- OpenBSC - lists.osmocom.org

AW: LCR crashes. Please help me!
by Andreas.Eversberg 16 Jul '10

16 Jul '10

hi luca, it seems that there is a change in the API. therefore i prepared a "snapshot" at www.linux-call-router.de to be downloaded. this version of openbsc and lcr is older, but works together. as soon as the current openbsc has been tested more (channel ressource handling has changed for example), i will make it work again. regards, andreas > #0 0x0805b009 in gsm0408_rcvmsg (msg=0x864ffb8, link_id=0 '\0') at openbsc/src/bsc_api.c:111 > 111 rc = api->compl_l3(lchan->conn, msg, 0); > > The problem seems to be the variable api, which is NULL. > > (gdb) print api > $1 = (struct bsc_api *) 0x0

4 5

OpenBSC+LCR+Asterisk: I got it!
by Luca Bertoncello 14 Jul '10

14 Jul '10

Hi, List! I got it! Now can I call mobile phones with Asterisk. If someone is interested, the problem was in the Asterisk-configuration (from HowTo). To use in extensions.conf is: [btsctrl] exten => _02X.,1,GotoIf($[${CALLERID(name)} != ""]?4) exten => _02X.,2,Set(CALLIDORIG=${CALLERID(num)}) exten => _02X.,3,Set(CALLERID(num)=02${CALLIDORIG}) exten => _02X.,4,Dial(LCR/GSM/${EXTEN:2},120) So can I define a prefix (02) for mobiles and forward the calls to LCR. Maybe help someone other having my problem. I suggest Andreas Everberg to update his HowTo... And now the next problem: calling a VoIP-phone from mobile phone... :) I'll get it, too! Bye -- _______________________________________________________________________ Luca Bertoncello Entwicklung Mail: bertoncello(a)netzing.de NETZING Solutions AG Tel.: 0351/41381 - 0 Kesselsdorfer Str. 216, 01169 Dresden Fax: 0351/41381 - 12 HRB 18926 / Ust.ID DE211326547 Mail: netzing.ag(a)netzing.de _______________________________________________________________________

1 0

Re: [BUG REPORT]: Unable to send SMS
by Dieter Spaar 14 Jul '10

14 Jul '10

Hello Luca, On Wed, 14 Jul 2010 11:22:05 +0200, "Luca Bertoncello" <bertoncello(a)netzing.de> wrote: > > I can confirm, that "neci 1" solve my problem! > I can call and send SMS without any problem, from the NE110, too! Great to hear that. I am not sure how well OpenBSC is tested with NECI set to "1" but I guess it should not cause too many problems, its only the channel allocator which behaves differently. Best regards, Dieter -- Dieter Spaar, Germany spaar(a)mirider.augusta.de

1 0

Re: [BUG REPORT]: Unable to send SMS
by Dieter Spaar 14 Jul '10

14 Jul '10

Hello Holger, On Wed, 14 Jul 2010 17:52:51 +0800, "Holger Hans Peter Freyther" <holger(a)freyther.de> wrote: > > maybe I was too blind and didn't see it in GSM 04.08, but setting NECI=1 > will not allow us to assign a TCH/F for MO Call. What about this: 0100xxxx Originating speech call from dual-rate mobile station when TCH/H is sufficient and supported by the MS for speech calls and the network sets NECI bit to 1 At least this is what the Nokia 3310 uses for a voice call. OpenBSC should than allocate a TCH for it. Best regards, Dieter -- Dieter Spaar, Germany spaar(a)mirider.augusta.de

2 1

Re: [BUG REPORT]: Unable to send SMS
by Dieter Spaar 14 Jul '10

14 Jul '10

Hello Harald, On Tue, 13 Jul 2010 16:06:55 +0200, "Harald Welte" <laforge(a)gnumonks.org> wrote: > > The Netzing NE110 (as well as other MTK phones) seem to have a bug in > their MO-SMS code. Dieter has told me he is looking at a workaround to this > (by sending dummy authentication request/response), and he will e-mail this > list once he has some results. This was a tough one. I tried Authentication and Ciphering commands in various combinations without any success. Finally I changed the allocated channel to SDCCH instead of TCH. And voila, it worked, even without Authentication and/or Ciphering. So it seems that there is a problem with MTK phones (and maybe others) and MO-SMS on a TCH channel. I am not sure yet if there is a workaround to solve this and still use the TCH. Best regards, Dieter -- Dieter Spaar, Germany spaar(a)mirider.augusta.de

3 2

Re: [BUG REPORT]: Unable to send SMS
by Dieter Spaar 14 Jul '10

14 Jul '10

Hello Holger, On Wed, 14 Jul 2010 14:56:27 +0800, "Holger Hans Peter Freyther" <holger(a)freyther.de> wrote: > > do you remember which kind of channel the phone is requesting? is it > really asking for a channel for calls? Yes, the phone always sets the mask according to this (from the spec): 111xxxxx Originating call and TCH/F is needed, or originating call and the network does not set NECI bit to 1, or procedures that can be completed with a SDCCH and the network does not set NECI bit to 1 This was the same for other phones, not just MTK based phones. NECI was set to 0, however I think that setting NECI to 1 might be the solution for the problem, without the need to change the Very Early Assignment procedure. I have not yet tested this in detail, so I might miss something. Best regards, Dieter -- Dieter Spaar, Germany spaar(a)mirider.augusta.de

2 1

Re: [BUG REPORT]: Unable to send SMS
by Dieter Spaar 14 Jul '10

14 Jul '10

Hello Luca, On Wed, 14 Jul 2010 10:04:05 +0200, "Luca Bertoncello" <bertoncello(a)netzing.de> wrote: > > Wow! It works! And I can send SMS with the NE110 phones too! > Maybe could it be a solution to set lctype to GSM_LCHAN_SDCCH if the > phone tried to send an SMS? Harald probably knows the details better, but the basic problem is that the phone does not tell you exactly what channels it needs when it sends the Channel Request, at least if NECI is set to 0 (new establishment causes are not supported). So the channel allocator allocates a TCH channel, to be able to handle everything (SMS and Voice). One solution for this is to switch from Very Early Assignment to other assignment procedures. I think Holger and Harald have plans to support this. However there also seems to be a different solution: You can set NECI to 1, this way the phone can tell you in more detail what it wants. I have not done extensive tests, but it seems that this could solve the problem without the need to change the code. You just have to set "neci 1" in the OpenBSC configuration file. Best regards, Dieter -- Dieter Spaar, Germany spaar(a)mirider.augusta.de

2 2

Re: [BUG REPORT]: Unable to send SMS
by Dieter Spaar 14 Jul '10

14 Jul '10

Hello Luca, On Wed, 14 Jul 2010 08:45:49 +0200, "Luca Bertoncello" <bertoncello(a)netzing.de> wrote: > > Of course! I tested OpenBSC with GSM1800, too. And, unfortunately, it > does not work... Attached the pcap-file and the config of OpenBSC. > Here some information from ipaccess-find: If you want to find out if the SMS sending problem is related to what I posted yesterday, you can try the following: In "/src/abis_rsl.c", function "rsl_rx_chan_rqd()" replace the line lctype = get_ctype_by_chreq(bts, rqd_ref->ra, bts->network->neci); with lctype = GSM_LCHAN_SDCCH; This will always allocate an SDCCH and you can check if SMS sending now works. However this is only for testing, voice calls will no longer work with this change. Best regards, Dieter -- Dieter Spaar, Germany spaar(a)mirider.augusta.de

2 1

Re: [BUG REPORT]: Unable to send SMS
by Luca Bertoncello 14 Jul '10

14 Jul '10

Am Wed, 23 Jun 2010 20:18:56 +0800 schrieb Holger Freyther <zecke(a)selfish.org>: > the first pointer is to the sha1/commit id[1]. So every commit has an > id, it happens to be the sha1 hash over some parts of the content. Now > git log will show all these as "commit". Now the openbsc version is > generated by the little "./git-version-gen" utility. In fact the > utility is calling "git describe" (man git-describe). > > E.g. it looks like this: > 0.9.0-891-gaf7edd9 > > 0.9.0 is the last tag we had.. > 891 is the number of commits since that tag > gaf7edd9 is the commit/sha1. This is what you want. > > I hope this helps Unfortunately not... I tried to give git-describe in the tree of the "sms-sending"-BSC (0.9.0-531-gb938d6b) and in the tree of the last version (0.9.0-890-g2788b96), then I tried to use git bisect: lucabert@Luca:/tmp/bsc/openbsc$ git bisect good gb938d6b Bad rev input: gb938d6b lucabert@Luca:/tmp/bsc/openbsc$ git bisect bad g2788b96 fatal: Needed a single revision Bad rev input: g2788b96 It seems not to work... Maybe I need vacancy? I can't understand what you suppose I can do... Thanks for your suggestion! -- _______________________________________________________________________ Luca Bertoncello Entwicklung Mail: bertoncello(a)netzing.de NETZING Solutions AG Tel.: 0351/41381 - 0 Kesselsdorfer Str. 216, 01169 Dresden Fax: 0351/41381 - 12 HRB 18926 / Ust.ID DE211326547 Mail: netzing.ag(a)netzing.de _______________________________________________________________________

4 28

[PATCH] * rewrite GSM 7bit default encoding/decoding based on a lookup table as the previous code produced wrong encodings for certain characters.
by Nico Golde 09 Jul '10

09 Jul '10

--- include/osmocore/gsm_utils.h | 28 ++++++++++++ src/gsm_utils.c | 99 +++++++++++++++++++++++++++++++++-------- tests/sms/sms_test.c | 17 +++++++- 3 files changed, 123 insertions(+), 21 deletions(-) diff --git a/include/osmocore/gsm_utils.h b/include/osmocore/gsm_utils.h index 7dc2388..64f9edc 100644 --- a/include/osmocore/gsm_utils.h +++ b/include/osmocore/gsm_utils.h @@ -3,6 +3,7 @@ * (C) 2008 by Daniel Willmann <daniel(a)totalueberwachung.de> * (C) 2009 by Holger Hans Peter Freyther <zecke(a)selfish.org> * (C) 2009-2010 by Harald Welte <laforge(a)gnumonks.org> + * (C) 2010 by Nico Golde <nico(a)ngolde.de> * * All Rights Reserved * @@ -53,6 +54,33 @@ enum gsm_band { GSM_BAND_810 = 0x80, }; +/* ETSI GSM 03.38 6.2.1 and 6.2.1.1 default alphabet + * Greek symbols at hex positions 0x10 and 0x12-0x1a + * left out as they can't be handled with a char and + * since most phones don't display or write these + * characters this would only needlessly make the code + * more complex +*/ +unsigned char gsm_7bit_alphabet[] = { + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x0a, 0xff, 0xff, 0x0d, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0x20, 0x21, 0x22, 0x23, 0x02, 0x25, 0x26, 0x27, 0x28, 0x29, 0x2a, 0x2b, 0x2c, + 0x2d, 0x2e, 0x2f, 0x30, 0x31, 0x32, 0x33, 0x34, 0x35, 0x36, 0x37, 0x38, 0x39, 0x3a, 0x3b, + 0x3c, 0x3d, 0x3e, 0x3f, 0x00, 0x41, 0x42, 0x43, 0x44, 0x45, 0x46, 0x47, 0x48, 0x49, 0x4a, + 0x4b, 0x4c, 0x4d, 0x4e, 0x4f, 0x50, 0x51, 0x52, 0x53, 0x54, 0x55, 0x56, 0x57, 0x58, 0x59, + 0x5a, 0x3c, 0x2f, 0x3e, 0x14, 0x11, 0xff, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, + 0x69, 0x6a, 0x6b, 0x6c, 0x6d, 0x6e, 0x6f, 0x70, 0x71, 0x72, 0x73, 0x74, 0x75, 0x76, 0x77, + 0x78, 0x79, 0x7a, 0x28, 0x40, 0x29, 0x3d, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0x0c, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x5e, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x40, 0xff, 0x01, 0xff, + 0x03, 0xff, 0x7b, 0x7d, 0xff, 0xff, 0xff, 0xff, 0xff, 0x5c, 0xff, 0xff, 0xff, 0xff, 0xff, + 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x5b, 0x7e, 0x5d, 0xff, 0x7c, 0xff, 0xff, 0xff, + 0xff, 0x5b, 0x0e, 0x1c, 0x09, 0xff, 0x1f, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x5d, + 0xff, 0xff, 0xff, 0xff, 0x5c, 0xff, 0x0b, 0xff, 0xff, 0xff, 0x5e, 0xff, 0xff, 0x1e, 0x7f, + 0xff, 0xff, 0xff, 0x7b, 0x0f, 0x1d, 0xff, 0x04, 0x05, 0xff, 0xff, 0x07, 0xff, 0xff, 0xff, + 0xff, 0x7d, 0x08, 0xff, 0xff, 0xff, 0x7c, 0xff, 0x0c, 0x06, 0xff, 0xff, 0x7e, 0xff, 0xff +}; + const char *gsm_band_name(enum gsm_band band); enum gsm_band gsm_band_parse(const char *mhz); diff --git a/src/gsm_utils.c b/src/gsm_utils.c index dc97cef..fb69377 100644 --- a/src/gsm_utils.c +++ b/src/gsm_utils.c @@ -2,6 +2,7 @@ * (C) 2008 by Daniel Willmann <daniel(a)totalueberwachung.de> * (C) 2009 by Holger Hans Peter Freyther <zecke(a)selfish.org> * (C) 2009-2010 by Harald Welte <laforge(a)gnumonks.org> + * (C) 2010 by Nico Golde <nico(a)ngolde.de> * * All Rights Reserved * @@ -34,52 +35,110 @@ #include "../config.h" -/* GSM 03.38 6.2.1 Charachter packing */ +/* GSM 03.38 6.2.1 Character lookup for decoding */ +static int gsm_septet_lookup(uint8_t ch) +{ + int i = 0; + for(; i < sizeof(gsm_7bit_alphabet); i++){ + if(gsm_7bit_alphabet[i] == ch) + return i; + } + return -1; +} + +/* GSM 03.38 6.2.1 Character unpacking */ int gsm_7bit_decode(char *text, const uint8_t *user_data, uint8_t length) { int i = 0; int l = 0; + uint8_t *rtext = calloc(length, sizeof(uint8_t)); + uint8_t tmp; - /* FIXME: We need to account for user data headers here */ + /* FIXME: We need to account for user data headers here */ i += l; - for (; i < length; i ++) - *(text ++) = + for (; i < length; i ++){ + rtext[i] = ((user_data[(i * 7 + 7) >> 3] << (7 - ((i * 7 + 7) & 7))) | (user_data[(i * 7) >> 3] >> ((i * 7) & 7))) & 0x7f; + } + for(i = 0; i < length; i++){ + /* this is an extension character */ + if(rtext[i] == 0x1b){ + tmp = rtext[i+1]; + *(text++) = gsm_7bit_alphabet[0x7f + tmp]; + i++; + continue; + } + + *(text++) = gsm_septet_lookup(rtext[i]); + } + *text = '\0'; + free(rtext); - return i - l; + return i; } +/* GSM 03.38 6.2.1 Prepare character packing */ +static int gsm_septet_encode(uint8_t *result, const char *data) +{ + int i, y = 0; + uint8_t ch; + for(i = 0; i < strlen(data); i++){ + ch = data[i]; + switch(ch){ + /* fall-through for extension characters */ + case 0x0c: + case 0x5e: + case 0x7b: + case 0x7d: + case 0x5c: + case 0x5b: + case 0x7e: + case 0x5d: + case 0x7c: + result[y++] = 0x1b; + default: + result[y] = gsm_7bit_alphabet[ch]; + break; + } + y++; + } -/* GSM 03.38 6.2.1 Charachter packing */ + return y; +} + +/* GSM 03.38 6.2.1 Character packing */ int gsm_7bit_encode(uint8_t *result, const char *data) { - int i,j = 0; - unsigned char ch1, ch2; + int i,y,z = 0; + /* prepare for the worst case, every character expanding to two bytes */ + uint8_t *rdata = calloc(strlen(data) * 2, sizeof(uint8_t)); + uint8_t cb, nb; int shift = 0; - for ( i=0; i<strlen(data); i++ ) { + y = gsm_septet_encode(rdata, data); - ch1 = data[i] & 0x7F; - ch1 = ch1 >> shift; - ch2 = data[(i+1)] & 0x7F; - ch2 = ch2 << (7-shift); + for(i = 0; i < y; i++) { + if(shift == 7 && i + 1 < y){ + shift = 0; + continue; + } - ch1 = ch1 | ch2; + cb = (rdata[i] & 0x7f) >> shift; + if(i + 1 < y){ + nb = (rdata[i + 1] & 0x7f) << (7 - shift); + cb = cb | nb; + } - result[j++] = ch1; + result[z++] = cb; shift++; - - if ((shift == 7) && (i+1<strlen(data))) { - shift = 0; - i++; - } } + free(rdata); return i; } diff --git a/tests/sms/sms_test.c b/tests/sms/sms_test.c index f5183d5..4daf003 100644 --- a/tests/sms/sms_test.c +++ b/tests/sms/sms_test.c @@ -1,5 +1,6 @@ /* * (C) 2008 by Daniel Willmann <daniel(a)totalueberwachung.de> + * (C) 2010 by Nico Golde <nico(a)ngolde.de> * All Rights Reserved * * This program is free software; you can redistribute it and/or modify @@ -32,7 +33,7 @@ int main(int argc, char** argv) uint8_t *sms; uint8_t i; - /* test 7-bit coding/decoding */ + /* test 7-bit coding/decoding */ const char *input = "test text"; uint8_t length; uint8_t coded[256]; @@ -43,5 +44,19 @@ int main(int argc, char** argv) if (strcmp(result, input) != 0) { printf("7 Bit coding failed... life sucks\n"); printf("Wanted: '%s' got '%s'\n", input, result); + return -1; } + + memset(coded, 0, sizeof(coded)); + memset(result, 0, sizeof(coded)); + input = strdup("!$ a more#^- complicated test@@?_\%! case"); + length = gsm_7bit_encode(coded, input); + gsm_7bit_decode(result, coded, length); + if (strcmp(result, input) != 0) { + printf("7 Bit coding failed... life sucks\n"); + printf("Wanted: '%s' got '%s'\n", input, result); + return -2; + } + + return 0; } -- 1.7.1 --ibTvN161/egqYuK8--

1 0