diff options
| author | Henri Chataing <henrichataing@google.com> | 2024-04-15 14:37:20 -0700 |
|---|---|---|
| committer | Henri Chataing <henrichataing@google.com> | 2024-04-15 15:03:18 -0700 |
| commit | 17e28905e389e0e866526ef4dd9d12018302a233 (patch) | |
| tree | 691fdbc8c24d0600d770e70899b74872742dc3d6 | |
| parent | f115bf79dcbd3f2e6e84902d18d7fd96755f2a38 (diff) | |
| download | bt-android11-gsi.tar.gz | |
RootCanal: Return correct event in response to LE Create Connection Cancelandroid11-gsi
Previously returned a Command Status event, the correct event
is HCI Command Complete, with no return parameters.
This is fixed on AOSP main, patch back-ported to android11-gsi
to address v2/android-gsi/test_mapping_presubmit_rvc test
failures / flakiness
Bug: 312314996
Bug: 312610336
Bug: 313441290
Test: TreeHugger
Change-Id: I9e7eaf4b4dcc3584fbef9e514ec8df9ccdcb1ea4
| -rw-r--r-- | vendor_libs/test_vendor_lib/model/controller/dual_mode_controller.cc | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/vendor_libs/test_vendor_lib/model/controller/dual_mode_controller.cc b/vendor_libs/test_vendor_lib/model/controller/dual_mode_controller.cc index 9ccaaff53..5ed60aea6 100644 --- a/vendor_libs/test_vendor_lib/model/controller/dual_mode_controller.cc +++ b/vendor_libs/test_vendor_lib/model/controller/dual_mode_controller.cc @@ -1602,8 +1602,8 @@ void DualModeController::LeConnectionCancel(CommandPacketView command) { gd_hci::LeConnectionManagementCommandView::Create(command)); ASSERT(command_view.IsValid()); link_layer_controller_.SetLeConnect(false); - auto packet = bluetooth::hci::LeCreateConnectionCancelStatusBuilder::Create( - ErrorCode::SUCCESS, kNumCommandPackets); + auto packet = bluetooth::hci::LeCreateConnectionCancelCompleteBuilder::Create( + kNumCommandPackets, ErrorCode::SUCCESS); send_event_(std::move(packet)); /* For testing Jakub's patch: Figure out a neat way to call this without recompiling. I'm thinking about a bad device. */ |