c | 40 ++++++++++++++++++++++++++++++++++++++ mm/page_owner.c | 30 ++++++++++++++++++++++++++++ 3 files changed, 72 insertions(+) diff --git a/include/linux/stackdepot.h b/include/linux/stackdepot.h index 5ee0cf5be88f..20f62039f23a 100644 --- a/include/linux/stackdepot.h +++ b/include/linux/stackdepot.h @@ -25,6 +25,8 @@ depot_stack_handle_t __stack_depot_save(unsigned long *en= tries, gfp_t gfp_flags, bool can_alloc, stack_action_t action); void stack_depot_dec_count(depot_stack_handle_t handle); +int stack_depot_print_stacks_threshold(char *buf, size_t size, loff_t *pos, + unsigned long *last_stack); =20 /* * Every user of stack depot has to call stack_depot_init() during its own= init diff --git a/lib/stackdepot.c b/lib/stackdepot.c index aeb59d3557e2..3090ae0f3958 100644 --- a/lib/stackdepot.c +++ b/lib/stackdepot.c @@ -526,3 +526,43 @@ depot_stack_handle_t stack_depot_save_action(unsigned = long *entries, return __stack_depot_save(entries, nr_entries, alloc_flags, true, action); } EXPORT_SYMBOL_GPL(stack_depot_save_action); + +int stack_depot_print_stacks_threshold(char *buf, size_t size, loff_t *pos, + unsigned long *last_stack) +{ + struct stack_record *stack =3D NULL, *last; + struct stack_record **stacks; + int i =3D *pos, ret =3D 0; + + /* Continue from the last week if we have one */ + if (*last_stack) { + last =3D (struct stack_record *)*last_stack; + stack =3D last->next; + } else { +new_table: + stacks =3D &stack_table[i]; + stack =3D (struct stack_record *)stacks; + } + + for (; stack; stack =3D stack->next) { + if (!stack->size || stack->size < 0 || + stack->size > size || stack->handle.valid !=3D 1 || + refcount_read(&stack->count) < 1) + continue; + + ret +=3D stack_trace_snprint(buf, size, stack->entries, stack->size, 0); + ret +=3D scnprintf(buf + ret, size - ret, "stack count: %d\n\n", + refcount_read(&stack->count)); + *last_stack =3D (unsigned long)stack; + return ret; + } + + i++; + *pos =3D i; + + /* Keep looking all tables for valid stacks */ + if (i < STACK_HASH_SIZE) + goto new_table; + + return 0; +} diff --git a/mm/page_owner.c b/mm/page_owner.c index 794f346d7520..8c67c7eb2451 100644 --- a/mm/page_owner.c +++ b/mm/page_owner.c @@ -43,6 +43,8 @@ static depot_stack_handle_t early_handle; =20 static void init_early_allocated_pages(void); =20 +static unsigned long last_stack =3D 0; + static int __init early_page_owner_param(char *buf) { int ret =3D kstrtobool(buf, &page_owner_enabled); @@ -663,6 +665,32 @@ static void init_early_allocated_pages(void) init_zones_in_node(pgdat); } =20 +static ssize_t read_page_owner_stacks(struct file *file, char __user *buf, + size_t count, loff_t *pos) +{ + char *kbuf; + int ret =3D 0; + + count =3D min_t(size_t, count, PAGE_SIZE); + kbuf =3D kmalloc(count, GFP_KERNEL); + if (!kbuf) + return ENOMEM; + + ret +=3D stack_depot_print_stacks_threshold(kbuf, count, pos, &last_stack= ); + if (copy_to_user(buf, kbuf, ret)) + ret =3D -EFAULT; + + if (!ret) + last_stack =3D 0; + + kfree(kbuf); + return ret; +} + +static const struct file_operations proc_page_owner_stacks =3D { + .read =3D read_page_owner_stacks, +}; + static const struct file_operations proc_page_owner_operations =3D { .read =3D read_page_owner, }; @@ -676,6 +704,8 @@ static int __init pageowner_init(void) =20 debugfs_create_file("page_owner", 0400, NULL, NULL, &proc_page_owner_operations); + debugfs_create_file("page_owner_stacks", 0400, NULL, NULL, + &proc_page_owner_stacks); =20 return 0; } --=20 2.35.3 From nobody Sat Apr 11 12:31:05 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 543B6ECAAD1 for ; Thu, 1 Sep 2022 04:43:57 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232887AbiIAEn4 (ORCPT ); Thu, 1 Sep 2022 00:43:56 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59016 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232827AbiIAEn3 (ORCPT ); Thu, 1 Sep 2022 00:43:29 -0400 Received: from smtp-out1.suse.de (smtp-out1.suse.de [IPv6:2001:67c:2178:6::1c]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 370A9ED01C for ; Wed, 31 Aug 2022 21:43:02 -0700 (PDT) Received: from imap2.suse-dmz.suse.de (imap2.suse-dmz.suse.de [192.168.254.74]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (P-521) server-digest SHA512) (No client certificate requested) by smtp-out1.suse.de (Postfix) with ESMTPS id D2024221EE; Thu, 1 Sep 2022 04:42:57 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=suse.de; s=susede2_rsa; t=1662007377; h=from:from:reply-to:date:date:message-id:message-id:to:to:cc:cc: mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=e8DWylbFwSCb/px9RHGyLSh+EhBiVeZp3bgNDtmW9IA=; b=S7u1rvndnGXFGe9gpIQARWw/22R+6XqNKcNzRqacOvUWansQ9BHdjWh8iRcKXrqgqLuOV2 v6DNKPXLbjMM5RWzIlyGrePp8e4xf2+G40n99CDJxSnS5QOyKG04Sjxj+U0/vhXgjzyXUW RP6Cdu7Ae7QcP7PhwIku/EEFbAb/KXE= DKIM-Signature: v=1; a=ed25519-sha256; c=relaxed/relaxed; d=suse.de; s=susede2_ed25519; t=1662007377; h=from:from:reply-to:date:date:message-id:message-id:to:to:cc:cc: mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=e8DWylbFwSCb/px9RHGyLSh+EhBiVeZp3bgNDtmW9IA=; b=BPF/tSxXMQDSvEmbgyPhuBR9rysuTMAjY41Rxsu8/nW5dEMdII4hStCGu0uptmYWj17dRU oJeH7qfBX4uymWBw== Received: from imap2.suse-dmz.suse.de (imap2.suse-dmz.suse.de [192.168.254.74]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (P-521) server-digest SHA512) (No client certificate requested) by imap2.suse-dmz.suse.de (Postfix) with ESMTPS id 32E2C139C4; Thu, 1 Sep 2022 04:42:57 +0000 (UTC) Received: from dovecot-director2.suse.de ([192.168.254.65]) by imap2.suse-dmz.suse.de with ESMTPSA id WKusCVE4EGPheAAAMHmgww (envelope-from ); Thu, 01 Sep 2022 04:42:57 +0000 From: Oscar Salvador To: Andrew Morton Cc: linux-kernel@vger.kernel.org, linux-mm@kvack.org, Michal Hocko , Vlastimil Babka , Eric Dumazet , Waiman Long , Suren Baghdasaryan , Oscar Salvador Subject: [PATCH 3/3] mm,page_owner: Filter out stacks by a threshold counter Date: Thu, 1 Sep 2022 06:42:49 +0200 Message-Id: <20220901044249.4624-4-osalvador@suse.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220901044249.4624-1-osalvador@suse.de> References: <20220901044249.4624-1-osalvador@suse.de> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" We want to be able to filter out the output on a threshold basis, in this way we can get rid of a lot of noise and focus only on those stacks which have an allegedly high counter. We can control the threshold value by a new file called 'page_owner_threshold', which is 0 by default. Signed-off-by: Oscar Salvador --- include/linux/stackdepot.h | 3 +- lib/stackdepot.c | 6 ++-- mm/page_owner.c | 61 +++++++++++++++++++++++++++++++++++++- 3 files changed, 66 insertions(+), 4 deletions(-) diff --git a/include/linux/stackdepot.h b/include/linux/stackdepot.h index 20f62039f23a..ee66be40a152 100644 --- a/include/linux/stackdepot.h +++ b/include/linux/stackdepot.h @@ -26,7 +26,8 @@ depot_stack_handle_t __stack_depot_save(unsigned long *en= tries, stack_action_t action); void stack_depot_dec_count(depot_stack_handle_t handle); int stack_depot_print_stacks_threshold(char *buf, size_t size, loff_t *pos, - unsigned long *last_stack); + unsigned long *last_stack, + unsigned long threshold); =20 /* * Every user of stack depot has to call stack_depot_init() during its own= init diff --git a/lib/stackdepot.c b/lib/stackdepot.c index 3090ae0f3958..b4a04f09a7b7 100644 --- a/lib/stackdepot.c +++ b/lib/stackdepot.c @@ -528,7 +528,8 @@ depot_stack_handle_t stack_depot_save_action(unsigned l= ong *entries, EXPORT_SYMBOL_GPL(stack_depot_save_action); =20 int stack_depot_print_stacks_threshold(char *buf, size_t size, loff_t *pos, - unsigned long *last_stack) + unsigned long *last_stack, + unsigned long threshold) { struct stack_record *stack =3D NULL, *last; struct stack_record **stacks; @@ -547,7 +548,8 @@ int stack_depot_print_stacks_threshold(char *buf, size_= t size, loff_t *pos, for (; stack; stack =3D stack->next) { if (!stack->size || stack->size < 0 || stack->size > size || stack->handle.valid !=3D 1 || - refcount_read(&stack->count) < 1) + refcount_read(&stack->count) < 1 || + refcount_read(&stack->count) < threshold) continue; =20 ret +=3D stack_trace_snprint(buf, size, stack->entries, stack->size, 0); diff --git a/mm/page_owner.c b/mm/page_owner.c index 8c67c7eb2451..ef10cf44aaec 100644 --- a/mm/page_owner.c +++ b/mm/page_owner.c @@ -44,6 +44,7 @@ static depot_stack_handle_t early_handle; static void init_early_allocated_pages(void); =20 static unsigned long last_stack =3D 0; +static unsigned long threshold_count =3D 0; =20 static int __init early_page_owner_param(char *buf) { @@ -676,7 +677,8 @@ static ssize_t read_page_owner_stacks(struct file *file= , char __user *buf, if (!kbuf) return ENOMEM; =20 - ret +=3D stack_depot_print_stacks_threshold(kbuf, count, pos, &last_stack= ); + ret +=3D stack_depot_print_stacks_threshold(kbuf, count, pos, &last_stack, + threshold_count); if (copy_to_user(buf, kbuf, ret)) ret =3D -EFAULT; =20 @@ -687,6 +689,61 @@ static ssize_t read_page_owner_stacks(struct file *fil= e, char __user *buf, return ret; } =20 +static ssize_t read_page_owner_threshold(struct file *file, char __user *b= uf, + size_t count, loff_t *pos) +{ + char *kbuf; + int ret =3D 0; + + count =3D min_t(size_t, count, PAGE_SIZE); + + if (*pos >=3D count) + return 0; + + kbuf =3D kmalloc(count, GFP_KERNEL); + if (!kbuf) + return ENOMEM; + + ret =3D scnprintf(kbuf, count, "%lu\n", threshold_count); + if (copy_to_user(buf, kbuf, ret)) + ret =3D -EFAULT; + + *pos +=3D count; + kfree(kbuf); + + return ret; +} + +static ssize_t write_page_owner_threshold(struct file *file, const char __= user *buf, + size_t count, loff_t *pos) +{ + char *kbuf; + int ret =3D 0; + + count =3D min_t(size_t, count, PAGE_SIZE); + kbuf =3D kmalloc(count, GFP_KERNEL); + if (!kbuf) + return -ENOMEM; + + if (copy_from_user(kbuf, buf, count)) { + ret =3D -EFAULT; + goto out; + } + + kbuf[count - 1] =3D '\0'; + + ret =3D kstrtoul(kbuf, 10, &threshold_count); + +out: + kfree(kbuf); + return ret ? ret : count; +} + +static const struct file_operations proc_page_owner_threshold =3D { + .read =3D read_page_owner_threshold, + .write =3D write_page_owner_threshold, +}; + static const struct file_operations proc_page_owner_stacks =3D { .read =3D read_page_owner_stacks, }; @@ -706,6 +763,8 @@ static int __init pageowner_init(void) &proc_page_owner_operations); debugfs_create_file("page_owner_stacks", 0400, NULL, NULL, &proc_page_owner_stacks); + debugfs_create_file("page_owner_threshold", 0600, NULL, NULL, + &proc_page_owner_threshold); =20 return 0; } --=20 2.35.3