1:16 p.m.
Use AT+CHLD=1x to release a specific call if AG supports that.
---
drivers/hfpmodem/voicecall.c | 51 +++++++++++++++++++++++++++++++++++++++++-
1 files changed, 50 insertions(+), 1 deletions(-)
diff --git a/drivers/hfpmodem/voicecall.c b/drivers/hfpmodem/voicecall.c
index 17811d6..6d7645e 100644
--- a/drivers/hfpmodem/voicecall.c
+++ b/drivers/hfpmodem/voicecall.c
@@ -452,6 +452,55 @@ static void hfp_release_all_active(struct ofono_voicecall *vc,
CALLBACK_WITH_FAILURE(cb, data);
}
+static void release_id_cb(gboolean ok, GAtResult *result,
+ gpointer user_data)
+{
+ struct release_id_req *req = user_data;
+ struct voicecall_data *vd = ofono_voicecall_get_data(req->vc);
+ struct ofono_error error;
+
+ dump_response("release_id_cb", ok, result);
+ decode_at_error(&error, g_at_result_final_response(result));
+
+ if (ok)
+ vd->local_release |= (1 << req->id);
+
+ req->cb(&error, req->data);
+}
+
+static void hfp_release_specific(struct ofono_voicecall *vc, int id,
+ ofono_voicecall_cb_t cb, void *data)
+{
+ struct voicecall_data *vd = ofono_voicecall_get_data(vc);
+ struct release_id_req *req;
+ char buf[32];
+
+ if (!(vd->ag_mpty_features & AG_CHLD_1x))
+ goto error;
+
+ req = g_try_new0(struct release_id_req, 1);
+
+ if (!req)
+ goto error;
+
+ req->vc = vc;
+ req->cb = cb;
+ req->data = data;
+ req->id = id;
+
+ sprintf(buf, "AT+CHLD=1%d", id);
+
+ if (g_at_chat_send(vd->chat, buf, none_prefix,
+ release_id_cb, req, g_free) > 0)
+ return;
+
+error:
+ if (req)
+ g_free(req);
+
+ CALLBACK_WITH_FAILURE(cb, data);
+}
+
static void hfp_send_dtmf(struct ofono_voicecall *vc, const char *dtmf,
ofono_voicecall_cb_t cb, void *data)
{
@@ -1010,7 +1059,7 @@ static struct ofono_voicecall_driver driver = {
.release_all_held = hfp_release_all_held,
.set_udub = hfp_set_udub,
.release_all_active = hfp_release_all_active,
- .release_specific = NULL,
+ .release_specific = hfp_release_specific,
.private_chat = NULL,
.create_multiparty = NULL,
.transfer = NULL,
--
1.6.5.2
1:16 p.m.
New subject: [PATCH 2/4] Add multiparty call supports for hfp modem
Add CHLD=3, 2x and 4 to create multiparty call, private chat and
call transfer.
---
drivers/hfpmodem/voicecall.c | 56 +++++++++++++++++++++++++++++++++++++++--
1 files changed, 53 insertions(+), 3 deletions(-)
diff --git a/drivers/hfpmodem/voicecall.c b/drivers/hfpmodem/voicecall.c
index 6d7645e..1d2188a 100644
--- a/drivers/hfpmodem/voicecall.c
+++ b/drivers/hfpmodem/voicecall.c
@@ -501,6 +501,56 @@ error:
CALLBACK_WITH_FAILURE(cb, data);
}
+static void hfp_private_chat(struct ofono_voicecall *vc, int id,
+ ofono_voicecall_cb_t cb, void *data)
+{
+ struct voicecall_data *vd = ofono_voicecall_get_data(vc);
+ char buf[32];
+
+ if (vd->ag_mpty_features & AG_CHLD_2x) {
+ sprintf(buf, "AT+CHLD=2%d", id);
+
+ hfp_template(buf, vc, generic_cb, 0, cb, data);
+
+ return;
+ }
+
+ CALLBACK_WITH_FAILURE(cb, data);
+}
+
+static void hfp_create_multiparty(struct ofono_voicecall *vc,
+ ofono_voicecall_cb_t cb, void *data)
+{
+ struct voicecall_data *vd = ofono_voicecall_get_data(vc);
+
+ if (vd->ag_mpty_features & AG_CHLD_3) {
+ hfp_template("AT+CHLD=3", vc, generic_cb, 0, cb, data);
+
+ return;
+ }
+
+ CALLBACK_WITH_FAILURE(cb, data);
+}
+
+static void hfp_transfer(struct ofono_voicecall *vc,
+ ofono_voicecall_cb_t cb, void *data)
+{
+ struct voicecall_data *vd = ofono_voicecall_get_data(vc);
+ /* Transfer can puts held & active calls together and disconnects
+ * from both. However, some networks support transfering of
+ * dialing/ringing calls as well.
+ */
+ unsigned int transfer = 0x1 | 0x2 | 0x4 | 0x8;
+
+ if (vd->ag_mpty_features & AG_CHLD_4) {
+ hfp_template("AT+CHLD=4", vc, generic_cb, transfer, cb, data);
+
+ return;
+ }
+
+ CALLBACK_WITH_FAILURE(cb, data);
+}
+
static void hfp_send_dtmf(struct ofono_voicecall *vc, const char *dtmf,
ofono_voicecall_cb_t cb, void *data)
{
@@ -1060,9 +1110,9 @@ static struct ofono_voicecall_driver driver = {
.set_udub = hfp_set_udub,
.release_all_active = hfp_release_all_active,
.release_specific = hfp_release_specific,
- .private_chat = NULL,
- .create_multiparty = NULL,
- .transfer = NULL,
+ .private_chat = hfp_private_chat,
+ .create_multiparty = hfp_create_multiparty,
+ .transfer = hfp_transfer,
.deflect = NULL,
.swap_without_accept = NULL,
.send_tones = hfp_send_dtmf
--
1.6.5.2
1:16 p.m.
New subject: [PATCH 3/4] Add poll_clcc if we are in multiparty call
If we are in multiparty call, remote AG won't update indicator
if any call in multiparty call is released by CHLD=1x. So we
have to poll it periodically.
---
drivers/hfpmodem/voicecall.c | 37 +++++++++++++++++++++++++++++++++++++
1 files changed, 37 insertions(+), 0 deletions(-)
diff --git a/drivers/hfpmodem/voicecall.c b/drivers/hfpmodem/voicecall.c
index 1d2188a..6b73e7c 100644
--- a/drivers/hfpmodem/voicecall.c
+++ b/drivers/hfpmodem/voicecall.c
@@ -39,6 +39,8 @@
#include "hfpmodem.h"
+#define POLL_CLCC_INTERVAL 2000
+
static const char *none_prefix[] = { NULL };
static const char *clcc_prefix[] = { "+CLCC:", NULL };
@@ -50,6 +52,7 @@ struct voicecall_data {
unsigned char cind_pos[HFP_INDICATOR_LAST];
int cind_val[HFP_INDICATOR_LAST];
unsigned int local_release;
+ unsigned int clcc_source;
};
struct release_id_req {
@@ -66,6 +69,8 @@ struct change_state_req {
int affected_types;
};
+static gboolean poll_clcc(gpointer user_data);
+
static GSList *find_dialing(GSList *calls)
{
GSList *c;
@@ -197,6 +202,8 @@ static void clcc_poll_cb(gboolean ok, GAtResult *result, gpointer
user_data)
GSList *calls;
GSList *n, *o;
struct ofono_call *nc, *oc;
+ unsigned int num_active = 0;
+ unsigned int num_held = 0;
dump_response("clcc_poll_cb", ok, result);
@@ -212,6 +219,14 @@ static void clcc_poll_cb(gboolean ok, GAtResult *result, gpointer
user_data)
nc = n ? n->data : NULL;
oc = o ? o->data : NULL;
+ if (nc) {
+ if (nc->status == CALL_STATUS_ACTIVE)
+ num_active++;
+
+ if (nc->status == CALL_STATUS_HELD)
+ num_held++;
+ }
+
if (oc && (!nc || (nc->id > oc->id))) {
enum ofono_disconnect_reason reason;
@@ -255,6 +270,25 @@ static void clcc_poll_cb(gboolean ok, GAtResult *result, gpointer
user_data)
g_slist_free(vd->calls);
vd->calls = calls;
+
+ /* If either active/held call is more than 1, we are in mpty calls.
+ * we won't get indicator update if any of them is released by CHLD=1x.
+ * So we have to poll it.
+ */
+ if (num_active > 1 || num_held > 1)
+ vd->clcc_source = g_timeout_add(POLL_CLCC_INTERVAL, poll_clcc,
+ vc);
+}
+
+static gboolean poll_clcc(gpointer user_data)
+{
+ struct ofono_voicecall *vc = user_data;
+ struct voicecall_data *vd = ofono_voicecall_get_data(vc);
+
+ g_at_chat_send(vd->chat, "AT+CLCC", clcc_prefix,
+ clcc_poll_cb, vc, NULL);
+
+ return FALSE;
}
static void generic_cb(gboolean ok, GAtResult *result, gpointer user_data)
@@ -1090,6 +1124,9 @@ static void hfp_voicecall_remove(struct ofono_voicecall *vc)
{
struct voicecall_data *vd = ofono_voicecall_get_data(vc);
+ if (vd->clcc_source)
+ g_source_remove(vd->clcc_source);
+
g_slist_foreach(vd->calls, (GFunc) g_free, NULL);
g_slist_free(vd->calls);
--
1.6.5.2
1:16 p.m.
New subject: [PATCH 4/4] Fix schedule clcc poll in the ring notify
Similar to atmodem, We need schedule CLCC poll in the ring notify in
case AG doesn't send CLIP notification to us. If CLIP comes right
after RING, we cancel the CLCC poll.
---
drivers/hfpmodem/voicecall.c | 15 +++++++++++++++
1 files changed, 15 insertions(+), 0 deletions(-)
diff --git a/drivers/hfpmodem/voicecall.c b/drivers/hfpmodem/voicecall.c
index 6b73e7c..dc9359a 100644
--- a/drivers/hfpmodem/voicecall.c
+++ b/drivers/hfpmodem/voicecall.c
@@ -40,6 +40,7 @@
#include "hfpmodem.h"
#define POLL_CLCC_INTERVAL 2000
+#define CLIP_INTERVAL 200
static const char *none_prefix[] = { NULL };
static const char *clcc_prefix[] = { "+CLCC:", NULL };
@@ -717,6 +718,12 @@ static void ring_notify(GAtResult *result, gpointer user_data)
if (!call)
ofono_error("Couldn't create call, call management is fubar!");
+
+ /* We don't know the number must wait for CLIP to arrive before
+ * announcing the call. So we schedule the clcc call. If the CLIP
+ * arrives earlier, we annouce the call there
+ */
+ vd->clcc_source = g_timeout_add(CLIP_INTERVAL, poll_clcc, vc);
}
static void clip_notify(GAtResult *result, gpointer user_data)
@@ -773,6 +780,14 @@ static void clip_notify(GAtResult *result, gpointer user_data)
call->clip_validity = validity;
ofono_voicecall_notify(vc, call);
+
+ /* We started a CLCC, but the CLIP arrived and the call number
+ * is known so we cancel the GSource
+ */
+ if (vd->clcc_source) {
+ g_source_remove(vd->clcc_source);
+ vd->clcc_source = 0;
+ }
}
static void ciev_call_notify(struct ofono_voicecall *vc,
--
1.6.5.2
4:14 p.m.
New subject: [PATCH 4/4] Fix schedule clcc poll in the ring notify
Hi,
Zhang, Zhenhua wrote:
Similar to atmodem, We need schedule CLCC poll in the ring
notify in case AG doesn't send CLIP notification to us. If
CLIP comes right after RING, we cancel the CLCC poll. ---
drivers/hfpmodem/voicecall.c | 15 +++++++++++++++
1 files changed, 15 insertions(+), 0 deletions(-)
Changed to add clip_timeout to notify the call directly if CLIP
notification doesn't arrive. Please review it.
Regards,
Zhenhua
5 p.m.
New subject: [PATCH 4/4] Fix schedule clcc poll in the ring notify
Hi,
Zhang, Zhenhua wrote:
Hi,
Zhang, Zhenhua wrote:
> Similar to atmodem, We need schedule CLCC poll in the ring notify in
> case AG doesn't send CLIP notification to us. If CLIP comes right
> after RING, we cancel the CLCC poll. ---
> drivers/hfpmodem/voicecall.c | 15 +++++++++++++++
> 1 files changed, 15 insertions(+), 0 deletions(-)
>
Changed to add clip_timeout to notify the call directly if
CLIP notification doesn't arrive. Please review it.
Regards,
Zhenhua
Replace clcc_source to clip_source to make code be more clear.
Regards,
Zhenhua
7:48 p.m.
New subject: [PATCH 4/4] Fix schedule clcc poll in the ring notify
Hi Zhenhua,
All your patches have been applied.
Regards,
-Denis
3:46 p.m.
New subject: [PATCH 3/4] Add poll_clcc if we are in multiparty call
Hi,
Zhang, Zhenhua wrote:
If we are in multiparty call, remote AG won't update indicator
if any call in multiparty call is released by CHLD=1x. So we have to
poll it periodically. ---
drivers/hfpmodem/voicecall.c | 37
+++++++++++++++++++++++++++++++++++++
1 files changed, 37 insertions(+), 0 deletions(-)
I updated this patch to reset vd->clcc_source in poll_clcc. And fix a style
issue for num_active, num_held in clcc_poll_cb. Updated patch is attached.
Regards,
Zhenhua
4:26 p.m.
New subject: [PATCH 3/4] Add poll_clcc if we are in multiparty call
Hi,
Zhang, Zhenhua wrote:
Hi,
Zhang, Zhenhua wrote:
> If we are in multiparty call, remote AG won't update indicator if any
> call in multiparty call is released by CHLD=1x. So we have to poll
> it periodically. --- drivers/hfpmodem/voicecall.c | 37
> +++++++++++++++++++++++++++++++++++++
> 1 files changed, 37 insertions(+), 0 deletions(-)
>
I updated this patch to reset vd->clcc_source in poll_clcc.
And fix a style issue for num_active, num_held in
clcc_poll_cb. Updated patch is attached.
Regards,
Zhenhua
Update the patch title since it should be 3/4.
Regards,
Zhenhua
4:12 p.m.
Hi,
Zhang, Zhenhua wrote:
Use AT+CHLD=1x to release a specific call if AG supports that. ---
drivers/hfpmodem/voicecall.c | 51
+++++++++++++++++++++++++++++++++++++++++-
1 files changed, 50 insertions(+), 1 deletions(-)
Update the patch to initlize 'struct release_id_req *req' as NULL.
Regards,
Zhenhua
4666
days inactive
4666
days old
9 comments
3 participants
participants (3)
-
Denis Kenzior -
Zhang, Zhenhua -
Zhenhua Zhang