All of lore.kernel.org
 help / color / mirror / Atom feed
From: andrey.konovalov@linux.dev
To: Marco Elver <elver@google.com>, Alexander Potapenko <glider@google.com>
Cc: Andrey Konovalov <andreyknvl@gmail.com>,
	Dmitry Vyukov <dvyukov@google.com>,
	Vlastimil Babka <vbabka@suse.cz>,
	kasan-dev@googlegroups.com, Evgenii Stepanov <eugenis@google.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	linux-mm@kvack.org, linux-kernel@vger.kernel.org,
	Andrey Konovalov <andreyknvl@google.com>
Subject: [PATCH 14/15] stackdepot: allow users to evict stack traces
Date: Tue, 29 Aug 2023 19:11:24 +0200	[thread overview]
Message-ID: <99cd7ac4a312e86c768b933332364272b9e3fb40.1693328501.git.andreyknvl@google.com> (raw)
In-Reply-To: <cover.1693328501.git.andreyknvl@google.com>

From: Andrey Konovalov <andreyknvl@google.com>

Add stack_depot_evict, a function that decrements a reference counter
on a stack record and removes it from the stack depot once the counter
reaches 0.

Internally, when removing a stack record, the function unlinks it from
the hash table bucket and returns to the freelist.

With this change, the users of stack depot can call stack_depot_evict
when keeping a stack trace in the stack depot is not needed anymore.
This allows avoiding polluting the stack depot with irrelevant stack
traces and thus have more space to store the relevant ones before the
stack depot reaches its capacity.

Signed-off-by: Andrey Konovalov <andreyknvl@google.com>
---
 include/linux/stackdepot.h | 11 ++++++++++
 lib/stackdepot.c           | 43 ++++++++++++++++++++++++++++++++++++++
 2 files changed, 54 insertions(+)

diff --git a/include/linux/stackdepot.h b/include/linux/stackdepot.h
index e58306783d8e..b14da6797714 100644
--- a/include/linux/stackdepot.h
+++ b/include/linux/stackdepot.h
@@ -121,6 +121,17 @@ depot_stack_handle_t stack_depot_save(unsigned long *entries,
 unsigned int stack_depot_fetch(depot_stack_handle_t handle,
 			       unsigned long **entries);
 
+/**
+ * stack_depot_evict - Drop a reference to a stack trace from stack depot
+ *
+ * @handle:	Stack depot handle returned from stack_depot_save()
+ *
+ * The stack trace gets fully removed from stack depot once all references
+ * to it has been dropped (once the number of stack_depot_evict calls matches
+ * the number of stack_depot_save calls for this stack trace).
+ */
+void stack_depot_evict(depot_stack_handle_t handle);
+
 /**
  * stack_depot_print - Print a stack trace from stack depot
  *
diff --git a/lib/stackdepot.c b/lib/stackdepot.c
index 641db97d8c7c..cf28720b842d 100644
--- a/lib/stackdepot.c
+++ b/lib/stackdepot.c
@@ -384,6 +384,13 @@ static struct stack_record *depot_fetch_stack(depot_stack_handle_t handle)
 	return stack;
 }
 
+/* Frees stack into the freelist. */
+static void depot_free_stack(struct stack_record *stack)
+{
+	stack->next = next_stack;
+	next_stack = stack;
+}
+
 /* Calculates the hash for a stack. */
 static inline u32 hash_stack(unsigned long *entries, unsigned int size)
 {
@@ -555,6 +562,42 @@ unsigned int stack_depot_fetch(depot_stack_handle_t handle,
 }
 EXPORT_SYMBOL_GPL(stack_depot_fetch);
 
+void stack_depot_evict(depot_stack_handle_t handle)
+{
+	struct stack_record *stack, **bucket;
+	unsigned long flags;
+
+	if (!handle || stack_depot_disabled)
+		return;
+
+	write_lock_irqsave(&pool_rwlock, flags);
+
+	stack = depot_fetch_stack(handle);
+	if (WARN_ON(!stack))
+		goto out;
+
+	if (refcount_dec_and_test(&stack->count)) {
+		/* Drop stack from the hash table. */
+		if (stack->next)
+			stack->next->prev = stack->prev;
+		if (stack->prev)
+			stack->prev->next = stack->next;
+		else {
+			bucket = &stack_table[stack->hash & stack_hash_mask];
+			*bucket = stack->next;
+		}
+		stack->next = NULL;
+		stack->prev = NULL;
+
+		/* Free stack. */
+		depot_free_stack(stack);
+	}
+
+out:
+	write_unlock_irqrestore(&pool_rwlock, flags);
+}
+EXPORT_SYMBOL_GPL(stack_depot_evict);
+
 void stack_depot_print(depot_stack_handle_t stack)
 {
 	unsigned long *entries;
-- 
2.25.1


  parent reply	other threads:[~2023-08-29 17:15 UTC|newest]

Thread overview: 50+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-08-29 17:11 [PATCH 00/15] stackdepot: allow evicting stack traces andrey.konovalov
2023-08-29 17:11 ` [PATCH 01/15] stackdepot: check disabled flag when fetching andrey.konovalov
2023-08-30  7:40   ` Alexander Potapenko
2023-08-29 17:11 ` [PATCH 02/15] stackdepot: simplify __stack_depot_save andrey.konovalov
2023-08-30  7:41   ` Alexander Potapenko
2023-08-29 17:11 ` [PATCH 03/15] stackdepot: drop valid bit from handles andrey.konovalov
2023-08-30  7:43   ` Alexander Potapenko
2023-08-29 17:11 ` [PATCH 04/15] stackdepot: add depot_fetch_stack helper andrey.konovalov
2023-08-30  7:47   ` Alexander Potapenko
2023-08-29 17:11 ` [PATCH 05/15] stackdepot: use fixed-sized slots for stack records andrey.konovalov
2023-08-30  8:21   ` Alexander Potapenko
2023-09-13 17:07     ` Andrey Konovalov
2023-09-15 10:36       ` Alexander Potapenko
2023-08-29 17:11 ` [PATCH 06/15] stackdepot: fix and clean-up atomic annotations andrey.konovalov
2023-08-30  8:34   ` Marco Elver
2023-09-04 18:45     ` Andrey Konovalov
2023-08-29 17:11 ` [PATCH 07/15] stackdepot: rework helpers for depot_alloc_stack andrey.konovalov
2023-08-29 17:11 ` [PATCH 08/15] stackdepot: rename next_pool_required to new_pool_required andrey.konovalov
2023-08-30  8:29   ` Alexander Potapenko
2023-08-29 17:11 ` [PATCH 09/15] stackdepot: store next pool pointer in new_pool andrey.konovalov
2023-08-29 17:11 ` [PATCH 10/15] stackdepot: store free stack records in a freelist andrey.konovalov
2023-08-29 17:11 ` [PATCH 11/15] stackdepot: use read/write lock andrey.konovalov
2023-08-30  9:13   ` Marco Elver
2023-09-04 18:46     ` Andrey Konovalov
2023-09-05 16:19       ` Marco Elver
2023-09-13 17:08         ` Andrey Konovalov
2024-01-02 12:59           ` Marco Elver
2024-01-09  3:27             ` Andrey Konovalov
2023-08-29 17:11 ` [PATCH 12/15] stackdepot: add refcount for records andrey.konovalov
2023-08-30  9:32   ` Marco Elver
2023-09-04 18:46     ` Andrey Konovalov
2023-09-04 18:55       ` Marco Elver
2023-09-13 17:07         ` Andrey Konovalov
2023-09-01 13:06   ` Kuan-Ying Lee (李冠穎)
2023-09-04 18:46     ` Andrey Konovalov
2023-08-29 17:11 ` [PATCH 13/15] stackdepot: add backwards links to hash table buckets andrey.konovalov
2023-08-30  9:24   ` Marco Elver
2023-09-13 17:07     ` Andrey Konovalov
2023-08-29 17:11 ` andrey.konovalov [this message]
2023-08-30  9:20   ` [PATCH 14/15] stackdepot: allow users to evict stack traces Marco Elver
2023-09-04 18:47     ` Andrey Konovalov
2023-08-29 17:11 ` [PATCH 15/15] kasan: use stack_depot_evict for tag-based modes andrey.konovalov
2023-08-30  9:38   ` Marco Elver
2023-09-04 18:48     ` Andrey Konovalov
2023-09-04 18:58       ` Marco Elver
2023-09-13 17:08         ` Andrey Konovalov
2023-08-30  7:46 ` [PATCH 00/15] stackdepot: allow evicting stack traces Vlastimil Babka
2023-09-04 18:45   ` Andrey Konovalov
2023-09-05  2:48     ` Kuan-Ying Lee (李冠穎)
2023-09-13 17:10       ` Andrey Konovalov

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=99cd7ac4a312e86c768b933332364272b9e3fb40.1693328501.git.andreyknvl@google.com \
    --to=andrey.konovalov@linux.dev \
    --cc=akpm@linux-foundation.org \
    --cc=andreyknvl@gmail.com \
    --cc=andreyknvl@google.com \
    --cc=dvyukov@google.com \
    --cc=elver@google.com \
    --cc=eugenis@google.com \
    --cc=glider@google.com \
    --cc=kasan-dev@googlegroups.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=vbabka@suse.cz \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.