|
|
BFX-AT100_PhantomII_issue_capture.mp4 (4,633,192 bytes) |
|
|
Hi Kinbay And I don't know how it relates to the BT audio loss issue and RCU(BT). If you know anything, please share it. Thank you YK.Jung |
|
|
Hi Kinbay SKB is requesting a schedule for this issue. Please update the schedule Thank you YK.Jung |
|
|
Hi,Younkwang Jung Could you catch and attach the bitsnoop hci log && log information for below case? Thank you. Smart3 + RCU(BA02) + BT(Phantom II) speaker : audio loss issue was observed Smart3 + BT(Phantom II) speaker : no problem |
|
|
Hi Da-Wen we are unable to capture the bitsnoop hci log. ( I'll let you know when we can capture it ) Currently, this issue is occurring when the BT RCU and speaker are connected at the same time. Please check the difference in A2DP audio transmission(BT) when only speaker is connected and when both are connected. And also check which part affects the TX queue overflow.) Thank you YK.Jung |
|
|
Hi Younkwang Thanks for your suggestion, we will check these parts. BTW, "2024-11-12_BTAudio.zip" log in jira is for the " Smart3 + RCU(BA02) + BT(Phantom II) speaker" ,right? This log.txt shows that after the info(btif_a2dp_source_enqueue_callback: TX queue buffer size) occurs, it will reads data such as RSSI (the value is 0?), which aligns with the btsnoop_hci.log. However, we haven't found any anomalies from the btsnoop log. Do you have any insights on this? image.png (52,863 bytes) image-2.png (41,629 bytes) image-3.png (48,139 bytes) |
|
|
Hi Da-Wen >> BTW, "2024-11-12_BTAudio.zip" log in jira is for the " Smart3 + RCU(BA02) + BT(Phantom II) speaker" ,right? No Phantom II cannot be rented and we have to make a reservation at the store for a test. it is too difficult to test. So Checking all BT audio devices owned by ALT The issue of "TX queue buffer size" was observed in one of them , so we're checking with the BT device first because It's a similar symptom. and I've checked the logs, but I haven't found any clear clues yet. Additionally, the issue is not being observed in INTEK Smart3 on the BT audio device. I don't know why RCU's BT connection affects a2dp Please let me know if you have any clue Thank you YK.Jung |
|
|
Hi,Younkwang We have checked the logs, and found some clues: 1. When "TX queue buffer size" occured, it read RSSI value and the value is 0. As we understand, this value is abnormal. 2. "btif_a2dp_source_read_callback: UNDERFLOW: ONLY READ 1412 BYTES OUT OF 4096" might case "TX queue buffer size". The system audio cannot be sent to BT in time. This problem is handled by audio hal, and it need Amlogic to help analyze this part. Best Regards. Liuling Wang |
|
|
Hi Liuling The analysis result of SoC is BT data is not transmitted or is delayed So SoC is proposing a review of BT controller or BT chipset performance or data transfer speed. Please review soc's request And this issue is not observed on INTEK Smart3 that use the same BT Stack and the same audio hal drivser . ( The BT chipset used by INTEK Smart3 is REALTEK , BT5.0 ) Thank you YK.Jung |
|
|
Hi Liuling SoC's comments have been updated on BPM-36973 https://jira.skbroadband.com/browse/BPM-36973 If there is any confirmed information, please let me know. - Is it a BT performance issue? BT chipset performance or data transfer speed.. etc.. - why RCU's BT connection affects a2dp I need to update the progress by DEC/4th Thank you YK.Jung |
|
|
Hi,Younkwang We have checked "2024-11-12_BTAudio log.txt", and found that the audio codec is SBC and the delay is 5000/10ms. We are analyzing the reason and the audio codec in our side is also SBC but the delay is 0/10ms. Best Regards. Liuling Wang image1.PNG (44,862 bytes) image2.PNG (34,546 bytes) |
|
|
Hi Liuling Wang Thank you for updating the contents. Please let me know if there is anything additionally confirmed. Thank you YK.Jung |
|
|
Hi,Younkwang We have checked 2024-11-12_BTAudio log.txt and btsnoop_hci.log, found that the deplay(500ms) is sent by speaker. Best Regards. Liuling Wang image2-2.png (43,777 bytes) image1-2.png (57,426 bytes) |
|
|
Hi Liuling Thank you for updating I'm checking some things based on the information you gave me. 1) Effect of remote_delay_report_ For certain BT(SBC codec) speaker, the remote_delay_report_value is 1500. For this BT speaker, if I repeat sleep/wakeup, audio loss issue was observed. So the value of remote_delay_report_ was forcibly set to 0 and tested. Test results: Audio loss issue is observed for both delay 0 and 1500. As a result of the test, the remote_delay_report_ value has an effect on sync and does not seem to be related to the audio loss issue. 2) I checked rssi / Tx Power periodically. The test results are consistent with rssi: 0 / Tx Power: 12. // It is outputting a certain value regardless of whether an issue occurs 12-05 13:57:26.271 W 3981 4480 bt_btif_a2dp_source: system/bt/btif/src/btif_a2dp_source.cc:1370 btm_read_rssi_cb: btm_read_rssi_cb: device: 00:42:79:a0:ec:50, rssi: 0 12-05 13:57:26.277 W 3981 4480 bt_btif_a2dp_source: system/bt/btif/src/btif_a2dp_source.cc:1411 btm_read_tx_power_cb: btm_read_tx_power_cb: device: 00:42:79:a0:ec:50, Tx Power: 12 Logs are as follows ============================================================================== > BT speaker [enter sleep] .. 12-05 13:56:48.552 D 4719 4719 BluetoothConnectionManager: [ mBluetoothReceiver ] action= android.bluetooth.device.action.ACL_DISCONNECTED, JBL Flip 4 , deviceBoundState= 12, deviceConnectionState= -1, profileState= -1, profiles= [A2DP_SINK, HFP, HSP] 12-05 13:56:48.552 D 4719 4719 BluetoothConnectionManager: [ mBluetoothReceiver ] ACTION_ACL_DISCONNECTED remove device. .. > BT speaker [wakeup] 12-05 13:57:17.895 D 4719 4719 BluetoothConnectionManager: [ mBluetoothReceiver ] action= android.bluetooth.a2dp.profile.action.CONNECTION_STATE_CHANGED, JBL Flip 4 , deviceBoundState= 12, deviceConnectionState= -1, profileState= 2, profiles= [A2DP_SINK, HFP, HSP] 12-05 13:57:17.895 D 4719 4719 BluetoothConnectionManager: BluetoothA2dp.ACTION_CONNECTION_STATE_CHANGED 12-05 13:57:17.895 I 4719 4719 BluetoothConnectionManager: BluetoothA2dp A2DP State: 2 .. * No audio issue .. > BT speaker [enter sleep] 12-05 13:57:44.003 D 4719 4719 BluetoothConnectionManager: [ mBluetoothReceiver ] action= android.bluetooth.device.action.ACL_DISCONNECTED, JBL Flip 4 , deviceBoundState= 12, deviceConnectionState= -1, profileState= -1, profiles= [A2DP_SINK, HFP, HSP] 12-05 13:57:44.003 D 4719 4719 BluetoothConnectionManager: [ mBluetoothReceiver ] ACTION_ACL_DISCONNECTED remove device. .. > BT speaker [wakeup] 12-05 13:58:08.261 D 4719 4719 BluetoothConnectionManager: [ mBluetoothReceiver ] action= android.bluetooth.a2dp.profile.action.CONNECTION_STATE_CHANGED, JBL Flip 4 , deviceBoundState= 12, deviceConnectionState= -1, profileState= 2, profiles= [A2DP_SINK, HFP, HSP] 12-05 13:58:08.261 D 4719 4719 BluetoothConnectionManager: BluetoothA2dp.ACTION_CONNECTION_STATE_CHANGED 12-05 13:58:08.261 I 4719 4719 BluetoothConnectionManager: BluetoothA2dp A2DP State: 2 .. * audio loss issue was observed .. > Battery is removed from RCU and disconnected 12-05 13:59:09.596 D 4719 4719 BluetoothConnectionManager: [ mBluetoothReceiver ] action= android.bluetooth.device.action.ACL_DISCONNECTED, BMM_BA01_A6C , deviceBoundState= 12, deviceConnectionState= -1, profileState= -1, profiles= [] 12-05 13:59:09.632 D 4719 4719 BluetoothConnectionManager: [ mBluetoothReceiver ] ACTION_ACL_DISCONNECTED remove device. .. * No audio issue .. ================================================================================== Please check and update me if you have any additional information Thank you YK.Jung |
|
|
Hi Liu-Ling Wang I have additional questions. 1) The following logs are being observed. 12-05 13:56:48.552 D 4719 4719 BluetoothConnectionManager: [ mBluetoothReceiver ] action= android.bluetooth.device.action.ACL_DISCONNECTED, JBL Flip 4 , deviceBoundState= 12, deviceConnectionState= -1, profileState= -1, profiles= [A2DP_SINK, HFP, HSP] > Do you know where "BluetoothConnectionManager" logs are outputting? I searched it, but I can't find it > The profiles information is displayed like [A2DP_SINK, HFP, HSP]. Is it possible to remove HFP and HSP from profiles information? 2) Is there a parameter that affects the A2DP data transmission when STB connects to RCUs? please check below ticket https://btracker-cns.fii-foxconn.com/vaas/view.php?id=1985 If you have any doubts or confirmations, please update us. SKB needs continuous updates. Thank you YK.Jung |
|
|
Hi,Younkwang 1) We do not know where "BluetoothConnectionManager" logs are outputting. 2) If you want to remove HFP and HSP from profiles information, maybe you can try like https://btracker-cns.fii-foxconn.com/vaas/view.php?id=2010#c17052. 3) From 2024-12-05_Test_logcat\2024-12-05_SBC_Codec_logcat.log, we found BMM_Audio_BA01 profiles is A2DP_SINK. 12-05 13:56:00.789 4719 4719 I BluetoothDeviceUtils: isAudioDevice: true, BMM_Audio_BA01, majorDeviceClass: AUDIO_VIDEO(1024), deviceClass: AUDIO_VIDEO_LOUDSPEAKER(1044), profiles: [A2DP_SINK] Best Regards. Liuling Wang |
|
|
Hi Liuling >> 1) We do not know where "BluetoothConnectionManager" logs are outputting. Please also check whether it is FXN's library or not. >> 2) If you want to remove HFP and HSP from profiles information, maybe you can try like https://btracker-cns.fii-foxconn.com/vaas/view.php?id=2010#c17052. It is already applied to the final FW. Even if it is applied, the issue is still being observed. Please contact BT chip company about audio loss issue , And please also contact BT chip company for this issue and let me know the feedback. FYI , SKB is asking for a schedule for this issue. Thank you YK.Jung |
|
|
Hi Liuling 1) I will update the results of the FW test 15.543.64(ATV12) : audio loss issue was observed after BT device was repeated sleep/wakeup 15.542.130(ATV12): audio loss issue was observed after BT device was repeated sleep/wakeup 15.541.66(ATV10) : audio loss issue was observed after BT device was repeated sleep/wakeup 15.540.16(ATV10) : audio loss issue was observed after BT device was repeated sleep/wakeup 15.532.20(ATV10) : audio loss issue was observed after BT device was repeated sleep/wakeup 15.520.8(ATV9) : audio loss issue was observed but too difficult to observe the issue (only observed once) Attached are two logs. - 15.542.36_log.zip : The BT device was repeated sleep/wakeup and audio loss issue was finally observed - 15.520.8_log.zip : The BT device was repeated sleep/wakeup and no issue was observed > compare the case where the issue occurs and the case where it does not occur. (Is there any difference ?) 2) please check the relationship between RCU BT connection and audio loss issue (the attached video and log.) 15.542.36_RCU_disconnect.zip 20241210_audio_loss_and_rcu.mp4 Please update me if there is anything or any clues. Thank you YK.Jung 20241210_audio_loss_and_rcu.mp4 (20,492,640 bytes) |
|
|
Hi Liuling Please update me on the progress and if there is anything confirmed Thank you YK.Jung |
|
|
Hi Liuling It's been more than a month since the issue was reported, but the cause has not been identified yet. SKB is requesting a quick resolution. 1) Please update us on the current analysis. 2) Did you report this issue to BT chipset? 3) Please give us your opinion on how to proceed to resolve this issue Thank you YK.Jung |
|
|
Hi Liuling If the BT speaker is repeatedly sleep/wakeup, an audio loss issue is being observed. I enabled the log of BT stack and tested it again. 1) About btsnoop_hci_2.log ==> BT speaker wakeup, no audio loss was observed after BT speaker connection (wakeup) 19120 6023.465403 controller host HCI_EVT 13 Rcvd Connect Request 19121 6023.466137 host controller HCI_CMD 11 Sent Accept Connection Request 19122 6023.469223 controller host HCI_EVT 7 Rcvd Command Status (Accept Connection Request) 19123 6023.645452 controller host HCI_EVT 11 Rcvd Role Change 19124 6023.655495 controller host HCI_EVT 14 Rcvd Connect Complete 19125 6023.656062 host controller HCI_CMD 8 Sent Write Link Policy Settings .. ==> BT speaker wakeup, audio loss was observed after BT BT speaker connection (wakeup) 19816 6050.389356 controller host HCI_EVT 13 Rcvd Connect Request 19817 6050.389993 host controller HCI_CMD 11 Sent Accept Connection Request 19818 6050.391976 controller host HCI_EVT 7 Rcvd Command Status (Accept Connection Request) 19819 6050.583579 controller host HCI_EVT 11 Rcvd Role Change 19820 6050.592049 controller host HCI_EVT 14 Rcvd Connect Complete * Please check btsnoop_hci for anything that could be a clue. 2) After activating the log of the BT stack, (logcat.log.0001) If an issue occurs, the num_left value remains as shown below. .. 12-12 13:40:57.475 I 3925 4410 bt_l2cap: system/bt/main/bte_logmsg.cc:201 LogMsg: L2CA_FlushChannel() flushed: 0 + 0, num_left: 3 12-12 13:40:57.507 I 3925 4410 bt_l2cap: system/bt/main/bte_logmsg.cc:201 LogMsg: L2CA_FlushChannel() flushed: 0 + 0, num_left: 3 12-12 13:40:57.507 I 3925 4410 bt_l2cap: system/bt/main/bte_logmsg.cc:201 LogMsg: L2CA_FlushChannel() flushed: 0 + 0, num_left: 4 12-12 13:40:57.539 I 3925 4410 bt_l2cap: system/bt/main/bte_logmsg.cc:201 LogMsg: L2CA_FlushChannel() flushed: 0 + 0, num_left: 2 12-12 13:40:57.539 I 3925 4410 bt_l2cap: system/bt/main/bte_logmsg.cc:201 LogMsg: L2CA_FlushChannel() flushed: 0 + 0, num_left: 3 .. Please check if this part is also related to the audio disconnection issue. Please update me on my previous request. https://btracker-cns.fii-foxconn.com/vaas/view.php?id=2011#c17159 Thank you YK.Jung |
|
|
Hi,Younkwang Sorry for the late reply. We checked FXN's library and could not find "BluetoothConnectionManager". We are analyzing these logs but have no clues. We have reported this issue to BT chipset and are awaiting a response. Best Regards. Liuling Wang |
|
|
Hi Liuling Thank you for update Please let me know if you have any feedback or updates. Thank you YK.Jung |
|
|
Hi,Younkwang Will you help capture btsnoop and sniffer logs? BT chipset need these logs to analize. Thank you. Best Regards. Liuling Wang |
|
|
Hi Liuling Did you check the log I attached (0017160) ? It includes the btsnoop And we don't have BT air log capture so I can't collect that air log. FYI, it may be good to check the BT Audio devices FXN have ( There may be a device where this issue is observed ) And SKB is requesting the schedule. Please update me. Thank you YK.Jung |
|
|
Hi Liuling I am checking btsnoop file again(0017160) There is a difference between when it is an issue or ok at HCI ACL Packet. when there is a problem, the source role is slave. OK case : 2024-12-17_ok_case.png issue case : 2024-12-17_issue_case.png Please check if this difference is related to the issue Thank you YK.Jung 2024-12-17_issue_case.png (172,405 bytes) 2024-12-17_ok_case.png (177,247 bytes) |
|
|
Hi Liuling I received a strong complaint from SKB. SKB has been asked to solve it by this month. It takes more than three days to test. Therefore, you have to solve it by next Tuesday. Please update me on every day that you checked and analyzed. Thank you YK.Jung |
|
|
Hi Liuling Please update the progress within an hour. Thank you YK.Jung |
|
|
Hi,Younkwang Sorry for the late reply. 1. BT chipset has created a case to analyze it. 2. So far we have no clues. Best Regards. Liuling Wang |
|
|
Hi Liuling I reported to SKB about the current situation. We have to solve it by next Tuesday. Please push to BT chipset and update me if there is anything confirmed Thank you YK.Jung |
|
|
Hi Liuling Please update the progress within an hour. Thank you YK.Jung |
|
|
Hi,Younkwang We have pushed to BT chipset. If there is anything confirmed, I will update you. Best Regards. Liuling Wang |
|
|
Hi Liuling I have a meeting with SKB today. 16:00 (KT) Therefore, please update the progress by 15:00 (KT) - what fxn checked, what was confirmed - feedback of BT chipset And please check if there is any way to fix the source role of STB as master. (0002011:0017170) Thank you. YK.Jung |
|
|
Hi,Younkwang BT chipset still analyzing. We checked logs, found that Controller sent Role Change command to Host. We are checking the reason. in logcat.log.0001 The JBL role is changed: peripheral -> central -> peripheral 12-12 13:40:29.958 3925 4410 I bt_bta_dm: system/bt/bta/dm/bta_dm_act.cc:2335 handle_role_change: Role change callback peer:xx:xx:xx:xx:ec:50 info:0x10 new_role:peripheral dev count:2 hci_status:Success 12-12 13:40:30.210 3925 4410 I bt_bta_dm: system/bt/bta/dm/bta_dm_act.cc:2335 handle_role_change: Role change callback peer:xx:xx:xx:xx:ec:50 info:0x10 new_role:central dev count:2 hci_status:Success 12-12 13:40:56.831 3925 4410 I bt_bta_dm: system/bt/bta/dm/bta_dm_act.cc:2335 handle_role_change: Role change callback peer:xx:xx:xx:xx:ec:50 info:0x30 new_role:peripheral dev count:2 hci_status:Success Before role change, the out mps is 0. 12-12 13:40:27.457 3925 4410 I bt_l2cap: system/bt/main/bte_logmsg.cc:201 LogMsg: l2cu_adjust_out_mps use 0 Based on peer_cfg.fcr.mps: 0 packet_size: 1011 12-12 13:40:27.471 3925 4410 I bt_l2cap: system/bt/main/bte_logmsg.cc:201 LogMsg: l2cu_adjust_out_mps use 0 Based on peer_cfg.fcr.mps: 0 packet_size: 1011 12-12 13:40:29.521 3925 4410 I bt_l2cap: system/bt/main/bte_logmsg.cc:201 LogMsg: l2cu_adjust_out_mps use 0 Based on peer_cfg.fcr.mps: 0 packet_size: 1011 Best Regards. Liuling Wang |
|
|
Hi Liuling On the schedule, there should be a rough solution today. Please update the progress within an hour. Thank you YK.Jung |
|
|
Hi,Younkwang We have pushed BT chipset to give a rough solution. Best Regards. Liuling Wang |
|
|
Hi Liuling SKB requested to fix this issue a long time ago , but we still haven't solved it. Based on the current status, it seems like it won't be possible to solve it within this year So SKB is constantly complaining to us. Please let me know if you have any comments on how we can make this issue resolved and update the progress within two hour. One more thing Who is the BT chipset you are in contact with? Please add me to the mail loop when contacting BT chipset. Thank you YK.Jung |
|
|
Hi,Younkwang We are in contact with Zenitron FAE Brad. And he created a case abaut this issue infineon website, the link is https://mycases.infineon.com/techsupport-edit/?id=d3ea127b-25bd-ef11-95f5-000d3a227332 Best Regards. Liuling Wang |
|
|
Hi Liuling I explained the current situation to SKB.. SKB is still complaining to us. First of all, please find with a solution as soon as possible. and I tried to access the site, but I couldn't access it. ( attached image.png) please check it Thank you YK.Jung image-4.png (32,327 bytes) |
|
|
Hi,Younkwang Infineon responded as follows: From the HCI log, we can see that the audio stops three times. 1. Packet 0000608~0000695, one time it is the suspend command actively sent by the upper layer protocol stack.(first-image.png) 2. Packet #19096~#19311, second time the speaker is disconnected automatically.(second-image.png) 3. Packet #19791~#20038, third time is the same as the second time, the speaker disconnected automatically.(third-image.png) From the logcat log, no sound refered to the last two times. Both times, the speaker actively disconnected, and the packets before the disconnection were normal. To solve this problem, we need to see why the speaker actively shuts down the AVDTP link after connecting the remote control. There should be no problem on the CYW20704 side. Best Regards. Liuling Wang first-image.png (54,745 bytes) second-image.png (94,259 bytes) third-image.png (94,515 bytes) |
|
|
Hi,Younkwang Brad added your email to the permissions, please check whether you can browse the case. Best Regards. Liuling Wang |
|
|
Hi Liuling I will explain once again how to reproduce the issue with "JBL Flip 4" when BT speaker on/off is repeated, this issue was observed. When I checked the log, it seems that you delivered the log file of 0002011:0017160. I will explain the btsnoop_hci_2.log file again. start_packet end_packet duration 1) 427 19095 288.000380 ( check audio , result : Issue not observed ) 2) Press the power button on the BT speaker to enter the BT speaker into off state 3) Press the power button on the BT speaker to enter the BT speaker into on state 4) 19316 19790 8.930888 ( check audio , result : Issue not observed ) 5) Press the power button on the BT speaker to enter the BT speaker into off state 6) Press the power button on the BT speaker to enter the BT speaker into on state 7) 20043 23448 103.671155 ( check audio , result : Issue being observed ) * The speaker disconnects because the BT speaker is turned off. * Issue is observed in 7). So far, the confirmation is The reason for the audio loss issue is that full occurs in the tx buffer in the BT stack. logcat.log.0001 : image1.png btsnoop_hci_2.log : image2.png When tx buffer is full in bt stack, tx buffer clear is done. In that case, these logs are observed in btsnoop. I think the cause of this issue is slow audio transmission when STB and BT are connected And one of the things that I suspect is BT speaker is Master if an issue is observed. image3.png Please request an analysis once again. In addition, please check with BT chipset if there have been any similar issues related to this. I tried logging in, but I couldn't understand the language of the issue page. I will monitor that page Thank you YK.Jung |
|
|
Hi,Younkwang BT chipset said that Switching role does not affect audio transmission. We are looking for a way to disable role switch to verify it. Do you have a way to collect the air logs? The air logs can help BT chipset to analyze this issue. They don't have any clues yet. Best Regards. Liuling Wang |
|
|
Hi Liuling >> BT chipset said that Switching role does not affect audio transmission. If I check the BT related content, it is affecting the transmission speed when it is master and slave. Master: Control the clock and frequency allocation of the Bluetooth piconet Set bandwidth Manage multiple slave devices Slave: Data is sent at the request of the Master I understand that the transmission speed is affected by BT speaker when STB is slave * Please let me know if there is anything I misunderstand about BT master/slave The current issue is occurring when the BT rcu and BT speaker are connected to the STB(slave) at the same time. >> Do you have a way to collect the air logs? The air logs can help BT chipset to analyze this issue. They don't have any clues yet. We don't have air logs capture equipment, we can't provide that information. Do FXN have air log capture equipment? and one more thing Is there a way to monitor the status of bt chipset? Thank you YK.Jung |
|
|
Hi Liuling SKB is asking us to find a solution about this issue by this week I know it's difficult situation, but please check it quickly. Thank you YK.Jung |
|
|
Hi Liuling I have received a new BT(AI700)speaker from SKB. The STB is continuously set to slave when connected to the BT(AI700) speaker. audio loss issue was observed [STB (BFX-AT100) + BT rcu + BT speaker] And when the BT RCU is disconnected, the audio is ok. [STB (BFX-AT100) + BT speaker] video : AI700_BT_SpeakerTest.mp4 logs : 2025-01-08_AI700_logs.zip ( include btsnoop / logcat.log / dumpsys ) capture : image.png And if I check the Bt stack code and change it as below, Then STB is always set to master. =============================================================== diff --git a/stack/acl/btm_acl.cc b/stack/acl/btm_acl.cc index 62a69870a..11f0f97ca 100644 --- a/stack/acl/btm_acl.cc +++ b/stack/acl/btm_acl.cc @@ -648,6 +648,14 @@ static void check_link_policy(tLINK_POLICY* settings) { *settings &= (~HCI_ENABLE_CENTRAL_PERIPHERAL_SWITCH); LOG_INFO("Role switch not supported (settings: 0x%04x)", *settings); } + + /* jglory test modify */ + if (1) { + *settings &= (~HCI_ENABLE_CENTRAL_PERIPHERAL_SWITCH); + LOG_INFO("jglory forced Role switch not supported (settings: 0x%04x)", *settings); + } + + if ((*settings & HCI_ENABLE_HOLD_MODE) && (!controller->supports_hold_mode())) { *settings &= (~HCI_ENABLE_HOLD_MODE); ==================================================================== As a result of testing so far,audio loss is not observed after the above test patch is applied Attach Log File ( 2025-01-08_always_set_master.zip ) However, I don't know what side effects there will be if we modify it as above. And In common sense, the device that transmits audio data should be the master.. It is strange that STB, which is an audio data transmission device, is set to slave. Please discuss that with Infineon. Thank you YK.Jung image-5.png (69,351 bytes) AI700_BT_SpeakerTest.mp4 (13,821,447 bytes) |
|
|
Hi Liuling When the FW including the test patch(0002011:0017241) is tested in QA, but it is sometimes set to slave. ( audio loss issue was also being observed.) That is, the test code above is not a WK solution. and Is there anything updated? Thank you YK.Jung image-6.png (131,968 bytes) |
|
|
Hi,Younkwang Infineon said the connection interval is 11.25ms, he suggested setting it to larger value for testing. capture : connection_interval.png Please modify it as set_connection_interval.jpg Best Regards. Liuling Wang connection_interval.png (164,037 bytes) set_connection_interval.jpg (142,909 bytes) |
|
|
Hi Liuling It is being tested as the guide you told me. But the issue seems to be observed. anyway I'll check more and update you And I can log-in to infineon site and see two issues. https://mycases.infineon.com/# - IFX-241218-1548924 - IFX-241008-1473117 Please add "jhjung@altech.kr" to the infineon tickets Thank you YK.Jung |
|
|
Hi,Younkwang Please use this mail "jhjung@altech.kr" to register and login https://mycases.infineon.com/ , the FAE add it fail. Best Regards. Liuling Wang |
|
|
Hi Liuling >> Please use this mail "jhjung@altech.kr" to register and login https://mycases.infineon.com/ , the FAE add it fail. please add it again Thank you YK.Jung |
|
|
Hi Liuling I tested the patch you mentioned. (https://btracker-cns.fii-foxconn.com/vaas/view.php?id=2011#c17246) - one more thing , in the part you told me, this part needs to be modified. BTM_BLE_CONN_INT_MIN_LIMIT is already defined in bdrop_buildcfg.h, so I deleted it. then the define below will be applied. // system/bt/stack/include/btm_ble_api_types.h #ifndef BTM_BLE_CONN_INT_MIN_LIMIT ( <= already defined in bdrop_buildcfg.h ) /* #define BTM_BLE_CONN_INT_MIN_LIMIT 0x0009 */ #define BTM_BLE_CONN_INT_MIN_LIMIT 24 #endif ================================================================================== diff --git a/wifi_bt/bluetooth/broadcom/btbuild/bdroid_buildcfg.h b/wifi_bt/bluetooth/broadcom/btbuild/bdroid_buildcfg.h index 51f893b..5cd3a85 100644 --- a/wifi_bt/bluetooth/broadcom/btbuild/bdroid_buildcfg.h +++ b/wifi_bt/bluetooth/broadcom/btbuild/bdroid_buildcfg.h @@ -40,7 +40,7 @@ #define BTM_BLE_CONN_TIMEOUT_DEF 500 //VaS Kerwin modified end for UEI BLE RCU, 2019/04/24 /* minimum acceptable connection interval */ -#define BTM_BLE_CONN_INT_MIN_LIMIT 0x0006 /*7.5ms=6*1.25*/ +/* #define BTM_BLE_CONN_INT_MIN_LIMIT 0x0006 */ /*7.5ms=6*1.25*/ #define BTM_BLE_SCAN_SLOW_INT_1 512 ================================================================================== As a result of the patch test, it seems that the audio loss issue is improved. - I'll have to do a little more testing at tomorrow However, when the BA01 RCU is connected, an unstable RCU connection is being observed. ( sometime BA01 RCU is unpaired , RCU key operation has slowed down ) I think we need to improve it or find another solution Thank you YK.Jung |
|
|
Hi Liuling I received a strong complaint from SKB today. We haven't found a solution to this issue yet. - It's been more than three months since the issue was reported Please support us quickly and check the contents. And SKB wants to know this clearly. Audio data is transmitted from the BFX-AT100 to the BT audio device At this time, sometimes it is set as follows. - Source Role of BFX-AT100 is slave - Source Role of BT audio device is master Please check whether the above status is a problem or not by tomorrow SKB manager's thinking is that BFX-AT100 should always be a master. And If there is a way to make BFX-AT100 a master all the time, please share it. Thank you YK.Jung |
|
|
Hi Liuling Please update the progress by 4:30 p.m (KT) today. Thank you YK.Jung |
|
|
Hi Younkwang >>And SKB wants to know this clearly. Audio data is transmitted from the BFX-AT100 to the BT audio device At this time, sometimes it is set as follows. - Source Role of BFX-AT100 is slave - Source Role of BT audio device is master Please check whether the above status is a problem or not by tomorrow I asked IFX and have not replyed yet. As far as I know, BFX-AT100 should always be a master, it is responsible for scanning devices and initiating connection requests. FAE added this mail "jhjung@altech.kr" to the cases permissions,please check it. Best Regards. Liuling Wang |
|
|
Hi Liuling Audio loss issue is also observed when the "connection interval" is 30ms (24 * 1.25) . so it's not solution When the source role of BFX-AT100 is master, the issue is not observed So please check that the source role of BFX-AT100 is always set to master Thank you YK.Jung |
|
|
Hi Liuling Please check the following items. > What the infineon needs to check - He should check why the BTRCU(ble) connection is associated with the BT audio loss issue. Is it a performance issue or a chipset issue.. BT version(?) - is there a way to monitor the all status of the BT controller. > What the FXN needs to check (WK) - Set the source role of the BFX-AT100 to always be the master when sending ACL packets It's BT stack, so FXN have to check it But best solution should be resolved by infineon. Please update the progress by 4:30 p.m. today. Thank you YK.Jung |
|
|
Hi Younkwang The IFX engineer has been on a business trip this week. He suggested checking the logic in the BT stack and disable the role switch. Best Regards. Liuling Wang |
|
|
Hi Liuling Please update the information you are analyzing by 4:30 p.m. today. - let me know in detail which side you are analyzing Thank you YK.Jung |
|
|
Hi Younkwang I am analyzing the connection request and link policy setting. I found the link policy setting packet has Role Switch property. when it was disabled , the issue was observed. Best Regards. Liuling Wang |
|
|
Hi Liuling I did a test with AISoundMax. Although it is set to slave, audio loss issue is not observed. Of course, the AISoundMax's bt_chipset is different That is Audio loss issue should not be observed even if STB is set to slave. And when transferring the audio data of BFX-AT100, I'm guessing the transfer speed will be slow when it's a slave. ( There is a difference between the master state and the slave state. ) first , please clarify the relationship between BT Rcu(ble) and audio transmission(A2DP). IFX engineer will be able to check using airlog. Please forward this information to IFX engineer. Thank you YK.Jung image-7.png (136,680 bytes) |
|
|
Hi Younkwang I found that STB is set to slave and the packet has not 'Disconnect in frame' when audio loss issue is observed. I think 'Disconnect in frame' is a clue, but IFX does not think that. Best Regards. Liuling Wang |
|
|
Hi Liuling Please explain "Disconnect in frame" in detail, I don't know what it is. And I checked the BID-AT200 btsnoop log of the competitor's product. When testing with BID-AT200, BID-AT200 is always connected to Master when transmitting data. If I look at the log (flip_log) , it's looks like below ( attached image.png ) ========================================================== Rcvd Role Change .. Sent Switch Role <--- it's different .. ========================================================== That is , BID-AT200 seems to switch back(Master) to "Sent Switch Role" when it gets "Rcvd Role Change". Please check it Thank you YK.Jung image-8.png (68,380 bytes) |
|
|
Hi Liuling I'm sorry. I misreported the information to you (#c17275) The "Sent Switch Role" is also being observed at BFX-AT100. Please review the btsnoop log of BID-AT200 to see what is different from BFX-AT100. Why does BID-AT200 always work as a master Thank you YK.Jung |
|
|
Hi Liuling Please update the progress by 4:30 p.m (KT) Thank you YK.Jung |
|
|
Hi Younkwang I think 'Disconnect in frame' maybe is not a clue, becasue I found BID-AT200 packet also has not it. In BID-AT200, It is Role Change -> Switch Role -> Role Change, and then the STB is master. In BFX-AT100, It is Role Change -> Switch Role, and the STB is slave. Before Role Change, host sent twice command 'Write link policy settings' to controller, their settings are slightly different. Best Regards. Liuling Wang BFX-AT100-sbc.png (79,877 bytes) UA300_B69A-sbc.png (115,989 bytes) BID-AT200-sbc.png (138,108 bytes) BFX-AT100-role_change.png (103,291 bytes) BID-AT200-role_change.png (157,099 bytes) BFX-AT100-1.png (134,615 bytes) BFX-AT100-2.png (118,253 bytes) BID-AT200-1.png (189,763 bytes) BID-AT200-2.png (172,255 bytes) |
|
|
Hi Liuling Please update the progress by 4:30 p.m (KT) Thank you YK.Jung |
|
|
Hi Younkwang I reproduced STB is set to slave, audio loss issue is not observed. But when I turned off/on speaker again, issue is observed. So Audio loss issue may not be observed even if STB is set to slave with BFX-AT100 Best Regards. Liuling Wang |
|
|
Hi Liuling Could you please share the btsnoop log file and dumpsys bluetooth_manager ? If it's set to slave, it looks like the processing A2dp speed/Performance has dropped. At this time, when the RCU is connected, an audio loss issue is observed. Of course, if the RCU is disconnected, the issue of audio loss is not being observed. I'm not sure exactly, but if STB is set to slave, there seems to be a possibility of a problem. (Of course, I don't think it's 100%.) Therefore, we are considering changing to a master (WK) If you have any other opinions, please let me know. Thank you YK.Jung |
|
|
Hi Younkwang In btsnoop log packet#12266, STB is slave and 'TX queue buffer size' is not observed in logcat log. In btsnoop log packet#29526, STB is slave and 'TX queue buffer size' is observed in logcat log. The audio loss issue is slightly observed. Best Regards. Liuling Wang |
|
|
Hi Liuling I don't know exactly, but I think there are cases where audio loss can't be observed even though it's a slave. And SKB asks us to update the progress on this issue. - Please update what FXN is currently reviewing. - Please check what the infineon is reviewing Please update the progress by 3:00 p.m (KT) Thank you YK.Jung |
|
|
Hi all I updated the progrerss with infineon Thank you YK.Jung image-9.png (86,156 bytes) image-10.png (55,934 bytes) |
|
|
I updated the progrerss with infineon |
|
|
I updated the progrerss with infineon |
|
|
update with infineon image-12.png (172,393 bytes) |
|
|
This is the progress with Infineon. image-13.png (185,808 bytes) |
|
|
This is the progress with infineon (IFX-241218-1548924) image-14.png (221,773 bytes) |
|
|
This is the progress with infineon (IFX-241218-1548924) image-15.png (153,274 bytes) image-16.png (245,152 bytes) |
|
|
This is the progress with infineon (IFX-241218-1548924) |
|
|
This is the progress with infineon (IFX-241218-1548924) |
- yk.jung ( (ALTech) Younkwang Jung ) developer
.png)
.png)
.png)
.png)
.png)
.png)
