From: Suren Baghdasaryan <surenb@google.com>
To: akpm@linux-foundation.org
Cc: kent.overstreet@linux.dev, vbabka@suse.cz, harry.yoo@oracle.com,
da.gomez@kernel.org, jonathanh@nvidia.com,
linux-kernel@vger.kernel.org, linux-mm@kvack.org,
Suren Baghdasaryan <surenb@google.com>
Subject: [PATCH 1/1] alloc_tag: skip rcu barrier on module unload if profiling was not enabled
Date: Thu, 4 Dec 2025 14:09:56 -0800 [thread overview]
Message-ID: <20251204220957.360838-1-surenb@google.com> (raw)
RCU barrier in the module unload path is needed to ensure all kfree_rcu
operations which modify tag counters are complete before we free the tags.
However this is not necessary if memory allocation profiling was never
enabled on the system. Introduce a sticky flag that records whether memory
profiling was ever enabled and make the RCU barrier conditional.
Suggested-by: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
---
Applies over mm-new
include/linux/codetag.h | 2 ++
lib/alloc_tag.c | 11 ++++++++++-
lib/codetag.c | 14 ++++++++++++--
3 files changed, 24 insertions(+), 3 deletions(-)
diff --git a/include/linux/codetag.h b/include/linux/codetag.h
index 8ea2a5f7c98a..8a07a83cabdd 100644
--- a/include/linux/codetag.h
+++ b/include/linux/codetag.h
@@ -95,6 +95,7 @@ void codetag_free_module_sections(struct module *mod);
void codetag_module_replaced(struct module *mod, struct module *new_mod);
int codetag_load_module(struct module *mod);
void codetag_unload_module(struct module *mod);
+void codetag_flush_rcu_on_module_unload(void);
#else /* defined(CONFIG_CODE_TAGGING) && defined(CONFIG_MODULES) */
@@ -109,6 +110,7 @@ static inline void codetag_free_module_sections(struct module *mod) {}
static inline void codetag_module_replaced(struct module *mod, struct module *new_mod) {}
static inline int codetag_load_module(struct module *mod) { return 0; }
static inline void codetag_unload_module(struct module *mod) {}
+static inline void codetag_flush_rcu_on_module_unload(void) {}
#endif /* defined(CONFIG_CODE_TAGGING) && defined(CONFIG_MODULES) */
diff --git a/lib/alloc_tag.c b/lib/alloc_tag.c
index 27fee57a5c91..173eebfacb1e 100644
--- a/lib/alloc_tag.c
+++ b/lib/alloc_tag.c
@@ -776,10 +776,16 @@ EXPORT_SYMBOL(page_alloc_tagging_ops);
static int proc_mem_profiling_handler(const struct ctl_table *table, int write,
void *buffer, size_t *lenp, loff_t *ppos)
{
+ int ret;
+
if (!mem_profiling_support && write)
return -EINVAL;
- return proc_do_static_key(table, write, buffer, lenp, ppos);
+ ret = proc_do_static_key(table, write, buffer, lenp, ppos);
+ if (!ret && write && mem_alloc_profiling_enabled())
+ codetag_flush_rcu_on_module_unload();
+
+ return ret;
}
@@ -829,6 +835,9 @@ static int __init alloc_tag_init(void)
return 0;
}
+ if (mem_alloc_profiling_enabled())
+ codetag_flush_rcu_on_module_unload();
+
if (!proc_create_seq_private(ALLOCINFO_FILE_NAME, 0400, NULL, &allocinfo_seq_op,
sizeof(struct allocinfo_private), NULL)) {
pr_err("Failed to create %s file\n", ALLOCINFO_FILE_NAME);
diff --git a/lib/codetag.c b/lib/codetag.c
index 545911cebd25..ed6d6425b62a 100644
--- a/lib/codetag.c
+++ b/lib/codetag.c
@@ -34,6 +34,7 @@ struct codetag_module {
static DEFINE_MUTEX(codetag_lock);
static LIST_HEAD(codetag_types);
+static bool flush_rcu_on_module_unload;
void codetag_lock_module_list(struct codetag_type *cttype, bool lock)
{
@@ -335,6 +336,11 @@ int codetag_load_module(struct module *mod)
return ret;
}
+void codetag_flush_rcu_on_module_unload(void)
+{
+ flush_rcu_on_module_unload = true;
+}
+
void codetag_unload_module(struct module *mod)
{
struct codetag_type *cttype;
@@ -342,8 +348,12 @@ void codetag_unload_module(struct module *mod)
if (!mod)
return;
- /* await any module's kfree_rcu() operations to complete */
- kvfree_rcu_barrier();
+ /*
+ * Await any module's kfree_rcu() operations to complete
+ * if profiling was ever enabled.
+ */
+ if (flush_rcu_on_module_unload)
+ kvfree_rcu_barrier();
mutex_lock(&codetag_lock);
list_for_each_entry(cttype, &codetag_types, link) {
base-commit: 3f43be96f919cc611dcb2a4e38dd464831f4513e
--
2.52.0.223.gf5cc29aaa4-goog
next reply other threads:[~2025-12-04 22:10 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-12-04 22:09 Suren Baghdasaryan [this message]
2025-12-06 4:45 ` kernel test robot
2025-12-06 5:08 ` kernel test robot
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=20251204220957.360838-1-surenb@google.com \
--to=surenb@google.com \
--cc=akpm@linux-foundation.org \
--cc=da.gomez@kernel.org \
--cc=harry.yoo@oracle.com \
--cc=jonathanh@nvidia.com \
--cc=kent.overstreet@linux.dev \
--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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox