[PATCH] Bluetooth: MGMT: Return RFKILLED for Set Powered on a blocked controller

[PATCH] Bluetooth: MGMT: Return RFKILLED for Set Powered on a blocked controller

[Tobias Wilken]

Since MGMT_OP_SET_POWERED runs through hci_cmd_sync, powering on a
controller that is blocked by rfkill fails with -ERFKILL from
hci_dev_open_sync(), and mgmt_set_powered_complete() reports it with
mgmt_status(err). mgmt_errno_status() has no case for -ERFKILL, so
userspace gets MGMT_STATUS_FAILED instead of MGMT_STATUS_RFKILLED.

Before that conversion, power on went through hci_power_on(), whose
error path mgmt_set_powered_failed() still maps -ERFKILL to
MGMT_STATUS_RFKILLED. bluetoothd relies on that status: for
RFKILLED it sets the adapter PowerState to "off-blocked" and returns
org.bluez.Error.Blocked for the Powered property, for any other status
it returns org.bluez.Error.Failed.

With the controller soft blocked, "bluetoothctl power on" printed
"Failed to set power on: org.bluez.Error.Failed" and btmon showed
"Set Powered (0x0005) Status: Failed (0x03)". With this change it
prints org.bluez.Error.Blocked and btmon shows "Status: RFKilled
(0x12)".

Map -ERFKILL to MGMT_STATUS_RFKILLED in mgmt_errno_status().

Fixes: cf75ad8b41d2 ("Bluetooth: hci_sync: Convert MGMT_SET_POWERED")
Assisted-by: LLM
Signed-off-by: Tobias Wilken <tooangel@tooangel.de>
---

Notes:
    Tested on Debian's 7.2.9 kernel with a soft blocked MediaTek MT7922
    (btusb) and on bluetooth-next in test-runner. mgmt-tester passes
    515/515 with this patch; the new test "Set powered on - Blocked by
    rfkill" fails without it (Failed (0x03) instead of RFKilled (0x12)).
    The test is sent to BlueZ separately as "mgmt-tester: Add Set Powered
    test for rfkill block".
    
    Not tested with a hard (hardware) rfkill block.
    
    This is my first kernel patch, so I'm happy about any feedback, also
    on whether it fits the conventions here.

 net/bluetooth/mgmt.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c
index 3b9c82bd2f0..0f93e9e2728 100644
--- a/net/bluetooth/mgmt.c
+++ b/net/bluetooth/mgmt.c
@@ -304,6 +304,8 @@ static u8 mgmt_errno_status(int err)
 		return MGMT_STATUS_DISCONNECTED;
 	case -ECANCELED:
 		return MGMT_STATUS_CANCELLED;
+	case -ERFKILL:
+		return MGMT_STATUS_RFKILLED;
 	}
 
 	return MGMT_STATUS_FAILED;
-- 
2.47.3

Re: [PATCH] Bluetooth: MGMT: Return RFKILLED for Set Powered on a blocked controller

[patchwork-bot+bluetooth]

Hello:

This patch was applied to bluetooth/bluetooth-next.git (master)
by Luiz Augusto von Dentz <luiz.von.dentz@intel.com>:

On Wed, 21 Oct 2026 08:58:33 +0200 you wrote:
> Since MGMT_OP_SET_POWERED runs through hci_cmd_sync, powering on a
> controller that is blocked by rfkill fails with -ERFKILL from
> hci_dev_open_sync(), and mgmt_set_powered_complete() reports it with
> mgmt_status(err). mgmt_errno_status() has no case for -ERFKILL, so
> userspace gets MGMT_STATUS_FAILED instead of MGMT_STATUS_RFKILLED.
> 
> Before that conversion, power on went through hci_power_on(), whose
> error path mgmt_set_powered_failed() still maps -ERFKILL to
> MGMT_STATUS_RFKILLED. bluetoothd relies on that status: for
> RFKILLED it sets the adapter PowerState to "off-blocked" and returns
> org.bluez.Error.Blocked for the Powered property, for any other status
> it returns org.bluez.Error.Failed.
> 
> [...]

Here is the summary with links:
  - Bluetooth: MGMT: Return RFKILLED for Set Powered on a blocked controller
    https://git.kernel.org/bluetooth/bluetooth-next/c/65683fbed5c8

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html