Re: [PATCH 3/5] ubloxmodem: add the netmon driver

[Denis Kenzior <denkenz@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 11424 bytes --]

Hi Djalal,

On 11/25/2016 08:00 AM, djalal(a)endocode.com wrote:
> From: Djalal Harouni <djalal@endocode.com>
>
> This adds a netmon driver for ublox. The driver support both +COPS and
> +CESQ commands to return the previously added ofono netmon types:
>
> RSCP: Received Signal Code Power
> ECN0: Received Energy Ratio
> RSRQ: Reference Signal Received Quality
> RSRP: Reference Signal Received Power
> OPERATOR: the operator
> ---
>   Makefile.am                 |   1 +
>   drivers/ubloxmodem/netmon.c | 339 ++++++++++++++++++++++++++++++++++++++++++++
>   2 files changed, 340 insertions(+)
>   create mode 100644 drivers/ubloxmodem/netmon.c
>
> diff --git a/Makefile.am b/Makefile.am
> index 3d7774b..07adeab 100644
> --- a/Makefile.am
> +++ b/Makefile.am
> @@ -383,6 +383,7 @@ builtin_sources += drivers/atmodem/atutil.h \
>   			drivers/ubloxmodem/ubloxmodem.h \
>   			drivers/ubloxmodem/ubloxmodem.c \
>   			drivers/ubloxmodem/gprs-context.c \
> +			drivers/ubloxmodem/netmon.c \
>   			drivers/ubloxmodem/lte.c
>
>
> diff --git a/drivers/ubloxmodem/netmon.c b/drivers/ubloxmodem/netmon.c
> new file mode 100644
> index 0000000..6b9d74f
> --- /dev/null
> +++ b/drivers/ubloxmodem/netmon.c
> @@ -0,0 +1,339 @@
> +/*
> + *
> + *  oFono - Open Source Telephony
> + *
> + *  Copyright (C) 2016  EndoCode AG. All rights reserved.
> + *
> + *  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, write to the Free Software
> + *  Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
> + *
> + */
> +
> +#ifdef HAVE_CONFIG_H
> +#include <config.h>
> +#endif
> +
> +#define _GNU_SOURCE
> +#include <stdio.h>
> +#include <stdlib.h>
> +#include <stdint.h>
> +#include <string.h>
> +#include <errno.h>
> +
> +#include <glib.h>
> +
> +#include <ofono/log.h>
> +#include <ofono/modem.h>
> +#include <ofono/netmon.h>
> +
> +#include "gatchat.h"
> +#include "gatresult.h"
> +
> +#include "common.h"
> +#include "netreg.h"
> +#include "ubloxmodem.h"
> +#include "drivers/atmodem/vendor.h"
> +
> +static const char *cops_prefix[] = { "+COPS:", NULL };
> +static const char *cesq_prefix[] = { "+CESQ:", NULL };
> +
> +struct netmon_driver_data {
> +	GAtChat *modem;
> +	GAtChat *aux;
> +};

Why do you need both the modem and the aux port?  I only see you using 
the aux port.

> +
> +struct req_cb_data {
> +	struct ofono_netmon *netmon;
> +
> +	ofono_netmon_cb_t cb;
> +	void *data;
> +
> +	struct ofono_network_operator op;
> +
> +	int rxlev;	/* CESQ: Received Signal Strength Indication */
> +	int ber;	/* CESQ: Bit Error Rate */
> +	int rscp;	/* CESQ: Received Signal Code Powe */
> +	int rsrp;	/* CESQ: Reference Signal Received Power */
> +	double ecn0;	/* CESQ: Received Energy Ratio */
> +	double rsrq;	/* CESQ: Reference Signal Received Quality */

Why are these doubles?

> +};
> +
> +/*
> + * Returns the appropriate radio access technology.
> + *
> + * If we can not resolve to a specific radio access technolgy
> + * we return OFONO_NETMON_CELL_TYPE_GSM by default.
> + */
> +static int ublox_map_radio_access_technology(int tech)
> +{
> +	switch (tech) {
> +	case ACCESS_TECHNOLOGY_GSM:
> +	case ACCESS_TECHNOLOGY_GSM_COMPACT:
> +		return OFONO_NETMON_CELL_TYPE_GSM;
> +	case ACCESS_TECHNOLOGY_UTRAN:
> +	case ACCESS_TECHNOLOGY_UTRAN_HSDPA:
> +	case ACCESS_TECHNOLOGY_UTRAN_HSUPA:
> +	case ACCESS_TECHNOLOGY_UTRAN_HSDPA_HSUPA:
> +		return OFONO_NETMON_CELL_TYPE_UMTS;
> +	case ACCESS_TECHNOLOGY_EUTRAN:
> +		return OFONO_NETMON_CELL_TYPE_LTE;
> +	}
> +
> +	return OFONO_NETMON_CELL_TYPE_GSM;
> +}
> +
> +static inline struct req_cb_data *req_cb_data_new0(void *cb, void *data, void *user)
> +{
> +	struct req_cb_data *ret = g_new0(struct req_cb_data, 1);
> +	if (ret == NULL)
> +		return NULL;
> +
> +
> +	ret->cb = cb;
> +	ret->data = data;
> +	ret->netmon = user;
> +	ret->rxlev = -1;
> +	ret->ber = -1;
> +	ret->rscp = -1;
> +	ret->rsrp = -1;
> +	ret->ecn0 = -1;
> +	ret->rsrq = -1;
> +
> +	return ret;
> +}
> +
> +static gboolean ublox_delayed_register(gpointer user_data)
> +{
> +	struct ofono_netmon *netmon = user_data;
> +
> +	ofono_netmon_register(netmon);
> +
> +	return FALSE;
> +}
> +
> +static void cesq_cb(gboolean ok, GAtResult *result, gpointer user_data)
> +{
> +	enum cesq_ofono_netmon_info {
> +		CESQ_RXLEV,
> +		CESQ_BER,
> +		CESQ_RSCP,
> +		CESQ_ECN0,
> +		CESQ_RSRQ,
> +		CESQ_RSRP,
> +		_MAX,
> +	};
> +
> +	struct req_cb_data *cbd = user_data;
> +	struct ofono_netmon *nm = cbd->netmon;
> +	struct ofono_error error;
> +	GAtResultIter iter;
> +	int idx, number;
> +
> +	DBG("ok %d", ok);
> +
> +	decode_at_error(&error, g_at_result_final_response(result));
> +
> +	if (!ok) {
> +		goto error;
> +	}

No need for parentheses

> +
> +	g_at_result_iter_init(&iter, result);
> +
> +	if (!g_at_result_iter_next(&iter, "+CESQ:"))
> +		return;

What happens to memory used by cbd? :)

> +
> +	for (idx = 0; idx < _MAX; idx++) {
> +
> +		ok = g_at_result_iter_next_number(&iter, &number);
> +		if (!ok) {
> +			DBG(" error at idx: %d", idx);
> +			goto error;
> +		}
> +
> +		switch (idx) {
> +		case CESQ_RXLEV:
> +			cbd->rxlev = number != 99 ? number:cbd->rxlev;
> +			break;
> +		case CESQ_BER:
> +			cbd->ber = number != 99 ? number:cbd->ber;
> +			break;
> +		case CESQ_RSCP:
> +			cbd->rscp= number != 255 ? number:cbd->rscp;
> +			break;
> +		case CESQ_ECN0:
> +			cbd->ecn0= number != 255 ? number:cbd->ecn0;
> +			break;
> +		case CESQ_RSRQ:
> +			cbd->rsrq= number != 255 ? number:cbd->rsrq;
> +			break;
> +		case CESQ_RSRP:
> +			cbd->rsrp= number != 255 ? number:cbd->rsrp;
> +			break;

space before / after equal sign.  See doc/coding-style.txt

> +		}
> +	}
> +
> +	DBG(" RXLEV	%d ", cbd->rxlev);
> +	DBG(" BER	%d ", cbd->ber);
> +	DBG(" RSCP	%d ", cbd->rscp);
> +	DBG(" ECN0	%f ", cbd->ecn0);
> +	DBG(" RSRQ	%f ", cbd->rsrq);
> +	DBG(" RSRP	%d ", cbd->rsrp);
> +
> +	ofono_netmon_serving_cell_notify(nm,
> +					 cbd->op.tech,
> +					 OFONO_NETMON_INFO_OPERATOR, cbd->op.name,
> +					 OFONO_NETMON_INFO_RXLEV, cbd->rxlev,
> +					 OFONO_NETMON_INFO_BER, cbd->ber,
> +					 OFONO_NETMON_INFO_RSCP, cbd->rscp,
> +					 OFONO_NETMON_INFO_ECN0, cbd->ecn0,
> +					 OFONO_NETMON_INFO_RSRQ, cbd->rsrq,
> +					 OFONO_NETMON_INFO_RSRP, cbd->rsrp,
> +					 OFONO_NETMON_INFO_INVALID);
> +
> +	CALLBACK_WITH_SUCCESS(cbd->cb, cbd->data);
> +	g_free(cbd);
> +	return;
> +
> +error:
> +	CALLBACK_WITH_FAILURE(cbd->cb, cbd->data);
> +	g_free(cbd);
> +}
> +
> +static void cops_cb(gboolean ok, GAtResult *result, gpointer user_data)
> +{
> +	struct req_cb_data *cbd = user_data;
> +	struct ofono_netmon *nm = cbd->netmon;
> +	struct netmon_driver_data *nmd = ofono_netmon_get_data(nm);
> +	struct ofono_error error;
> +	GAtResultIter iter;
> +	int tech;
> +	const char *name;
> +
> +	DBG("ok %d", ok);
> +
> +	decode_at_error(&error, g_at_result_final_response(result));
> +
> +	if (!ok) {
> +		goto error;
> +	}

No need for parentheses

> +
> +	g_at_result_iter_init(&iter, result);
> +
> +	if (!g_at_result_iter_next(&iter, "+COPS:"))
> +		return;

What happens to memory used by cbd?

> +
> +	g_at_result_iter_skip_next(&iter);
> +	g_at_result_iter_skip_next(&iter);
> +
> +	if (g_at_result_iter_next_string(&iter, &name) == FALSE)
> +		goto error;
> +
> +	strncpy(cbd->op.name, name, OFONO_MAX_OPERATOR_NAME_LENGTH);
> +	cbd->op.name[OFONO_MAX_OPERATOR_NAME_LENGTH] = '\0';
> +

As mentioned before, not sure string representation of the operator is 
useful...

> +	/* Ignored for now. TODO: maybe read format but value
> +	 * wouldn't be forwarder anywhere
> +	 */
> +	cbd->op.mcc[0] = '\0';
> +	cbd->op.mnc[0] = '\0';
> +
> +	/* Default to GSM */
> +	if (g_at_result_iter_next_number(&iter, &tech) == FALSE)
> +		cbd->op.tech = ublox_map_radio_access_technology(ACCESS_TECHNOLOGY_GSM);
> +	else
> +		cbd->op.tech = ublox_map_radio_access_technology(tech);
> +
> +	if (g_at_chat_send(nmd->aux, "AT+CESQ", cesq_prefix,
> +			   cesq_cb, cbd, NULL)) {
> +		return;
> +	}
> +
> +error:
> +	CALLBACK_WITH_FAILURE(cbd->cb, cbd->data);
> +	g_free(cbd);
> +}
> +
> +

double empty line

> +static void ublox_netmon_request_update(struct ofono_netmon *netmon,
> +					ofono_netmon_cb_t cb, void *data)
> +{
> +	struct netmon_driver_data *nmd = ofono_netmon_get_data(netmon);
> +	struct req_cb_data *cbd;
> +
> +	DBG("ublox netmon request update");
> +
> +	cbd = req_cb_data_new0(cb, data, netmon);
> +	if (!cbd) {
> +		CALLBACK_WITH_FAILURE(cb, data);
> +		return;
> +	}
> +
> +	if (g_at_chat_send(nmd->aux, "AT+COPS?", cops_prefix,
> +			   cops_cb, cbd, NULL) == 0) {

You should be providing a destructor for cbd.  E.g. last argument to 
g_at_chat_send.  Avoids memory leaks if the modem gets hot-unplugged for 
example.  I think some of the drivers still get this wrong when 
implementing multiple-command transactions.

Might be safer to simply issue AT+COPS;+CESQ and write a single callback 
that parses both responses.

> +		CALLBACK_WITH_FAILURE(cb, data);
> +		g_free(cbd);
> +	}
> +}
> +
> +static int ublox_netmon_probe(struct ofono_netmon *netmon,
> +			      unsigned int vendor, void *user)
> +{
> +	struct netmon_driver_data *n = user;

This is evil and is going to bite you at some point.  In 
ofono_netmon_create inside ublox.c you pass in ublox_data.  So if 
someone changes this structure or ublox_data, things will explode.

> +	struct netmon_driver_data *nmd;
> +
> +	DBG("ublox netmon probe");
> +
> +	nmd = g_try_new0(struct netmon_driver_data, 1);
> +	if (nmd == NULL)
> +		return -ENOMEM;
> +
> +	nmd->modem = g_at_chat_clone(n->modem);
> +	nmd->aux = g_at_chat_clone(n->aux);

If you truly need both ports, use the g_at_chat_set_slave functionality.

> +
> +	ofono_netmon_set_data(netmon, nmd);
> +
> +	g_idle_add(ublox_delayed_register, netmon);
> +
> +	return 0;
> +}
> +
> +static void ublox_netmon_remove(struct ofono_netmon *netmon)
> +{
> +	struct netmon_driver_data *nmd = ofono_netmon_get_data(netmon);
> +
> +	DBG("ublox netmon remove");
> +
> +	ofono_netmon_set_data(netmon, NULL);
> +
> +	g_at_chat_unref(nmd->modem);
> +	g_at_chat_unref(nmd->aux);
> +
> +	g_free(nmd);
> +}
> +
> +static struct ofono_netmon_driver driver = {
> +	.name			= UBLOXMODEM,
> +	.probe			= ublox_netmon_probe,
> +	.remove			= ublox_netmon_remove,
> +	.request_update		= ublox_netmon_request_update,
> +};
> +
> +void ublox_netmon_init(void)
> +{
> +	ofono_netmon_driver_register(&driver);
> +}
> +
> +void ublox_netmon_exit(void)
> +{
> +	ofono_netmon_driver_unregister(&driver);
> +}
>

Regards,
-Denis