Bluetooth: hci_h4: Fix race during initialization
PATCH 1/1 Bluetooth: hci_h4: Fix race during initialization
Commit 5df5dafc171b ("Bluetooth: hci_uart: Fix another race during initialization") fixed a race for hci commands sent during initialization. However, there is still a race that happens if an hci event from one of these commands is received before HCI_UART_REGISTERED has been set at the end of hci_uart_register_dev(). The event will be ignored which causes the command to fail with a timeout in the log:
"Bluetooth: hci0: command 0x1003 tx timeout"
This is because the hci event receive path (hci_uart_tty_receive -> h4_recv) requires HCI_UART_REGISTERED to be set in h4_recv(), while the hci command transmit path (hci_uart_send_frame -> h4_enqueue) only requires HCI_UART_PROTO_INIT to be set in hci_uart_send_frame().
The check for HCI_UART_REGISTERED was originally added in commit c2578202919a ("Bluetooth: Fix H4 crash from incoming UART packets") to fix a crash caused by hu->hdev being null dereferenced. That can no longer happen: once HCI_UART_PROTO_INIT is set in hci_uart_register_dev() all pointers (hu, hu->priv and hu->hdev) are valid, and hci_uart_tty_receive() already calls h4_recv() on HCI_UART_PROTO_INIT or HCI_UART_PROTO_READY.
Remove the check for HCI_UART_REGISTERED in h4_recv() to fix the race condition.
--- drivers/bluetooth/hci_h4.c | 3 --- 1 file changed, 3 deletions(-)
drivers/bluetooth/hci_h4.c +0 -3
@@ -109,9 +109,6 @@ static int h4_recv(struct hci_uart *hu, const void *data, int count) {struct h4_struct *h4 = hu->priv; -if (!test_bit (HCI_UART_REGISTERED , &hu->flags)) -return -EUNATCH ; - h4->rx_skb =h4_recv_buf (hu, h4->rx_skb, data, count, h4_recv_pkts,ARRAY_SIZE (h4_recv_pkts));if (IS_ERR (h4->rx_skb)) {
---
base-commit: 05f7e89ab9731565d8a62e3b5d1ec206485eeb0b
change-id: 20260303-hci-init-fix-9657128a0104
Best regards,
Signature · 1 line
Jonathan Rissanen <[email protected]>
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/project/bluetooth/list/?series=1069873
---Test result---
Test Summary:
CheckPatch PENDING 0.28 seconds GitLint PENDING 0.22 seconds SubjectPrefix PASS 0.11 seconds BuildKernel PASS 26.60 seconds CheckAllWarning PASS 29.28 seconds CheckSparse PASS 27.58 seconds BuildKernel32 PASS 25.94 seconds TestRunnerSetup PASS 569.53 seconds TestRunner_l2cap-tester PASS 28.30 seconds TestRunner_iso-tester FAIL 36.61 seconds TestRunner_bnep-tester PASS 6.29 seconds TestRunner_mgmt-tester FAIL 115.04 seconds TestRunner_rfcomm-tester PASS 9.29 seconds TestRunner_sco-tester FAIL 14.40 seconds TestRunner_ioctl-tester PASS 10.10 seconds TestRunner_mesh-tester FAIL 12.55 seconds TestRunner_smp-tester PASS 8.65 seconds TestRunner_userchan-tester PASS 8.98 seconds IncrementalBuild PENDING 0.67 seconds
Details
##############################
Test: CheckPatch - PENDING
Desc: Run checkpatch.pl script
Output:
##############################
Test: GitLint - PENDING
Desc: Run gitlint
Output:
##############################
Test: TestRunner_iso-tester - FAIL
Desc: Run iso-tester with test-runner
Output:
BUG: KASAN: slab-use-after-free in le_read_features_complete+0x7e/0x2b0
Total: 141, Passed: 141 (100.0%), Failed: 0, Not Run: 0
##############################
Test: TestRunner_mgmt-tester - FAIL
Desc: Run mgmt-tester with test-runner
Output:
Total: 494, Passed: 489 (99.0%), Failed: 1, Not Run: 4
Failed Test Cases
Read Exp Feature - Success Failed 0.108 seconds
##############################
Test: TestRunner_sco-tester - FAIL
Desc: Run sco-tester with test-runner
Output:
WARNING: possible circular locking dependency detected
BUG: sleeping function called from invalid context at net/core/sock.c:3782
Total: 30, Passed: 30 (100.0%), Failed: 0, Not Run: 0
##############################
Test: TestRunner_mesh-tester - FAIL
Desc: Run mesh-tester with test-runner Output:
Total: 10, Passed: 8 (80.0%), Failed: 2, Not Run: 0
Failed Test Cases
Mesh - Send cancel - 1 Timed out 2.740 seconds Mesh - Send cancel - 2 Timed out 1.996 seconds
##############################
Test: IncrementalBuild - PENDING
Desc: Incremental build with the patches in the series Output:
---
Regards,
Linux Bluetooth
Luiz Augusto von Dentz
maintainerHi Jonathan,
Quoted 37 lines · Jonathan Rissanen
Commit
5df5dafc171b("Bluetooth:hci_uart: Fix another race during initialization") fixed a race for hci commands sent during initialization. However, there is still a race that happens if an hci event from one of these commands is received beforeHCI_UART_REGISTEREDhas been set at the end ofhci_uart_register_dev(). The event will be ignored which causes the command to fail with a timeout in the log:"Bluetooth: hci0: command
0x1003tx timeout"This is because the hci event receive path (
hci_uart_tty_receive->h4_recv) requiresHCI_UART_REGISTEREDto be set inh4_recv(), while the hci command transmit path (hci_uart_send_frame->h4_enqueue) only requiresHCI_UART_PROTO_INITto be set inhci_uart_send_frame().The check for
HCI_UART_REGISTEREDwas originally added in commitc2578202919a("Bluetooth: Fix H4 crash from incoming UART packets") to fix a crash caused byhu->hdevbeing null dereferenced. That can no longer happen: onceHCI_UART_PROTO_INITis set inhci_uart_register_dev()all pointers (hu,hu->privandhu->hdev) are valid, andhci_uart_tty_receive()already callsh4_recv()onHCI_UART_PROTO_INITorHCI_UART_PROTO_READY.Remove the check for
HCI_UART_REGISTEREDinh4_recv()to fix the race condition.Signed-off-by Jonathan Rissanen <[email protected]>--- drivers/bluetooth/hci_h4.c | 3 --- 1 file changed, 3 deletions(-)drivers/bluetooth/hci_h4.c +0
-0@@ -109,9 +109,6 @@ static int h4_recv(struct hci_uart *hu, const void *data, int count) {
drivers/bluetooth/hci_h4.c static int h4_recv(struct hci_uart *hu, const void *data, int count)
Quoted 10 lines
drivers/bluetooth/hci_h4.c +0
-4struct h4_struct *h4 = hu->priv; -if (!test_bit (HCI_UART_REGISTERED , &hu->flags)) -return -EUNATCH ; - h4->rx_skb =h4_recv_buf (hu, h4->rx_skb, data, count, h4_recv_pkts,ARRAY_SIZE (h4_recv_pkts));if (IS_ERR (h4->rx_skb)) { ---
There is some interesting comments on:
https://sashiko.dev/#/patchset/20260320-hci-init-fix-v1-1-e1960a41baf2%40axis.com
It seems the issues pointed out there are unrelated to this change, but I guess it is worth double checking just in case.
Signature · 1 line
Luiz Augusto von Dentz
Hello Luiz,
Quoted 52 lines · Luiz Augusto von Dentz
[You don't often get email from [email protected]. Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ]
Hi Jonathan,
Quoted 46 lines · Jonathan Rissanen
Commit
5df5dafc171b("Bluetooth:hci_uart: Fix another race during initialization") fixed a race for hci commands sent during initialization. However, there is still a race that happens if an hci event from one of these commands is received beforeHCI_UART_REGISTEREDhas been set at the end ofhci_uart_register_dev(). The event will be ignored which causes the command to fail with a timeout in the log:"Bluetooth: hci0: command
0x1003tx timeout"This is because the hci event receive path (
hci_uart_tty_receive->h4_recv) requiresHCI_UART_REGISTEREDto be set inh4_recv(), while the hci command transmit path (hci_uart_send_frame->h4_enqueue) only requiresHCI_UART_PROTO_INITto be set inhci_uart_send_frame().The check for
HCI_UART_REGISTEREDwas originally added in commitc2578202919a("Bluetooth: Fix H4 crash from incoming UART packets") to fix a crash caused byhu->hdevbeing null dereferenced. That can no longer happen: onceHCI_UART_PROTO_INITis set inhci_uart_register_dev()all pointers (hu,hu->privandhu->hdev) are valid, andhci_uart_tty_receive()already callsh4_recv()onHCI_UART_PROTO_INITorHCI_UART_PROTO_READY.Remove the check for
HCI_UART_REGISTEREDinh4_recv()to fix the race condition.Signed-off-by Jonathan Rissanen <[email protected]>--- drivers/bluetooth/hci_h4.c | 3 --- 1 file changed, 3 deletions(-)drivers/bluetooth/hci_h4.c +0
-3@@ -109,9 +109,6 @@ static int h4_recv(struct hci_uart *hu, const void *data, int count) {struct h4_struct *h4 = hu->priv; -if (!test_bit (HCI_UART_REGISTERED , &hu->flags)) -return -EUNATCH ; - h4->rx_skb =h4_recv_buf (hu, h4->rx_skb, data, count, h4_recv_pkts,ARRAY_SIZE (h4_recv_pkts));if (IS_ERR (h4->rx_skb)) {
Quoted 5 lines
Quoted 1 line
---
There is some interesting comments on:
https://sashiko.dev/#/patchset/20260320-hci-init-fix-v1-1-e1960a41baf2%40axis.com
I see. Yeah there seem to be some valid comments:
Quoted 3 lines
If
hci_register_dev()fails, the error path callshu->proto->close(hu)which frees the protocol-private data inhu->privand sets it to NULL. However, the error path fails to clear theHCI_UART_PROTO_INITflag.
I think regardless of this patch it makes sense to clear HCI_UART_PROTO_INIT if hci_register_dev() fails, since we're no longer in the initializing state. It becomes more important with this patch since it will lead to a null pointer dereference if h4_recv() is called after hci_register_dev() fails.
Quoted 4 lines
Additionally, since
hci_uart_register_dev()is called without holding the write-side ofhu->proto_lock, can concurrent incoming UART data during the failure window trigger a use-after-free whilehu->privis actively being freed byh4_close()?
I believe adding a write lock and clearing the bit would solve these issues:
drivers/bluetooth/hci_ldisc.c +3 -0
@@ -692,6 +692,9 @@ static int hci_uart_register_dev(struct hci_uart *hu)if (hci_register_dev (hdev) <0 ) {BT_ERR ("Can't register HCI device" ); +percpu_down_write (&hu->proto_lock); +clear_bit (HCI_UART_PROTO_INIT , &hu->flags); +percpu_up_write (&hu->proto_lock); hu->proto->close (hu); hu->hdev =NULL ;hci_free_dev (hdev);
Quoted 2 lines
It seems the issues pointed out there are unrelated to this change, but I guess it is worth double checking just in case.
I think it's best to fix it before applying this change as it can cause null pointer dereference in error path. I can send a new patchset with the fix.
Signature · 1 line
Best regards, Jonathan
Luiz Augusto von Dentz
maintainerHi Jonathan,
Quoted 86 lines · Jonathan Rissanen
Hello Luiz,
Quoted 52 lines · Luiz Augusto von Dentz
[You don't often get email from [email protected]. Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ]
Hi Jonathan,
Quoted 46 lines · Jonathan Rissanen
Commit
5df5dafc171b("Bluetooth:hci_uart: Fix another race during initialization") fixed a race for hci commands sent during initialization. However, there is still a race that happens if an hci event from one of these commands is received beforeHCI_UART_REGISTEREDhas been set at the end ofhci_uart_register_dev(). The event will be ignored which causes the command to fail with a timeout in the log:"Bluetooth: hci0: command
0x1003tx timeout"This is because the hci event receive path (
hci_uart_tty_receive->h4_recv) requiresHCI_UART_REGISTEREDto be set inh4_recv(), while the hci command transmit path (hci_uart_send_frame->h4_enqueue) only requiresHCI_UART_PROTO_INITto be set inhci_uart_send_frame().The check for
HCI_UART_REGISTEREDwas originally added in commitc2578202919a("Bluetooth: Fix H4 crash from incoming UART packets") to fix a crash caused byhu->hdevbeing null dereferenced. That can no longer happen: onceHCI_UART_PROTO_INITis set inhci_uart_register_dev()all pointers (hu,hu->privandhu->hdev) are valid, andhci_uart_tty_receive()already callsh4_recv()onHCI_UART_PROTO_INITorHCI_UART_PROTO_READY.Remove the check for
HCI_UART_REGISTEREDinh4_recv()to fix the race condition.Signed-off-by Jonathan Rissanen <[email protected]>--- drivers/bluetooth/hci_h4.c | 3 --- 1 file changed, 3 deletions(-)drivers/bluetooth/hci_h4.c +0
-3@@ -109,9 +109,6 @@ static int h4_recv(struct hci_uart *hu, const void *data, int count) {struct h4_struct *h4 = hu->priv; -if (!test_bit (HCI_UART_REGISTERED , &hu->flags)) -return -EUNATCH ; - h4->rx_skb =h4_recv_buf (hu, h4->rx_skb, data, count, h4_recv_pkts,ARRAY_SIZE (h4_recv_pkts));if (IS_ERR (h4->rx_skb)) {Quoted 5 lines
Quoted 1 line
---
There is some interesting comments on:
https://sashiko.dev/#/patchset/20260320-hci-init-fix-v1-1-e1960a41baf2%40axis.com
I see. Yeah there seem to be some valid comments:
Quoted 3 lines
If
hci_register_dev()fails, the error path callshu->proto->close(hu)which frees the protocol-private data inhu->privand sets it to NULL. However, the error path fails to clear theHCI_UART_PROTO_INITflag.I think regardless of this patch it makes sense to clear
HCI_UART_PROTO_INITifhci_register_dev()fails, since we're no longer in the initializing state. It becomes more important with this patch since it will lead to a null pointer dereference ifh4_recv()is called afterhci_register_dev()fails.Quoted 4 lines
Additionally, since
hci_uart_register_dev()is called without holding the write-side ofhu->proto_lock, can concurrent incoming UART data during the failure window trigger a use-after-free whilehu->privis actively being freed byh4_close()?I believe adding a write lock and clearing the bit would solve these issues:
drivers/bluetooth/hci_ldisc.c +0
-0
drivers/bluetooth/hci_ldisc.c static int hci_uart_register_dev(struct hci_uart *hu) · line 700
Quoted 9 lines
drivers/bluetooth/hci_ldisc.c +3
-0@@ -692,6 +692,9 @@ static int hci_uart_register_dev(struct hci_uart *hu)if (hci_register_dev (hdev) <0 ) {BT_ERR ("Can't register HCI device" ); +percpu_down_write (&hu->proto_lock); +clear_bit (HCI_UART_PROTO_INIT , &hu->flags); +percpu_up_write (&hu->proto_lock); hu->proto->close (hu); hu->hdev =NULL ;hci_free_dev (hdev);
LGTM
Quoted 8 lines
Quoted 4 lines
It seems the issues pointed out there are unrelated to this change, but I guess it is worth double checking just in case.
I think it's best to fix it before applying this change as it can cause null pointer dereference in error path. I can send a new patchset with the fix.
Please do.
Quoted 3 lines
--
Best regards, Jonathan
Signature · 1 line
Luiz Augusto von Dentz
This patch was applied to bluetooth/bluetooth-next.git (master) by Luiz Augusto von Dentz <[email protected]>:
Hello:
This patch was applied to bluetooth/bluetooth-next.git (master) by Luiz Augusto von Dentz <[email protected]>:
Quoted 8 lines · On Fri, 20 Mar 2026 13:09:49 +0100 you wrote:
Commit
5df5dafc171b("Bluetooth:hci_uart: Fix another race during initialization") fixed a race for hci commands sent during initialization. However, there is still a race that happens if an hci event from one of these commands is received beforeHCI_UART_REGISTEREDhas been set at the end ofhci_uart_register_dev(). The event will be ignored which causes the command to fail with a timeout in the log:[...]
Here is the summary with links:
- Bluetooth: hci_h4: Fix race during initialization
https://git.kernel.org/bluetooth/bluetooth-next/c/42ee54fcd11e
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html