Linux/WL1805MOD: hciattach fails often (often means 1 out of 2) when WiFi is active !

Noel Vellemans

Prodigy 230 points

Part Number: WL18xxMOD

Tool/software: Linux

Hello,

We are experiencing a strange issue, when we enable WiFi+Bluetooth ( at the same time)

WiFi operation alone , works , Bluetooth alone works fine.

BUT When we are using Wifi and we enable Bluetooth (by means of enabling hardware enable pin and attach Bluez via hciattach, after that WiFi is already running and actively connected ! ) at that time hciattach bails out ( literally 1 on 2 hciattach operations works )

/usr/sbin/hciattach /dev/ttymxc3 texas -s 115200

hcInitialization timed out.

When looking at the serial lines ( with logger , when ever haci attach fails I don't get the '0x04' byte where read_hci_event is waiting on !

COM30 [0004]: 01 01 10 00

Timeout !!!!!

Nothing is rxd from the WL18xx module !

At the second hciattach .. all is working fine !

/usr/sbin/hciattach /dev/ttymxc3 texas -s 115200

Found a Texas Instruments' chip!
Firmware file : /lib/firmware/TIInit_11.8.32.bts
Loaded BTS script version 1
texas: changing baud rate to 3000000, flow control to 1

COM30 [0001]: 01
COM30 [0003]: 01 10 00
COM31 [0001]: 04
COM30 [0018]: A5 AA 5A 01 01 10 00 01 36 FF 04 C0 C6 2D 00 FC EC 24
COM31 [0021]: 0E 0C 01 01 10 00 06 00 00 06 0D 00 20 AC 04 0E 04 01 36 FF 00
COM30 [0001]: 84

all working fine!.

At the 3third hciattach it will fail gain , 4th will work , 5th will fail .. etc etc..

When adding debug code and looking at the logger output in the serial lines ( i must come to the same conclusion 0x04 on the read_hci_event is missing, what should come from the WL18xx module)

REMARK : This is only seen when WiFi is active !!!

Any idea , why this is going wrong in 50% of the cases !

I want to remark ( once more) ,in Bluetooth only operation it always works, and thus never fails !

Regards

Noel

over 6 years ago

0 Hari Nagalla over 6 years ago

TI__Mastermind 27080 points

How are resetting and re-initializing for each hciattach?
Can, you try resetting the BT_EN i.e toggle the BT_EN GPIO two times, with a delay in between them around 10 secs or so, before you try a new hciattach.

Thanks

0 Noel Vellemans over 6 years ago in reply to Hari Nagalla

Prodigy 230 points

Hello

I'm resetting and re-initializing by driving BT_EN ( to disabled) ( kill / abort hciattach, wait 3 seconds) and drive BT_EN (to ENABLE) launch a new hciattach.

{ PS: i've set the hciattach timeout to 3 seconds iso default 10 seconds }

all works from the first time if wifi is disabled, if wifi is enabled I have to do this TWICE.

I've tried the 10 second scenario , and for some reason it sometimes works from the first time , and sometimes it does not !

0 Hari Nagalla over 6 years ago in reply to Noel Vellemans

TI__Mastermind 27080 points

Do, you mean it the delay is long between the 2 resets , it did not work?

We, tried with the below script and it worked fine on AM437x EVM, with WiFi running and pinging a remote device..

#!/bin/bash
echo 3 > /sys/class/gpio/export
echo out > /sys/class/gpio/gpio3/direction
echo 0 > /sys/class/gpio/gpio3/value
sleep 1
echo 1 > /sys/class/gpio/gpio3/value
for i in {1..10}
do
set -x #echo on
hciconfig hci0 down
kill -9 `pgrep hciattach`
kill -9 `pgrep bluetoothd`
rmmod hci_uart
rmmod bluetooth
rmmod serdev
echo 0 > /sys/class/gpio/gpio3/value
usleep 1000
echo 1 > /sys/class/gpio/gpio3/value
usleep 5000
echo 0 > /sys/class/gpio/gpio3/value
usleep 1000
echo 1 > /sys/class/gpio/gpio3/value
usleep 5000
hciattach /dev/ttyS1 texas 3000000
hcitool scan
echo "----------------------"
echo "---iteration:$i Done--"
echo "----------------------"
set +x #echo off
done

0 Hari Nagalla over 6 years ago in reply to Hari Nagalla

TI__Mastermind 27080 points

Yes, it is recommended to reset the BT EN twice, as the above script depicts... I hope this solved your issue..

Thanks

Wi-Fi

Wi-Fi forum

Linux/WL1805MOD: hciattach fails often (often means 1 out of 2) when WiFi is active !