osmo-bts-trx: "CHAN RQD: no resources for SDCCH" on master

Submitted by Tom Tsou on June 15, 2016, 7:38 p.m.

Details

Message ID CAATyM+adxM9-db+tZ+s2kvn7bdpY9CKJ-JoGOv2TMiTWk5kObQ@mail.gmail.com
State New
Series "osmo-bts-trx: "CHAN RQD: no resources for SDCCH" on master"
Headers show

Commit Message

Tom Tsou June 15, 2016, 7:38 p.m.
On Wed, Jun 15, 2016 at 10:08 AM, Max <msuraev@sysmocom.de> wrote:
> I've tried it with manually sending POWERON - the issue with CHAN RQD is
> gone but the phone still does not work. I'll look into OML/RSL/whatever
> is preventing it from working.

Make sure that the static variable 'settsc_enabled' is set.

Patch hide | download patch | download mbox

diff --git a/src/osmo-bts-trx/trx_if.c b/src/osmo-bts-trx/trx_if.c
index a4c16dc..8c9a839 100644
--- a/src/osmo-bts-trx/trx_if.c
+++ b/src/osmo-bts-trx/trx_if.c
@@ -47,7 +47,7 @@ 
 //#define TOA_RSSI_DEBUG

 int transceiver_available = 0;
-int settsc_enabled = 0;
+int settsc_enabled = 1;
 int setbsic_enabled = 0;

  -TT

Comments

Alexander Chemeris June 15, 2016, 7:50 p.m.
On Jun 15, 2016 10:38 PM, "Tom Tsou" <tom@tsou.cc> wrote:
>
> On Wed, Jun 15, 2016 at 10:08 AM, Max <msuraev@sysmocom.de> wrote:
> > I've tried it with manually sending POWERON - the issue with CHAN RQD is
> > gone but the phone still does not work. I'll look into OML/RSL/whatever
> > is preventing it from working.
>
> Make sure that the static variable 'settsc_enabled' is set.
>
> diff --git a/src/osmo-bts-trx/trx_if.c b/src/osmo-bts-trx/trx_if.c
> index a4c16dc..8c9a839 100644
> --- a/src/osmo-bts-trx/trx_if.c
> +++ b/src/osmo-bts-trx/trx_if.c
> @@ -47,7 +47,7 @@
>  //#define TOA_RSSI_DEBUG
>
>  int transceiver_available = 0;
> -int settsc_enabled = 0;
> +int settsc_enabled = 1;
>  int setbsic_enabled = 0;

Could you explain why is that required?

I don't remember the code exactly, but IIRC this is equivalent to setting
or removing 'tsc' line in the config file.

Please excuse typos. Written with a touchscreen keyboard.

--
Regards,
Alexander Chemeris
CEO Fairwaves, Inc.
https://fairwaves.co
Tom Tsou June 15, 2016, 8:19 p.m.
On Wed, Jun 15, 2016 at 12:50 PM, Alexander Chemeris
<alexander.chemeris@gmail.com> wrote:
>> diff --git a/src/osmo-bts-trx/trx_if.c b/src/osmo-bts-trx/trx_if.c
>> index a4c16dc..8c9a839 100644
>> --- a/src/osmo-bts-trx/trx_if.c
>> +++ b/src/osmo-bts-trx/trx_if.c
>> @@ -47,7 +47,7 @@
>>  //#define TOA_RSSI_DEBUG
>>
>>  int transceiver_available = 0;
>> -int settsc_enabled = 0;
>> +int settsc_enabled = 1;
>>  int setbsic_enabled = 0;
>
> Could you explain why is that required?

Because osmo-trx only supports setting the training sequence through
the SETTSC command. There is some history of the code that I am not
familiar with. There is a SETBSIC command on the osmo-bts side of the
interface, but that command does not exist in osmo-trx side. Perhaps
SETBSIC previously existed in OpenBTS at one point in time?

> I don't remember the code exactly, but IIRC this is equivalent to setting or
> removing 'tsc' line in the config file.

I'm not aware SETTSC vs SETBSIC configuration outside of those static variables.

  -TT
msuraev@sysmocom.de June 16, 2016, 8:45 a.m.
Neat, the test call passed through. Now we got to find a place to send
poweron automatically at the right moment instead of manual activation.

On 06/15/2016 09:38 PM, Tom Tsou wrote:
> On Wed, Jun 15, 2016 at 10:08 AM, Max <msuraev@sysmocom.de> wrote:
>> I've tried it with manually sending POWERON - the issue with CHAN RQD is
>> gone but the phone still does not work. I'll look into OML/RSL/whatever
>> is preventing it from working.
> Make sure that the static variable 'settsc_enabled' is set.
>
> diff --git a/src/osmo-bts-trx/trx_if.c b/src/osmo-bts-trx/trx_if.c
> index a4c16dc..8c9a839 100644
> --- a/src/osmo-bts-trx/trx_if.c
> +++ b/src/osmo-bts-trx/trx_if.c
> @@ -47,7 +47,7 @@
>  //#define TOA_RSSI_DEBUG
>
>  int transceiver_available = 0;
> -int settsc_enabled = 0;
> +int settsc_enabled = 1;
>  int setbsic_enabled = 0;
>
>   -TT
msuraev@sysmocom.de June 16, 2016, 8:52 a.m.
Curiously I've noticed following error messages while testing osmo-bts-trx:

<0010> lapd_core.c:1508 N(S) sequence error: N(S)=0, V(R)=1
<0010> lapd_core.c:1508 N(S) sequence error: N(S)=1, V(R)=2
<0010> lapd_core.c:1508 N(S) sequence error: N(S)=2, V(R)=3
<0010> lapd_core.c:1508 N(S) sequence error: N(S)=3, V(R)=4
<0010> lapd_core.c:1508 N(S) sequence error: N(S)=4, V(R)=5
<0010> lapd_core.c:1508 N(S) sequence error: N(S)=5, V(R)=6

Not sure if it's related.

On 06/15/2016 09:38 PM, Tom Tsou wrote:
> On Wed, Jun 15, 2016 at 10:08 AM, Max <msuraev@sysmocom.de> wrote:
>> I've tried it with manually sending POWERON - the issue with CHAN RQD is
>> gone but the phone still does not work. I'll look into OML/RSL/whatever
>> is preventing it from working.
> Make sure that the static variable 'settsc_enabled' is set.
>
> diff --git a/src/osmo-bts-trx/trx_if.c b/src/osmo-bts-trx/trx_if.c
> index a4c16dc..8c9a839 100644
> --- a/src/osmo-bts-trx/trx_if.c
> +++ b/src/osmo-bts-trx/trx_if.c
> @@ -47,7 +47,7 @@
>  //#define TOA_RSSI_DEBUG
>
>  int transceiver_available = 0;
> -int settsc_enabled = 0;
> +int settsc_enabled = 1;
>  int setbsic_enabled = 0;
>
>   -TT
Alexander Chemeris June 21, 2016, 4:44 p.m.
On Wed, Jun 15, 2016 at 11:19 PM, Tom Tsou <tom@tsou.cc> wrote:
> On Wed, Jun 15, 2016 at 12:50 PM, Alexander Chemeris
> <alexander.chemeris@gmail.com> wrote:
>>> diff --git a/src/osmo-bts-trx/trx_if.c b/src/osmo-bts-trx/trx_if.c
>>> index a4c16dc..8c9a839 100644
>>> --- a/src/osmo-bts-trx/trx_if.c
>>> +++ b/src/osmo-bts-trx/trx_if.c
>>> @@ -47,7 +47,7 @@
>>>  //#define TOA_RSSI_DEBUG
>>>
>>>  int transceiver_available = 0;
>>> -int settsc_enabled = 0;
>>> +int settsc_enabled = 1;
>>>  int setbsic_enabled = 0;
>>
>> Could you explain why is that required?
>
> Because osmo-trx only supports setting the training sequence through
> the SETTSC command. There is some history of the code that I am not
> familiar with. There is a SETBSIC command on the osmo-bts side of the
> interface, but that command does not exist in osmo-trx side. Perhaps
> SETBSIC previously existed in OpenBTS at one point in time?

SETBSIC was introduced for osmocom-bb-trx (an OsmocomBB phone
operating as a BTS). These phones didn't have a function to set TSC,
but had a function to set BSIC, so Andreas added SETBSIC command to
work around this. OsmoBTS has sent both commands, because it doesn't
know whether it's connected to osmo-trx or to osmocom-bb-trx - an
unsupported command was historically just ignored. Has this behaviour
changed recently? I don't think there is any issues with sending
SETBSIC to osmo-trx except a warning in the log.

>> I don't remember the code exactly, but IIRC this is equivalent to setting or
>> removing 'tsc' line in the config file.
>
> I'm not aware SETTSC vs SETBSIC configuration outside of those static variables.

I was sure that I saw that removing "bsic" parameter stopped osmo-bts
from sending SETBSIC, but now I don't see that in the code. That
should be the best way to handle this - only send TSC of BSIC based on
configuration.