Hi,
The commit header needs a bit of tweaking. It should be prefixed by the
driver name or core atom name. E.g.
atmodem: Handle broken +CMT notifications of some cinterion modems
On 08/28/2018 08:41 AM, [email protected] wrote:
Hi Denis,
I've recently found a bug in the +CMT notification parsing on some cinterion
modems. Some of them (ALS3, PLS8, maybe others...) have a leading coma right
after "+CTM : ", before the PDU size, which makes the parser rejecting the
message since the PDU size is not found.
Please find attached my proposed patch.
Could you fix the patch Author information to be something that I could
I apply? E.g. in a format like the AUTHORS file in ofono.git.
---
drivers/atmodem/sms.c | 17 +++++++++++++++--
1 file changed, 15 insertions(+), 2 deletions(-)
diff --git a/drivers/atmodem/sms.c b/drivers/atmodem/sms.c
index 68b89862..d650c0f6 100644
--- a/drivers/atmodem/sms.c
+++ b/drivers/atmodem/sms.c
@@ -412,8 +412,21 @@ static void at_cmt_notify(GAtResult *result, gpointer
user_data)
switch (data->vendor) {
case OFONO_VENDOR_CINTERION:
- if (!g_at_result_iter_next_number(&iter, &tpdu_len))
- goto err;
+ if (!g_at_result_iter_next_number(&iter, &tpdu_len)) {
+ /*
+ * Can't get the PDU length, we need to skip
+ * the leading coma for some modems (ALS3,PLS8...).
+ */
You might want to put an illustrative example of a broken +CMT
notification above. Doesn't have to be an actual valid message, just
something to showcase the broken syntax.
+ DBG("Retrying to find the PDU length");
+
+ if (!g_at_result_iter_skip_next(&iter))
+ goto err;
+
+ /* Next attempt at finding the PDU length. */
+ if (!g_at_result_iter_next_number(&iter, &tpdu_len))
+ goto err;
+ }
+
break;
default:
if (!g_at_result_iter_skip_next(&iter))
Otherwise this looks good to me.
Regards,
-Denis
_______________________________________________
ofono mailing list
[email protected]
https://lists.ofono.org/mailman/listinfo/ofono
