Stephen-F opened a new pull request, #18847:
URL: https://github.com/apache/nuttx/pull/18847
Summary
bluetooth_sendto() uses netdev_findbyindex(conn->bc_ldev + 1) to locate the
Bluetooth network device for a BTPROTO_HCI socket. netdev_findbyindex
searches
by global interface index, but bc_ldev is a 0-based count among Bluetooth
devices only. The two are only coincidentally equal when a BT device
happens
to be registered at global index 1 (immediately after loopback). On any
system
where a non-BT netdev (e.g. Ethernet, loopback) is registered before the
BT
device, bc_ldev + 1 will not match the BT device's global index and the
lookup
will silently return NULL or the wrong device.
This was noted in the original code with a /* TODO: should actually look
among
BT devices */ comment that was never resolved.
The fix replaces netdev_findbyindex with a netdev_foreach walk using a
small
private callback (bluetooth_dev_byidx_callback) that counts only
NET_LL_BLUETOOTH devices and returns the bc_ldev-th one, making the lookup
independent of global registration order.
Impact
Users of BTPROTO_HCI sockets on systems with more than one netdev (e.g.
loopback + BT, or Ethernet + BT) were silently getting a failed or
incorrect
device lookup. This fix corrects that without any API or ABI changes. No
Kconfig changes, no build system impact, no documentation changes required.
Testing
Tested on a custom ARM Cortex-M7 board (STM32H7-based) with an STM32WB BLE
co-processor connected over HCI UART. The board registers a loopback netdev
before the BT netdev, which caused netdev_findbyindex(conn->bc_ldev + 1) to
return NULL consistently. With this fix, netdev_foreach correctly locates
the
BT device by position among NET_LL_BLUETOOTH devices, and HCI socket send
succeeds.
Host: Linux x86_64, GCC ARM toolchain.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
