From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: stable@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
patches@lists.linux.dev, Mark Rutland <mark.rutland@arm.com>,
Mathieu Desnoyers <mathieu.desnoyers@efficios.com>,
Jiri Olsa <jolsa@kernel.org>, Alexei Starovoitov <ast@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>,
"Masami Hiramatsu (Google)" <mhiramat@kernel.org>,
"Steven Rostedt (Google)" <rostedt@goodmis.org>
Subject: [PATCH 6.6 45/49] ftrace: Fix modification of direct_function hash while in use
Date: Wed, 3 Jan 2024 17:56:05 +0100 [thread overview]
Message-ID: <20240103164841.969983746@linuxfoundation.org> (raw)
In-Reply-To: <20240103164834.970234661@linuxfoundation.org>
6.6-stable review patch. If anyone has any objections, please let me know.
------------------
From: Steven Rostedt (Google) <rostedt@goodmis.org>
commit d05cb470663a2a1879277e544f69e660208f08f2 upstream.
Masami Hiramatsu reported a memory leak in register_ftrace_direct() where
if the number of new entries are added is large enough to cause two
allocations in the loop:
for (i = 0; i < size; i++) {
hlist_for_each_entry(entry, &hash->buckets[i], hlist) {
new = ftrace_add_rec_direct(entry->ip, addr, &free_hash);
if (!new)
goto out_remove;
entry->direct = addr;
}
}
Where ftrace_add_rec_direct() has:
if (ftrace_hash_empty(direct_functions) ||
direct_functions->count > 2 * (1 << direct_functions->size_bits)) {
struct ftrace_hash *new_hash;
int size = ftrace_hash_empty(direct_functions) ? 0 :
direct_functions->count + 1;
if (size < 32)
size = 32;
new_hash = dup_hash(direct_functions, size);
if (!new_hash)
return NULL;
*free_hash = direct_functions;
direct_functions = new_hash;
}
The "*free_hash = direct_functions;" can happen twice, losing the previous
allocation of direct_functions.
But this also exposed a more serious bug.
The modification of direct_functions above is not safe. As
direct_functions can be referenced at any time to find what direct caller
it should call, the time between:
new_hash = dup_hash(direct_functions, size);
and
direct_functions = new_hash;
can have a race with another CPU (or even this one if it gets interrupted),
and the entries being moved to the new hash are not referenced.
That's because the "dup_hash()" is really misnamed and is really a
"move_hash()". It moves the entries from the old hash to the new one.
Now even if that was changed, this code is not proper as direct_functions
should not be updated until the end. That is the best way to handle
function reference changes, and is the way other parts of ftrace handles
this.
The following is done:
1. Change add_hash_entry() to return the entry it created and inserted
into the hash, and not just return success or not.
2. Replace ftrace_add_rec_direct() with add_hash_entry(), and remove
the former.
3. Allocate a "new_hash" at the start that is made for holding both the
new hash entries as well as the existing entries in direct_functions.
4. Copy (not move) the direct_function entries over to the new_hash.
5. Copy the entries of the added hash to the new_hash.
6. If everything succeeds, then use rcu_pointer_assign() to update the
direct_functions with the new_hash.
This simplifies the code and fixes both the memory leak as well as the
race condition mentioned above.
Link: https://lore.kernel.org/all/170368070504.42064.8960569647118388081.stgit@devnote2/
Link: https://lore.kernel.org/linux-trace-kernel/20231229115134.08dd5174@gandalf.local.home
Cc: stable@vger.kernel.org
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Fixes: 763e34e74bb7d ("ftrace: Add register_ftrace_direct()")
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
kernel/trace/ftrace.c | 100 +++++++++++++++++++++++---------------------------
1 file changed, 47 insertions(+), 53 deletions(-)
--- a/kernel/trace/ftrace.c
+++ b/kernel/trace/ftrace.c
@@ -1183,18 +1183,19 @@ static void __add_hash_entry(struct ftra
hash->count++;
}
-static int add_hash_entry(struct ftrace_hash *hash, unsigned long ip)
+static struct ftrace_func_entry *
+add_hash_entry(struct ftrace_hash *hash, unsigned long ip)
{
struct ftrace_func_entry *entry;
entry = kmalloc(sizeof(*entry), GFP_KERNEL);
if (!entry)
- return -ENOMEM;
+ return NULL;
entry->ip = ip;
__add_hash_entry(hash, entry);
- return 0;
+ return entry;
}
static void
@@ -1349,7 +1350,6 @@ alloc_and_copy_ftrace_hash(int size_bits
struct ftrace_func_entry *entry;
struct ftrace_hash *new_hash;
int size;
- int ret;
int i;
new_hash = alloc_ftrace_hash(size_bits);
@@ -1366,8 +1366,7 @@ alloc_and_copy_ftrace_hash(int size_bits
size = 1 << hash->size_bits;
for (i = 0; i < size; i++) {
hlist_for_each_entry(entry, &hash->buckets[i], hlist) {
- ret = add_hash_entry(new_hash, entry->ip);
- if (ret < 0)
+ if (add_hash_entry(new_hash, entry->ip) == NULL)
goto free_hash;
}
}
@@ -2536,7 +2535,7 @@ ftrace_find_unique_ops(struct dyn_ftrace
#ifdef CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS
/* Protected by rcu_tasks for reading, and direct_mutex for writing */
-static struct ftrace_hash *direct_functions = EMPTY_HASH;
+static struct ftrace_hash __rcu *direct_functions = EMPTY_HASH;
static DEFINE_MUTEX(direct_mutex);
int ftrace_direct_func_count;
@@ -2555,39 +2554,6 @@ unsigned long ftrace_find_rec_direct(uns
return entry->direct;
}
-static struct ftrace_func_entry*
-ftrace_add_rec_direct(unsigned long ip, unsigned long addr,
- struct ftrace_hash **free_hash)
-{
- struct ftrace_func_entry *entry;
-
- if (ftrace_hash_empty(direct_functions) ||
- direct_functions->count > 2 * (1 << direct_functions->size_bits)) {
- struct ftrace_hash *new_hash;
- int size = ftrace_hash_empty(direct_functions) ? 0 :
- direct_functions->count + 1;
-
- if (size < 32)
- size = 32;
-
- new_hash = dup_hash(direct_functions, size);
- if (!new_hash)
- return NULL;
-
- *free_hash = direct_functions;
- direct_functions = new_hash;
- }
-
- entry = kmalloc(sizeof(*entry), GFP_KERNEL);
- if (!entry)
- return NULL;
-
- entry->ip = ip;
- entry->direct = addr;
- __add_hash_entry(direct_functions, entry);
- return entry;
-}
-
static void call_direct_funcs(unsigned long ip, unsigned long pip,
struct ftrace_ops *ops, struct ftrace_regs *fregs)
{
@@ -4223,8 +4189,8 @@ enter_record(struct ftrace_hash *hash, s
/* Do nothing if it exists */
if (entry)
return 0;
-
- ret = add_hash_entry(hash, rec->ip);
+ if (add_hash_entry(hash, rec->ip) == NULL)
+ ret = -ENOMEM;
}
return ret;
}
@@ -5266,7 +5232,8 @@ __ftrace_match_addr(struct ftrace_hash *
return 0;
}
- return add_hash_entry(hash, ip);
+ entry = add_hash_entry(hash, ip);
+ return entry ? 0 : -ENOMEM;
}
static int
@@ -5410,7 +5377,7 @@ static void remove_direct_functions_hash
*/
int register_ftrace_direct(struct ftrace_ops *ops, unsigned long addr)
{
- struct ftrace_hash *hash, *free_hash = NULL;
+ struct ftrace_hash *hash, *new_hash = NULL, *free_hash = NULL;
struct ftrace_func_entry *entry, *new;
int err = -EBUSY, size, i;
@@ -5436,17 +5403,44 @@ int register_ftrace_direct(struct ftrace
}
}
- /* ... and insert them to direct_functions hash. */
err = -ENOMEM;
+
+ /* Make a copy hash to place the new and the old entries in */
+ size = hash->count + direct_functions->count;
+ if (size > 32)
+ size = 32;
+ new_hash = alloc_ftrace_hash(fls(size));
+ if (!new_hash)
+ goto out_unlock;
+
+ /* Now copy over the existing direct entries */
+ size = 1 << direct_functions->size_bits;
+ for (i = 0; i < size; i++) {
+ hlist_for_each_entry(entry, &direct_functions->buckets[i], hlist) {
+ new = add_hash_entry(new_hash, entry->ip);
+ if (!new)
+ goto out_unlock;
+ new->direct = entry->direct;
+ }
+ }
+
+ /* ... and add the new entries */
+ size = 1 << hash->size_bits;
for (i = 0; i < size; i++) {
hlist_for_each_entry(entry, &hash->buckets[i], hlist) {
- new = ftrace_add_rec_direct(entry->ip, addr, &free_hash);
+ new = add_hash_entry(new_hash, entry->ip);
if (!new)
- goto out_remove;
+ goto out_unlock;
+ /* Update both the copy and the hash entry */
+ new->direct = addr;
entry->direct = addr;
}
}
+ free_hash = direct_functions;
+ rcu_assign_pointer(direct_functions, new_hash);
+ new_hash = NULL;
+
ops->func = call_direct_funcs;
ops->flags = MULTI_FLAGS;
ops->trampoline = FTRACE_REGS_ADDR;
@@ -5454,17 +5448,17 @@ int register_ftrace_direct(struct ftrace
err = register_ftrace_function_nolock(ops);
- out_remove:
- if (err)
- remove_direct_functions_hash(hash, addr);
-
out_unlock:
mutex_unlock(&direct_mutex);
- if (free_hash) {
+ if (free_hash && free_hash != EMPTY_HASH) {
synchronize_rcu_tasks();
free_ftrace_hash(free_hash);
}
+
+ if (new_hash)
+ free_ftrace_hash(new_hash);
+
return err;
}
EXPORT_SYMBOL_GPL(register_ftrace_direct);
@@ -6309,7 +6303,7 @@ ftrace_graph_set_hash(struct ftrace_hash
if (entry)
continue;
- if (add_hash_entry(hash, rec->ip) < 0)
+ if (add_hash_entry(hash, rec->ip) == NULL)
goto out;
} else {
if (entry) {
next prev parent reply other threads:[~2024-01-03 17:14 UTC|newest]
Thread overview: 72+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-01-03 16:55 [PATCH 6.6 00/49] 6.6.10-rc1 review Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 01/49] ksmbd: Remove unused field in ksmbd_user struct Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 02/49] ksmbd: reorganize ksmbd_iov_pin_rsp() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 03/49] ksmbd: fix kernel-doc comment of ksmbd_vfs_setxattr() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 04/49] ksmbd: fix missing RDMA-capable flag for IPoIB device in ksmbd_rdma_capable_netdev() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 05/49] ksmbd: add support for surrogate pair conversion Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 06/49] ksmbd: no need to wait for binded connection termination at logoff Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 07/49] ksmbd: fix kernel-doc comment of ksmbd_vfs_kern_path_locked() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 08/49] ksmbd: prevent memory leak on error return Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 09/49] ksmbd: separately allocate ci per dentry Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 10/49] ksmbd: move oplock handling after unlock parent dir Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 11/49] ksmbd: release interim response after sending status pending response Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 12/49] ksmbd: move setting SMB2_FLAGS_ASYNC_COMMAND and AsyncId Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 13/49] ksmbd: dont update ->op_state as OPLOCK_STATE_NONE on error Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 14/49] ksmbd: set epoch in create context v2 lease Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 15/49] ksmbd: set v2 lease capability Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 16/49] ksmbd: downgrade RWH lease caching state to RH for directory Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 17/49] ksmbd: send v2 lease break notification " Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 18/49] ksmbd: lazy v2 lease break on smb2_write() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 19/49] ksmbd: avoid duplicate opinfo_put() call on error of smb21_lease_break_ack() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 20/49] fs: new accessor methods for atime and mtime Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 21/49] client: convert to new timestamp accessors Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 22/49] fs: cifs: Fix atime update check Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 23/49] virtio_ring: fix syncs DMA memory with different direction Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 24/49] kexec: fix KEXEC_FILE dependencies Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 25/49] kexec: select CRYPTO from KEXEC_FILE instead of depending on it Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 26/49] linux/export: Fix alignment for 64-bit ksymtab entries Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 27/49] linux/export: Ensure natural alignment of kcrctab array Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 28/49] mptcp: refactor sndbuf auto-tuning Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 29/49] mptcp: fix possible NULL pointer dereference on close Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 30/49] mptcp: fix inconsistent state on fastopen race Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 31/49] block: renumber QUEUE_FLAG_HW_WC Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 32/49] platform/x86/intel/pmc: Add suspend callback Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 33/49] platform/x86/intel/pmc: Allow reenabling LTRs Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 34/49] platform/x86/intel/pmc: Move GBE LTR ignore to suspend callback Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 35/49] ksmbd: fix slab-out-of-bounds in smb_strndup_from_utf16() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 36/49] platform/x86: p2sb: Allow p2sb_bar() calls during PCI device probe Greg Kroah-Hartman
2024-01-04 9:01 ` Shinichiro Kawasaki
2024-01-03 16:55 ` [PATCH 6.6 37/49] maple_tree: do not preallocate nodes for slot stores Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 38/49] selftests: secretmem: floor the memory size to the multiple of page_size Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 6.6 39/49] mm/filemap: avoid buffered read/write race to read inconsistent data Greg Kroah-Hartman
2024-01-03 16:56 ` [PATCH 6.6 40/49] mm: migrate high-order folios in swap cache correctly Greg Kroah-Hartman
2024-01-03 16:56 ` [PATCH 6.6 41/49] mm/memory-failure: cast index to loff_t before shifting it Greg Kroah-Hartman
2024-01-03 16:56 ` [PATCH 6.6 42/49] mm/memory-failure: check the mapcount of the precise page Greg Kroah-Hartman
2024-01-03 16:56 ` [PATCH 6.6 43/49] Revert "nvme-fc: fix race between error recovery and creating association" Greg Kroah-Hartman
2024-01-03 16:56 ` [PATCH 6.6 44/49] ring-buffer: Fix wake ups when buffer_percent is set to 100 Greg Kroah-Hartman
2024-01-03 16:56 ` Greg Kroah-Hartman [this message]
2024-01-03 16:56 ` [PATCH 6.6 46/49] tracing: Fix blocked reader of snapshot buffer Greg Kroah-Hartman
2024-01-03 16:56 ` [PATCH 6.6 47/49] wifi: cfg80211: fix CQM for non-range use Greg Kroah-Hartman
2024-01-03 16:56 ` [PATCH 6.6 48/49] wifi: nl80211: fix deadlock in nl80211_set_cqm_rssi (6.6.x) Greg Kroah-Hartman
2024-01-03 16:56 ` [PATCH 6.6 49/49] netfilter: nf_tables: skip set commit for deleted/destroyed sets Greg Kroah-Hartman
2024-01-03 17:44 ` [PATCH 6.6 00/49] 6.6.10-rc1 review Nam Cao
2024-01-03 18:57 ` SeongJae Park
2024-01-03 22:04 ` Florian Fainelli
2024-01-03 23:35 ` Kelsey Steele
2024-01-04 0:18 ` Shuah Khan
2024-01-04 2:24 ` Takeshi Ogasawara
2024-01-04 4:10 ` Daniel Díaz
2024-01-04 7:15 ` Daniel Díaz
2024-01-04 7:58 ` Greg Kroah-Hartman
2024-01-04 8:21 ` Johannes Berg
2024-01-04 12:39 ` Naresh Kamboju
2024-01-04 12:58 ` Greg Kroah-Hartman
2024-01-04 5:20 ` Bagas Sanjaya
2024-01-04 7:55 ` Luna Jernberg
2024-01-04 7:57 ` Greg Kroah-Hartman
2024-01-04 10:26 ` Ron Economos
2024-01-04 11:53 ` Harshit Mogalapalli
2024-01-04 16:52 ` Jon Hunter
2024-01-04 17:12 ` Allen
2024-01-05 1:04 ` Guenter Roeck
2024-01-05 2:43 ` Namjae Jeon
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=20240103164841.969983746@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=ast@kernel.org \
--cc=daniel@iogearbox.net \
--cc=jolsa@kernel.org \
--cc=mark.rutland@arm.com \
--cc=mathieu.desnoyers@efficios.com \
--cc=mhiramat@kernel.org \
--cc=patches@lists.linux.dev \
--cc=rostedt@goodmis.org \
--cc=stable@vger.kernel.org \
/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