Skip to contentSkip to conversation

a2dp: Fix UAF after rejected SetConfiguration

PATCH Applied Luiz Augusto von Dentz 3 participants 5 messages Oct 7 – Oct 8 Bluetooth drivers

lore
Share & tools
lore

Luiz Augusto von Dentz

author maintainer
#0 lore raw

PATCH v2 1/3 a2dp: Fix UAF after rejected SetConfiguration

From: Eduardo Alves <[email protected]>

When the MediaEndpoint1 rejects SetConfiguration, auto_config() calls setconf_cb() with an error and avdtp frees the avdtp_stream, but the a2dp_stream wrapping it is left on sep->streams with a dangling stream pointer. The next stream configured on the same session finds it in a2dp_config() and reads the freed avdtp_stream:

ERROR: AddressSanitizer: heap-use-after-free READ of size 4
#0 avdtp_stream_get_state profiles/audio/avdtp.c:3952 #1 a2dp_config profiles/audio/a2dp.c:3282
#2 select_complete profiles/audio/source.c:200 #3 finalize_select profiles/audio/a2dp.c:486
freed by thread T0 here:
#1 stream_free profiles/audio/avdtp.c:747
#2 setconf_cb profiles/audio/avdtp.c:1504
#3 auto_config profiles/audio/a2dp.c:752
#4 endpoint_setconf_cb profiles/audio/a2dp.c:768

Fix it by destroying the a2dp_stream when the configuration is rejected, which also drops the session reference it holds.

This can be reproduced with test-functional using two VMs over btvirt, where the A2DP Sink endpoint replies to SetConfiguration with an error and the sink then connects to the source on the same session.

Assisted-by Claude:claude-opus-5-5
---
 profiles/audio/a2dp.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)
profiles/audio/a2dp.c +9 -1
index 3c991de6acb7..08a9e1c96e4f 100644
@@ -748,9 +748,17 @@ static gboolean auto_config(gpointer data)
 	}
 
 done:
-	if (setup->setconf_cb)
+	if (setup->setconf_cb) {
+		/* Rejecting the configuration frees the avdtp_stream */
+		if (setup->err)
+			a2dp_stream_destroy(setup->sep, setup->stream);
+
 		setup->setconf_cb(setup->session, setup->stream, setup->err);
 
+		if (setup->err)
+			setup->stream = NULL;
+	}
+
 	finalize_config(setup);
 
 	setup_error_set(setup, NULL);
Signature · 1 line
2.55.0
author maintainer
#1 lore raw

PATCH v2 2/3 a2dp: Fix crash on NULL session in auto_config

From: Eduardo Alves <[email protected]>

If the AVDTP channel is disconnected while bluetoothd is still waiting for the MediaEndpoint1 to reply to SetConfiguration, channel_free() sets setup->session to NULL but the setup stays alive since the pending endpoint request holds a reference to it. Once the endpoint replies, or the call times out and the request is canceled, endpoint_setconf_cb() calls auto_config() which passes the NULL session to avdtp_get_device() and then dereferences the resulting NULL device:

bluetoothd[820]: profiles/audio/media.c:endpoint_reply() Endpoint
replied with an error: org.freedesktop.DBus.Error.NoReply
kernel: bluetoothd[820]: segfault at 1a0 ip 00005633dd741323

    sp 00007ffe51923650 error 4 in bluetoothd
  #0  auto_config (data=0x56340361a120) at profiles/audio/a2dp.c:723
  #1  endpoint_setconf_cb (setup=...) at profiles/audio/a2dp.c:768
  #2  media_endpoint_cancel (request=...) at profiles/audio/media.c:208
  #3  media_endpoint_cancel_all () at profiles/audio/media.c:216
  #4  clear_endpoint () at profiles/audio/media.c:379
  #5  endpoint_reply (user_data=...) at profiles/audio/media.c:407

(gdb) p *setup
$1 = {chan = 0x0, session = 0x0, ...,

        setconf_cb = 0x5633dd74c780 <setconf_cb>, ..., ref = 2}

Commit 14750a2e48f4 ("audio/a2dp: Fix Access session device only when its valid") fixed the same crash, but commit 77932f2dac1a ("profiles/audio: add nullity checks") moved avdtp_get_device() back ahead of the checks. Just reordering is not enough though: the a2dp_stream is still on sep->streams, so the aborted check passes and setconf_cb() would then be called with a NULL session as well.

Fix it by making finalize_all() abort the pending Set Configuration while the session is still valid: clearing the endpoint configuration sends ClearConfiguration, removes the MediaTransport created for the request and cancels the request, so auto_config() rejects the configuration and frees the pending avdtp_stream and a2dp_stream, both previously leaked together with the session reference held by the latter. If no request is pending, the configuration is rejected directly. auto_config() bails out when setup->session is NULL, as it may still run from idle, and setconf_cb is cleared once called so it cannot be called twice.

Removing a transport now cancels the endpoint requests pending for it, so a late reply is ignored instead of leaving the transport registered, which made the next connection fail with "Resource temporarily unavailable".

Assisted-by Claude:claude-opus-5-5
---
 profiles/audio/a2dp.c  | 132 +++++++++++++++++++++++++++--------------
 profiles/audio/media.c |  20 ++++++-
 2 files changed, 106 insertions(+), 46 deletions(-)
profiles/audio/a2dp.c +87 -45
index 08a9e1c96e4f..37f6d4ae12ac 100644
@@ -507,41 +507,6 @@ static void finalize_discover(struct a2dp_setup *s)
 	}
 }
 
-static gboolean finalize_all(gpointer data)
-{
-	struct a2dp_setup *s = data;
-	struct avdtp_stream *stream = s->err ? NULL : s->stream;
-	GSList *l;
-
-	for (l = s->cb; l != NULL; ) {
-		struct a2dp_setup_cb *cb = l->data;
-
-		l = l->next;
-
-		if (cb->discover_cb) {
-			cb->discover_cb(s->session, s->seps,
-					error_to_errno(s->err), cb->user_data);
-		} else if (cb->select_cb) {
-			cb->select_cb(s->session, s->sep, s->caps,
-					error_to_errno(s->err), cb->user_data);
-		} else if (cb->suspend_cb) {
-			cb->suspend_cb(s->session,
-					error_to_errno(s->err), cb->user_data);
-		} else if (cb->resume_cb) {
-			cb->resume_cb(s->session,
-					error_to_errno(s->err), cb->user_data);
-		} else if (cb->config_cb) {
-			cb->config_cb(s->session, s->sep, stream,
-					error_to_errno(s->err), cb->user_data);
-		} else
-			warn("setup_cb doesn't have any callback function");
-
-		setup_cb_free(cb);
-	}
-
-	return FALSE;
-}
-
 static struct a2dp_setup *find_setup_by_session(struct avdtp *session)
 {
 	GSList *l;
@@ -711,6 +676,86 @@ static void stream_state_changed(struct avdtp_stream *stream,
 		sep->endpoint->clear_configuration(sep, dev, sep->user_data);
 }
 
+static void setup_setconf_reply(struct a2dp_setup *setup,
+					struct avdtp_error *err)
+{
+	avdtp_set_configuration_cb cb = setup->setconf_cb;
+
+	if (!cb)
+		return;
+
+	setup->setconf_cb = NULL;
+
+	/* Rejecting the configuration frees the avdtp_stream */
+	if (err)
+		a2dp_stream_destroy(setup->sep, setup->stream);
+
+	cb(setup->session, setup->stream, err);
+
+	if (err)
+		setup->stream = NULL;
+}
+
+/* Reject a pending Set Configuration while setup->session is still valid */
+static void setup_abort_setconf(struct a2dp_setup *setup)
+{
+	struct a2dp_sep *sep = setup->sep;
+	struct avdtp_error err;
+
+	if (!setup->setconf_cb)
+		return;
+
+	/* Clearing the endpoint configuration cancels its pending request
+	 * which rejects the configuration via auto_config().
+	 */
+	if (sep->endpoint && sep->endpoint->clear_configuration)
+		sep->endpoint->clear_configuration(sep,
+					avdtp_get_device(setup->session),
+					sep->user_data);
+
+	/* Reject it if it was not pending on the endpoint */
+	avdtp_error_init(&err, AVDTP_MEDIA_CODEC,
+					AVDTP_UNSUPPORTED_CONFIGURATION);
+	setup_setconf_reply(setup, &err);
+}
+
+static gboolean finalize_all(gpointer data)
+{
+	struct a2dp_setup *s = data;
+	struct avdtp_stream *stream = s->err ? NULL : s->stream;
+	GSList *l;
+
+	for (l = s->cb; l != NULL; ) {
+		struct a2dp_setup_cb *cb = l->data;
+
+		l = l->next;
+
+		if (cb->discover_cb) {
+			cb->discover_cb(s->session, s->seps,
+					error_to_errno(s->err), cb->user_data);
+		} else if (cb->select_cb) {
+			cb->select_cb(s->session, s->sep, s->caps,
+					error_to_errno(s->err), cb->user_data);
+		} else if (cb->suspend_cb) {
+			cb->suspend_cb(s->session,
+					error_to_errno(s->err), cb->user_data);
+		} else if (cb->resume_cb) {
+			cb->resume_cb(s->session,
+					error_to_errno(s->err), cb->user_data);
+		} else if (cb->config_cb) {
+			cb->config_cb(s->session, s->sep, stream,
+					error_to_errno(s->err), cb->user_data);
+		} else
+			warn("setup_cb doesn't have any callback function");
+
+		setup_cb_free(cb);
+	}
+
+	setup_abort_setconf(s);
+
+	return FALSE;
+}
+
 static gboolean auto_config(gpointer data)
 {
 	struct a2dp_setup *setup = data;
@@ -718,6 +763,12 @@ static gboolean auto_config(gpointer data)
 	struct btd_service *service;
 	struct a2dp_stream *stream;
 
+	/* Check if the channel has been disconnected, in which case
+	 * channel_free() has already rejected the configuration.
+	 */
+	if (!setup->session)
+		goto done;
+
 	dev = avdtp_get_device(setup->session);
 
 	if (setup->sep->type == AVDTP_SEP_TYPE_SOURCE)
@@ -748,16 +799,7 @@ static gboolean auto_config(gpointer data)
 	}
 
 done:
-	if (setup->setconf_cb) {
-		/* Rejecting the configuration frees the avdtp_stream */
-		if (setup->err)
-			a2dp_stream_destroy(setup->sep, setup->stream);
-
-		setup->setconf_cb(setup->session, setup->stream, setup->err);
-
-		if (setup->err)
-			setup->stream = NULL;
-	}
+	setup_setconf_reply(setup, setup->err);
 
 	finalize_config(setup);
 
profiles/audio/media.c +19 -1
index e8418280e60b..34ff251c7ef2 100644
@@ -359,12 +359,26 @@ static struct media_adapter *find_adapter(struct btd_device *device)
 	return NULL;
 }
 
+static int request_transport_cmp(gconstpointer data, gconstpointer user_data)
+{
+	const struct endpoint_request *request = data;
+
+	return request->transport == user_data ? 0 : -1;
+}
+
 static void endpoint_remove_transport(struct media_endpoint *endpoint,
 					struct media_transport *transport)
 {
+	GSList *l;
+
 	if (!endpoint || !transport)
 		return;
 
+	/* Cancel pending requests for the transport */
+	while ((l = g_slist_find_custom(endpoint->requests, transport,
+						request_transport_cmp)))
+		media_endpoint_cancel(l->data);
+
 	endpoint->transports = g_slist_remove(endpoint->transports, transport);
 	media_transport_destroy(transport);
 }
@@ -431,7 +445,11 @@ static void endpoint_reply(DBusPendingCall *call, void *user_data)
 		if (dbus_message_is_method_call(request->msg,
 					MEDIA_ENDPOINT_INTERFACE,
 					"SetConfiguration")) {
-			endpoint_remove_transport(endpoint, request->transport);
+			struct media_transport *transport = request->transport;
+
+			/* Detach so the request is not canceled */
+			request->transport = NULL;
+			endpoint_remove_transport(endpoint, transport);
 			error_code = a2dp_parse_config_error(err.name);
 			ret = &error_code;
 			size = 1;
Signature · 1 line
2.55.0
author maintainer
#2 lore raw

PATCH v2 3/3 test: Cover A2DP disconnection during SetConfiguration

In reply to Luiz Augusto von Dentz

From: Luiz Augusto von Dentz <[email protected]>

Add test_a2dp_disconnect_during_setconf, where the source disconnects while the sink endpoint has not replied to SetConfiguration yet, and the late reply, either accepting or rejecting, must be ignored and the stream configured again on reconnection.

Assisted-by Claude:claude-opus-5-5
---
 doc/functional-a2dp.rst      | 37 +++++++++++++++++++++++
 test/functional/test_a2dp.py | 58 ++++++++++++++++++++++++++++++++++++
 2 files changed, 95 insertions(+)
doc/functional-a2dp.rst +37 -0
index 03b18a7201c2..f38a458bb7c2 100644
@@ -97,3 +97,40 @@ test_a2dp_transport_acquire
 
 :Notes: Acquiring on the source side starts the stream, so the
 	peripheral does not have to acquire its own transport.
+
+test_a2dp_disconnect_during_setconf[accept|reject]
+--------------------------------------------------
+
+:Setup: As above, except the peripheral registers its A2DP Sink
+	endpoint manually with ``Auto Accept: no``, so that `bluetoothd`
+	is left waiting for the reply to
+	``org.bluez.MediaEndpoint1.SetConfiguration``.
+
+:Steps:
+	1. Pair and trust as above.
+	2. Central: ``connect <peripheral bdaddr>``.
+	3. Peripheral: wait for the ``Accept (yes/no):`` prompt, without
+	   answering it.
+	4. Central: ``disconnect <peripheral bdaddr>``.
+	5. Peripheral: answer the pending prompt with ``yes``
+	   (``accept``) or ``no`` (``reject``).
+	6. Central: ``connect <peripheral bdaddr>`` again, and the
+	   peripheral answers ``yes`` to the new prompt.
+
+:Expected:
+	1. ``Pairing successful`` and ``trust succeeded`` on both hosts.
+	2. The peripheral receives ``Endpoint: SetConfiguration``.
+	3. The prompt is shown.
+	4. ``Disconnection successful`` on the central, and the transport
+	   created on the peripheral for the pending configuration is
+	   removed (``[DEL] Transport``).
+	5. The late reply is ignored.
+	6. ``Connection successful``, and a transport appears on both
+	   hosts.
+
+:Notes: Regression test for a crash of `bluetoothd` on the peripheral:
+	the setup of the pending configuration outlived the AVDTP session
+	and the late reply dereferenced a NULL session in
+	``auto_config()``. A late ``yes`` used to also leave the stale
+	transport registered, failing the next connection with
+	``Resource temporarily unavailable``.
test/functional/test_a2dp.py +58 -0

61 lines not on this page · raw message

Signature · 1 line
2.55.0
#3 lore raw
In reply to Luiz Augusto von Dentz

This is automated email and please do not reply to this email!

Dear submitter,

Thank you for submitting the patches to the linux bluetooth mailing list. This is a CI test results with your patch series:
PW Link:https://patchwork.kernel.org/series/1181105/

---Test result---

Test Summary:

CheckPatch                    PASS      1.62 seconds
GitLint                       PASS      0.69 seconds
BuildEll                      PASS      13.67 seconds
BluezMake                     PASS      213.68 seconds
MakeCheck                     PASS      13.06 seconds
MakeDistcheck                 PASS      94.41 seconds
CheckValgrind                 PASS      142.41 seconds
CheckSmatch                   PASS      180.59 seconds
bluezmakeextell               PASS      66.42 seconds
TestFunctional                PASS      813.30 seconds
IncrementalBuild              PASS      220.31 seconds
ScanBuild                     PASS      594.43 seconds

https://github.com/bluez/bluez/pull/2633

---
Regards,
Linux Bluetooth

#4 lore raw
In reply to Luiz Augusto von Dentz
This series was applied to bluetooth/bluez.git (master) by Luiz Augusto von Dentz <[email protected]>:

Hello:

This series was applied to bluetooth/bluez.git (master) by Luiz Augusto von Dentz <[email protected]>:

Quoted 9 lines · On Wed, 7 Oct 2026 15:52:04 -0400 you wrote:

From: Eduardo Alves <[email protected]>

When the MediaEndpoint1 rejects SetConfiguration, auto_config() calls setconf_cb() with an error and avdtp frees the avdtp_stream, but the a2dp_stream wrapping it is left on sep->streams with a dangling stream pointer. The next stream configured on the same session finds it in a2dp_config() and reads the freed avdtp_stream:

[...]

Here is the summary with links:
- [BlueZ,v2,1/3] a2dp: Fix UAF after rejected SetConfiguration

    https://git.kernel.org/pub/scm/bluetooth/bluez.git/?id=ef54133c69e9
  - [BlueZ,v2,2/3] a2dp: Fix crash on NULL session in auto_config
    https://git.kernel.org/pub/scm/bluetooth/bluez.git/?id=79aedd33f8fa
  - [BlueZ,v2,3/3] test: Cover A2DP disconnection during SetConfiguration
    https://git.kernel.org/pub/scm/bluetooth/bluez.git/?id=ab5df7ba3019

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html