diff options
author | Serhii Niukalov (GitHub) <36993782+SNiukalov@users.noreply.github.com> | 2020-04-17 16:56:01 +0300 |
---|---|---|
committer | GitHub <noreply@github.com> | 2020-04-17 09:56:01 -0400 |
commit | 5220f686f8de414e2b8e281da2d573cb0ba79333 (patch) | |
tree | b67629d0e089439023608033a6157952ee838156 /src/components/application_manager/rpc_plugins/vehicle_info_plugin/test/commands/hmi/vi_get_vehicle_data_response_test.cc | |
parent | 589ac12f05b6e55cea443e1ddc693f1e6a24d236 (diff) | |
download | sdl_core-5220f686f8de414e2b8e281da2d573cb0ba79333.tar.gz |
Fix/Deadlock during transport re-connect via USB (#3332)
* Rework the use of AbortConnection for a usb connection
Sometimes we get a deadlock because we use the same lock in the same thread.
The situation is possible when calling the SendData, OnOutTransfer or AbortConnection functions.
They block the mutex at the start.
If in the PostOutTransfer function, which is called in SendData or indirectly in OnOutTransfer via PopOutMessage, we get the error LIBUSB_ERROR_NO_DEVICE,
this will call AbortConnection.
Which will cause a deadlock.
* fixup! Rework the use of AbortConnection for a usb connection
Co-authored-by: sniukalov <sniukaov@luxoft.com>
Diffstat (limited to 'src/components/application_manager/rpc_plugins/vehicle_info_plugin/test/commands/hmi/vi_get_vehicle_data_response_test.cc')
0 files changed, 0 insertions, 0 deletions