I can only state that:
ssp_x86Linux_boot: QERROR_WARNING(2) qkernerr.h 41400 qkernerr.h 41003
Points towards failure to open a Gstream because that stream id is already open. Meaning the gstream is used to pass the data down from host library to firmware, and seems it was not cleanup properly from prior usage.
Maybe there is some issue on the tear down of faxes once they complete?
Does this happen all the time, or only once in a while?
Jeff
Hi Chris,
Thank you.
I'm going to verify these points.
Yvan
Yes, Chris raises a good point. I have seen this occur on newer SUs because OpenSSL libs were recently upgraded in HMP. Thus if system has other software installed which also uses OpenSSL it may cause problems/conflicts if the PATH leads HMP to find those older files first as opposed to loading the ones expected.
Meaning within gc_start, it will dynamically loads the two files (libeay32.dll and ssleay32.dll) within the libgch3r library. If it fails you may see the following typw of error returned:
Error: gc_Start(), GC ErrorValue: 0x126 - Invalid Call control library - may be corrupted, CCLibID: 0 - GLOBALCALL,CC ErrorValue: 0x126 - Invalid Call control library Error opening board device: (null)
Jeff
Hi,
I have a problem with HMP3.0 SU361 on windows 7 pro 64 bits
I have a version of HMP3.0 SU 347 which run very good. I have update this with HMP with SU361 and after a moment HMP crash!
In a debug mode, HMP crash in Cheetah thread.
Is this problem is known by everybody?
Thank's for your collaboration
Yvan
Hi,
it's a good reason if my crash arrive while HMP start.
But my problem arrive while HMP is running. My application run normaly and, after a moment, my application stop. In debug mode, my application is stopped in Cheetah tthread!
May be a problem with develop version of HMP.
I'm going to verify my develop environment.
Many thank's.
Yvan
My understanding is that 64 bit OS is supported but 64 bit applications are not supported by HMP, thus it is required to force the compiler to generate 32 bit code (-m32 option in gcc).
Assuming my understading is correct I wonder if a full support for 64 bit application is in roadmap for HMP, does anyone know the answer to this question?
Hello,
your code looks OK, except of 2 parts, First, you don't need to use dev_Connect for this app, dx_listen/ipm_Listen pair would do all you need. Second: In your snippet I don't see how sc_tsinfo is defined, as long as the timeslot pointer. It looks like the structure gets overwritten before xx_Listen is called. I'm attaching here some working code what is doing what your app needs, it plays, records, sends digits etc. Please give it a try, the interface should be self-explanatory. Please let me know if you have questions on using or reviewing it. It's a quite old demo, compiled with VS 6.0, but it's easy to turn to VS 9+,.
Forgot to add, that the code I sent is for Windows VS, but it can be used in Linux apps as far as the IPM and voice resources are concerned.
I'm running HMP 4.1 on Linux. I try to record voice data coming from a RTP endpoint (ipm device), but the record file contains only silence.
I'm not confident with functions like dev_Connect(), dx_listen() and ipm_Listen. I thought that dev_Connect would connect my ipm device to my voice device in both ways and then I could record voice received over IP with dx_rec(), but as said I only record silence.
Here is my pseudo-code, I do the following:
1. create 2 imp devices for 2 RTP endpoints
2. create 2 voice devices for play/record functions
3. connect them together
4. start media bothway of 2 IP endpoints (first endpoint sends data over IP to second endpoint, loop over IP)
5. listen on second endpoint
6. start recording on voice device 2
7. play wave file on vocie device 1
A wireshark trace shows that play is working fine, data is sent over IP to the second endpoint but I record only silence. All funtions return successfully.
long m_ipmCCH1 = ipm_Open("ipmB1C1", 0, EV_SYNC);
long m_ipmCCH2 = ipm_Open("ipmB1C2", 0, EV_SYNC);
long m_voxH1 = dx_open("dxxxB1C1", 0);
long m_voxH2 = dx_open("dxxxB1C2", 0);
dev_Connect(m_ipmCCH1, m_voxH1, DM_FULLDUP, EV_SYNC);
dev_Connect(m_ipmCCH2, m_voxH2, DM_FULLDUP, EV_SYNC);
ipm_StartMedia(m_ipmCCH1, &MediaInfo1, DATA_IP_TDM_BIDIRECTIONAL, EV_SYNC);
ipm_StartMedia(m_ipmCCH2, &MediaInfo2, DATA_IP_TDM_BIDIRECTIONAL, EV_SYNC);
sc_tsinfo.sc_numts = 1;
sc_tsinfo.sc_tsarrayp = &lVoiceTS;
dx_getxmitslot(m_voxH2, &sc_tsinfo);
sc_tsinfo.sc_numts = 1;
sc_tsinfo.sc_tsarrayp = &lLineTS;
ipm_GetXmitSlot(m_ipmCCH2, &sc_tsinfo, EV_SYNC);
sc_tsinfo.sc_numts = 1;
sc_tsinfo.sc_tsarrayp = &lVoiceTS;
ipm_Listen(m_ipmCCH2, &sc_tsinfo, EV_SYNC);
sc_tsinfo.sc_numts = 1;
sc_tsinfo.sc_tsarrayp = &lLineTS;
dx_listen(m_voxH2, &sc_tsinfo);
iott_r.io_type = IO_DEV|IO_EOT; // record to file
iott_r.io_bufp = 0;
iott_r.io_offset = 0;
iott_r.io_length = -1;
iott_r.io_fhandle = open("record.wav", O_RDWR|O_CREAT|O_TRUNC, 0666);
xpb_r.wFileFormat = FILE_FORMAT_WAVE;
xpb_r.wDataFormat = DATA_FORMAT_PCM;
xpb_r.nSamplesPerSec = DRT_8KHZ;
xpb_r.wBitsPerSample = 16;
dx_reciottdata(m_voxH2,&iott_r,NULL,&xpb_r,EV_ASYNC);
iott_p.io_type = IO_DEV|IO_EOT; // play from file
iott_p.io_bufp = 0;
iott_p.io_offset = 0;
iott_p.io_length = -1; /* play till end of file */
iott_p.io_fhandle = open("play.wav", O_RDONLY);
xpb_p.wFileFormat = FILE_FORMAT_WAVE;
xpb_p.wDataFormat = DATA_FORMAT_PCM;
xpb_p.nSamplesPerSec = DRT_8KHZ;
xpb_p.wBitsPerSample = 16;
dx_playiottdata(m_voxH,&iott_p,NULL,&xpb_p,EV_ASYNC);
What is wrong/missing?
Any idea?
I don't see any code, where did you send code?
I believe we have tested up to RH 6.7 in this case with the 2.6.x kernel.
There are no plans for RH 7 or 64-bit support right now, as the no new feature are being added currently.
The recommendation in this case, is to port to HMP 4.1 baseline for either new O/S distribution or 64 bit support.
Jeff
What is the latest version of RHEL supported in SR6.1 for Linux? The last update i saw was SU320 indicating support added for "Red hat Enterprise Linux 6". Does that mean version 6.0 only, or are other 6.x releases built upon the 2.6 kernel supported also? Are there plans to support version 7.0 or 7.1?
Also, was 64 bit support ever added to SR6.1 for Linux?
I have the following products:
HMP 4.1 Linux
DNI1210TEPE2HMP (4x E1)
License
Voice 150, Enhanced_RTP 30, Fax 150, RTP_G_711 30, AMR_NB 30, Native_Audio 30, Native_RTP 30, AMR_WB 30, G_722 30
When detecting the boards/devices with the SRL functions, I get the following:
2 physical boards:
Board name: brdB1
Board name: brdB2
Virtual board 1 on phys board 1: dxxxB1
Virtual board 2 on phys board 1: dxxxB2
Sub device on virtual board 1 of phys board 1: dxxxB1C1, dxxxB1C2, dxxxB1C3, dxxxB1C4
Sub device on virtual board 2 of phys board 1: dxxxB2C1, dxxxB2C2, dxxxB2C3, dxxxB2C4
Virtual board 1 on phys board 2: dtiB1
Virtual board 2 on phys board 2: dtiB2
Sub device on virtual board 1 of phys board 2: 31 devices
Sub device on virtual board 2 of phys board 2: 31 devices
Does that mean that I have only max 8 voice channels (dxxx)? According the license I should have 150...
The DNI board has 4 E1s, why do I see only 2?
Hi,
It stoped happening , it seems it wasn't issue on HMP but other software problem .
Thank You for Your reply.
Hello Mario,
Actually, you don't need to use 2 IPT boards to take an advantage of multiple NICs for RTP. Please take a look ch. 1.18 at www.dialogic.com/.../release_update.pdf; is this what you are looking for?
Hi,
Thank You for Your answer.
At first i didn't use 2 IPT virtboards and i did receive calls on both eth0 and eth1 but i had same problem with RTP so i assumed
when i opened 2 boards that automaticly RTP will belong to coresponding ethernet card but it didn't . I will look into this pdf You sugested .
Thanks
Hi,
I am using Linux HMP 4.1 dialogic release .
I am trying to build application with 2 ethernet cards with different IP adresses .
I opened 2 virtual boards (first one to ip adress of eth0 ,second one to ip adress of eth1) , succefully and each board has half of all IP channels (they all opened ok ).
When i receive calls on eth0 it is ok , in SDP i see RTP adress of eth0 .
When i receive calls on eth1 it is not ok , in SDP i see RTP adrress of eth0 and i would like to see adress of eth1 .
Does anyone know how to set this ?
Thanks
HI
I am facing the "Start pending" while I try to start the VFX Dialogic cards on Windows 7, I am using Release 6.0 DCM for windows 7.
Can someone let me know if VFX cards are compatible windows 7 at all? If yes then what I am I doing wrong?
Thanks
Hi again. I have been trying to send an out of session SIP message as you described using this code:
bool success = false;
GC_PARM_BLKP parameterBlock = nullptr;
GC_PARM_BLKP mimeBlock = nullptr;
GC_PARM_BLKP returnBlock = nullptr;
char *contentType = "Content-Type: text/plain";
char *messageBody = "Message body goes here";
char *userAgent = "User-Agent: VoiceGuideHMP";
std::ostringstream contactUriSS;
contactUriSS << "Contact: <sip:" << contactUri << ">;methods=\"INVITE, INFO, MESSAGE, SUBSCRIBE, BYE, CANCEL, NOTIFY, ACK, REFER\"";
const std::string formattedContactUri = contactUriSS.str();
if (gc_util_insert_parm_val(¶meterBlock, IPSET_MSG_SIP, IPPARM_MSGTYPE, sizeof(int), IP_MSGTYPE_SIP_MESSAGE) < 0 ||
gc_util_insert_parm_ref_ex(¶meterBlock, IPSET_SIP_MSGINFO, IPPARM_REQUEST_URI, static_cast<unsigned long>(formattedContactUri.length() + 1), const_cast<char *>(formattedContactUri.c_str())) < 0 ||
gc_util_insert_parm_ref(¶meterBlock, IPSET_SIP_MSGINFO, IPPARM_SIP_HDR, (strlen(userAgent) + 1), userAgent) < 0 ||
gc_util_insert_parm_ref_ex(&mimeBlock, IPSET_MIME, IPPARM_MIME_PART_TYPE, static_cast<unsigned long>(strlen(contentType)) + 1, contentType) < 0 ||
gc_util_insert_parm_val(&mimeBlock, IPSET_MIME, IPPARM_MIME_PART_BODY_SIZE, sizeof(unsigned long), strlen(messageBody)) < 0 ||
gc_util_insert_parm_val(&mimeBlock, IPSET_MIME, IPPARM_MIME_PART_BODY, sizeof(unsigned long), reinterpret_cast<unsigned long>(messageBody)) < 0 ||
gc_util_insert_parm_val(¶meterBlock, IPSET_MIME, IPPARM_MIME_PART, sizeof(unsigned long), reinterpret_cast<unsigned long>(mimeBlock)) < 0)
{
// log error
}
else if (gc_Extension(GCTGT_GCLIB_CHAN, ldid, IPEXTID_SENDMSG, parameterBlock, &returnBlock, EV_ASYNC) != GC_SUCCESS)
{
// log error
}
else
{
success = true;
}
if (mimeBlock != nullptr) gc_util_delete_parm_blk(mimeBlock);
if (parameterBlock != nullptr) gc_util_delete_parm_blk(parameterBlock);
if (returnBlock != nullptr) gc_util_delete_parm_blk(returnBlock);
return success;
However this is failing when gc_Extension() is called. I retrieve error info and am logging this:
[2016-03-01 11:08:06] (0x00000d50) [Err] gc_Extension() failed in Extension_SipMessage() with LDID 3
[2016-03-01 11:08:06] (0x00000d50) [Err] gc_ErrorInfo() results - GCValue: 68, GCMessage Invalid parameter, CCLibID: 8, CCLibName: GC_H3R_LIB, CCLibValue: 48, CCLibMessage: IPERR_INVALID_ID.
In the Dialogic RTF logs I see this:
03/01/2016 10:52:56.738 8784 12824 gc APPL gclib <:::: gc_Extension(target_type:12, target_id:40, exit_id:1, mode:async)
03/01/2016 10:52:56.738 8784 12824 gc_h3r SH_UNP..ER DEBG unpack.cpp:1339 ! 0 ! >> Unpack::reqExtension : pIptDevice=0x5a9cf98,hCrn=0,extId=1,pInfoParmBlk=0x65ea498,pRetParmBlk=0x6edade8
03/01/2016 10:52:56.738 8784 12824 gc_h3r ERR1 unpack.cpp:1364 ! 0 ! << Unpack::reqExtension : [48] - hCrn=0 is invalid
03/01/2016 10:52:56.738 8784 12824 gc_h3r WARN sm_callcontrol.:1196 ! 0 ! << Extension:reqExtension failed
03/01/2016 10:52:56.738 8784 12824 gc ERR1 gclib iptB1T19 ::::> gc_Extension(target_type:12, target_id:40, exit_id:1, mode:async) - returns:-1
03/01/2016 10:52:56.738 8784 12824 gc_h3r SH_MGR DEBG sm_error.cpp:198 ! 0 ! ShmErrorValue leaving OK...
I am wondering what I am doing wrong that would mean it was saying "hCrn=0 is invalid" when I am specifying GCTGT_GCLIB_CHAN.
Do you see any issues with my code? Should I be using IPPARAM_CONTACT_URI rather then IPPARAM_REQUEST_URI?
Thanks