BPI-R2 Kernel bluetooth module

frank-w · November 26, 2018, 6:52pm

started new attempt to get BT working…

currently produces this output:

most interesting should be this:

[   42.614751] [WMT-CONF][E]wmt_conf_parse_pair(323):unknown field 'mt6620.defAnt'.
[   42.622142] [WMT-CONF][E]wmt_conf_parse_pair(323):unknown field 'mt6628.defAnt'.
[   42.629537] [WMT-CONF][E]wmt_conf_parse_pair(323):unknown field 'mt6630.defAnt'.
[   42.637254] [WMT-CORE][E]wmt_core_stp_init(648):WMT-CORE: no hif info!
[   42.643739] drivers/misc/mediatek/connectivity/common/conn_soc/core/wmt_core.c, 649, (0)
[   42.651807] [WMT-CORE][E]opfunc_pwr_on(918):WMT-CORE: wmt_core_stp_init fail (-1)
[   42.659267] drivers/misc/mediatek/connectivity/common/conn_soc/core/wmt_core.c, 919, (0)
[   42.667330] [WMT-CORE][E]wmt_core_ctrl(500):WMT-CORE: wmt_core_ctrl failed: id(5), type(1), value(0) iRet:(-1)
[   42.677280] drivers/misc/mediatek/connectivity/common/conn_soc/core/wmt_core.c, 501, (0)
[   42.685313] [WMT-CORE][E]wmt_core_ctrl(500):WMT-CORE: wmt_core_ctrl failed: id(5), type(0), value(0) iRet:(-1)
[   42.695255] drivers/misc/mediatek/connectivity/common/conn_soc/core/wmt_core.c, 501, (0)
[   42.703316] [WMT-CORE][E]wmt_core_ctrl(500):WMT-CORE: wmt_core_ctrl failed: id(3), type(0), value(0) iRet:(-2)
[   42.713261] drivers/misc/mediatek/connectivity/common/conn_soc/core/wmt_core.c, 501, (0)

first lead to

github.com

frank-w/BPI-R2-4.14/blob/8529be1dab2e25de8bc8102f93d938160993db53/drivers/misc/mediatek/connectivity/common/conn_soc/core/wmt_core.c#L648


	INT32 iRet = -1;
	unsigned long ctrlPa1;
	unsigned long ctrlPa2;
	UINT8 co_clock_type;
	P_WMT_CTX pctx = &gMtkWmtCtx;
	P_WMT_GEN_CONF pWmtGenConf = NULL;


	wmt_conf_read_file();
	pWmtGenConf = wmt_conf_get_cfg();
	if (!(pctx->wmtInfoBit & WMT_OP_HIF_BIT)) {
		WMT_ERR_FUNC("WMT-CORE: no hif info!\n");
		osal_assert(0);
		return -1;
	}
	/* 4 <1> open stp */
	ctrlPa1 = 0;
	ctrlPa2 = 0;
	iRet = wmt_core_ctrl(WMT_CTRL_STP_OPEN, &ctrlPa1, &ctrlPa2);
	if (iRet) {
		WMT_ERR_FUNC("WMT-CORE: wmt open stp\n");
		return -2;

did not know how critical it is…before it tries to read from config-file, but this seems to broke init-procedure

first three lines are from the /system/etc/firmware/WMT_SOC.cfg

mt6620.defAnt=mt6620_ant_m3.cfg
mt6628.defAnt=mt6628_ant_m1.cfg
mt6630.defAnt=mt6630_ant_m1.cfg

but the check itself seems not to check the config

it relies on a struct where the bit is set in

github.com

frank-w/BPI-R2-4.14/blob/8529be1dab2e25de8bc8102f93d938160993db53/drivers/misc/mediatek/connectivity/common/conn_soc/core/wmt_core.c#L862


	if (iret) {
		WMT_ERR_FUNC("chip id(0x%x) ver_check error:%d\n", chipid, iret);
		return -5;
	}


	WMT_DBG_FUNC("chip id(0x%x) ver_check ok\n", chipid);
	gMtkWmtCtx.p_ic_ops = p_ops;
	return 0;
}


static INT32 opfunc_hif_conf(P_WMT_OP pWmtOp)
{
	if (!(pWmtOp->u4InfoBit & WMT_OP_HIF_BIT)) {
		WMT_ERR_FUNC("WMT-CORE: no HIF_BIT in WMT_OP!\n");
		return -1;
	}


	if (gMtkWmtCtx.wmtInfoBit & WMT_OP_HIF_BIT) {
		WMT_ERR_FUNC("WMT-CORE: WMT HIF already exist. overwrite! old (%d), new(%d))\n",
			     gMtkWmtCtx.wmtHifConf.hifType, pWmtOp->au4OpData[0]);
	} else {

if not breaked before (but i don’t see that message in my log…maybe i need define-debug here too)

this looks also interesting:

github.com

frank-w/BPI-R2-4.14/blob/8529be1dab2e25de8bc8102f93d938160993db53/drivers/misc/mediatek/connectivity/common/conn_soc/linux/pri/wmt_dev.c#L1914


				iRet = -EFAULT;
				kfree(pBuffer);
				break;
			}
			pBuffer[NAME_MAX] = '\0';
			wmt_lib_set_patch_name(pBuffer);
			kfree(pBuffer);
		}
		break;


	case WMT_IOCTL_SET_STP_MODE:	/* stp/hif/fm mode */


		/* set hif conf */
		do {
			P_OSAL_OP pOp;
			MTK_WCN_BOOL bRet;
			P_OSAL_SIGNAL pSignal = NULL;
			P_WMT_HIF_CONF pHif = NULL;


			iRet = wmt_lib_set_hif(arg);
			if (0 != iRet) {

maybe we have to call stp_uart_launcher with different parameters to activate hif…a bit later i see this:

[   43.439140] DEBUG: Passed mtk_bt_hci_open 314 
[   43.443555] mtk_bt_hci_open func on failed with 0
[   44.848357] [WMT-CORE][E]opfunc_hif_conf(874):WMT-CORE: WMT HIF info added
[   44.856498] DEBUG: Passed _mtk_wcn_wmt_func_on 160

frank-w · December 31, 2018, 5:06am

anyone who can help here?

for me it looks like opfunc_hif_conf (function where WMT_OP_HIF_BIT is set) is called to late (after wmt_core_stp_init where it is needed) but i dont understand the flow…

opfunc_hif_conf is called over a struct wmt_core_opfunc

github.com

frank-w/BPI-R2-4.14/blob/8529be1dab2e25de8bc8102f93d938160993db53/drivers/misc/mediatek/connectivity/common/conn_soc/core/wmt_core.c#L234


static UINT8 WMT_ANT_RAM_DWN_EVT[] = { 0x02, 0x15, 0x01, 0x00	/*length */
	, 0x00
};
#endif


/* GeorgeKuo: Use designated initializers described in
* http://gcc.gnu.org/onlinedocs/gcc-4.0.4/gcc/Designated-Inits.html
*/


static const WMT_OPID_FUNC wmt_core_opfunc[] = {
	[WMT_OPID_HIF_CONF] = opfunc_hif_conf,
	[WMT_OPID_PWR_ON] = opfunc_pwr_on,
	[WMT_OPID_PWR_OFF] = opfunc_pwr_off,
	[WMT_OPID_FUNC_ON] = opfunc_func_on,
	[WMT_OPID_FUNC_OFF] = opfunc_func_off,
	[WMT_OPID_REG_RW] = opfunc_reg_rw,	/* TODO:[ChangeFeature][George] is this OP obsoleted? */
	[WMT_OPID_EXIT] = opfunc_exit,
	[WMT_OPID_PWR_SV] = opfunc_pwr_sv,
	[WMT_OPID_DSNS] = opfunc_dsns,
	[WMT_OPID_LPBK] = opfunc_lpbk,
	[WMT_OPID_CMD_TEST] = opfunc_cmd_test,

and used here

github.com

frank-w/BPI-R2-4.14/blob/4da7b54690665353f469799733d689e23d5b7407/drivers/misc/mediatek/connectivity/common/conn_soc/linux/pri/wmt_dev.c#L1935


				WMT_INFO_FUNC("wmt_lib_set_hif fail\n");
				break;
			}


			pOp = wmt_lib_get_free_op();
			if (!pOp) {
				WMT_INFO_FUNC("get_free_lxop fail\n");
				break;
			}
			pSignal = &pOp->signal;
			pOp->op.opId = WMT_OPID_HIF_CONF;


			pHif = wmt_lib_get_hif();


			osal_memcpy(&pOp->op.au4OpData[0], pHif, sizeof(WMT_HIF_CONF));
			pOp->op.u4InfoBit = WMT_OP_HIF_BIT;
			pSignal->timeoutValue = 0;


			bRet = wmt_lib_put_act_op(pOp);
			WMT_DBG_FUNC("WMT_OPID_HIF_CONF result(%d)\n", bRet);
			iRet = (MTK_WCN_BOOL_FALSE == bRet) ? -EFAULT : 0;

but where is it called?? @Jackzeng any idea?

bluetooth should mainly be handled in drivers/misc/mediatek/connectivity/common/conn_soc/linux/pub/stp_chrdev_bt.c but i don’t know the complete flow yet

frank-w · January 11, 2019, 7:05pm

with a patch from alxchk i got BT working (successful pairing)

first call wifi.sh (or at least call wmt_loader + stp-uart-launcher)
then load BT-module “modprobe stp_chrdev_bt”

here my log:

root@bpi-r2:~# bluetoothctl                                                     
[NEW] Controller 00:00:46:85:90:01 bpi-r2 [default]                             
[bluetooth]# power on                                                           
Changing power on succeeded                                                     
[bluetooth]# scan on                                                            
Discovery started                                                               
[CHG] Controller 00:00:46:85:90:01 Discovering: yes                             
[NEW] Device 00:14:03:05:08:AE 00-14-03-05-08-AE                                
[NEW] Device 9C:8C:6E:4A:F8:23 [TV] Samsung 8 Series (65)                       
[NEW] Device B4:9D:0B:5F:2A:B9 Aquaris X5 Plus                                  
[CHG] Device 00:14:03:05:08:AE LegacyPairing: no                                
[CHG] Device 00:14:03:05:08:AE Name: BPI-R2                                     
[CHG] Device 00:14:03:05:08:AE Alias: BPI-R2                                    
[bluetooth]# scan off                                                           
[CHG] Device B4:9D:0B:5F:2A:B9 RSSI is nil                                      
[CHG] Device 9C:8C:6E:4A:F8:23 RSSI is nil                                      
[CHG] Device 00:14:03:05:08:AE RSSI is nil                                      
[CHG] Controller 00:00:46:85:90:01 Discovering: no                              
Discovery stopped                                                               
[bluetooth]# pair B4:9D:0B:5F:2A:B9                                             
Attempting to pair with B4:9D:0B:5F:2A:B9                                       
[CHG] Device B4:9D:0B:5F:2A:B9 Connected: yes                                   
[CHG] Device B4:9D:0B:5F:2A:B9 Modalias: bluetooth:v000Fp1200d1436              
...
[CHG] Device B4:9D:0B:5F:2A:B9 ServicesResolved: yes                            
[CHG] Device B4:9D:0B:5F:2A:B9 Paired: yes                                      
Pairing successful                                                              
...                             
[bluetooth]# exit                                                               
[DEL] Controller 00:00:46:85:90:01 bpi-r2 [default]

yoursunny · February 21, 2019, 7:33pm

I tested this kernel revision. Pairing with cellphone works, but Bluetooth Low Energy does not work.

Client program running on BPI-R2:

from bluepy.btle import Scanner

scanner = Scanner()
devices = scanner.scan(10.0)

for dev in devices:
    print("Device %s (%s), RSSI=%d dB" % (dev.addr, dev.addrType, dev.rssi))
    for (adtype, desc, value) in dev.getScanData():
        print("  %s = %s" % (desc, value))

When I run this program on RPI-3B, several BLE devices are discovered. BPI-R2 cannot discover any device. bluepy library internally uses bluez-5.47, so bluez version isn’t a problem.

BPI-R2 kernel log: https://pastebin.com/uBwrjucW

frank-w · February 25, 2019, 4:42pm

@jackzeng do you have an idea where the problem can be? in the BT driver i’ve found no reference to BTLE

github.com

frank-w/BPI-R2-4.14/blob/4.19-bt/drivers/misc/mediatek/connectivity/common/conn_soc/linux/pub/stp_chrdev_bt.c

/*
* Copyright (C) 2011-2014 MediaTek Inc.
*
* This program is free software: you can redistribute it and/or modify it under the terms of the
* GNU General Public License version 2 as published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY;
* without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
* See the GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License along with this program.
* If not, see <http://www.gnu.org/licenses/>.
*/

#include <linux/init.h>
#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/fs.h>
#include <linux/cdev.h>

This file has been truncated. show original

frank-w · June 15, 2019, 8:56am

@Jackzeng @sinovoip @garywang

Datasheet posted here lists btle-support of mt6625l. Can you help me to implement it?

frank-w · June 30, 2019, 11:35am

have anyone a technical document of mt6625l with register-value-infos for btle-support?