mirror of
https://github.com/lwvmobile/dsd-fme.git
synced 2026-11-16 10:47:21 -05:00
DMR: PDU / LRRP Misc Fixes, Tweaks, and Cleanup;
This commit is contained in:
@@ -168,6 +168,8 @@ void dmr_dheader (dsd_opts * opts, dsd_state * state, uint8_t dheader[], uint8_t
|
||||
else if (sap == 5) sprintf (sap_string, "%s", "ARP Prot"); //Address Resoution Protocol (ARP)
|
||||
else if (sap == 9) sprintf (sap_string, "%s", "EXTD HDR"); //Extended Header (Proprietary)
|
||||
else if (sap == 10) sprintf (sap_string, "%s", "Short DT"); //Short Data
|
||||
else if (sap == 1 && p_mfid == 0x10)
|
||||
sprintf (sap_string, "%s", "Moto EXT"); //motorola extended format
|
||||
else sprintf (sap_string, "%s", "Reserved"); //reserved, or err/unk
|
||||
|
||||
//mfid string handling
|
||||
@@ -317,29 +319,32 @@ void dmr_dheader (dsd_opts * opts, dsd_state * state, uint8_t dheader[], uint8_t
|
||||
//The SAP found here is the actual SAP of the message (like a P25 ndary SAP, and can chain together according to ETSI)
|
||||
fprintf (stderr, " - SAP %02d [%s] - MFID %02X [%s]", p_sap, sap_string, p_mfid, mfid_string);
|
||||
|
||||
//p_sap 1 on mfid 10 (moto) is unknown, and the messages don't pass the CRC32
|
||||
//two extra duplicate blocks observed only happens on one sample, could have been radio or repeater errors?
|
||||
// if (p_sap != 1)
|
||||
//p_sap 1 on mfid 10 (moto) has been observed as the first block of LRRP data on a Moto system but doesnt' pass the CRC for some reason (included or not included)
|
||||
if (p_mfid == 0x10 && p_sap == 1)
|
||||
{
|
||||
//this method will include this extended header since its observed to be MOTO LRRP first block
|
||||
//NOTE: This is broken on R34 data (expects only R12U data)
|
||||
//TODO: Redo block assembly at some point (or store this seperately?)
|
||||
int blocks = state->data_header_blocks[slot]-1;
|
||||
if (blocks > 127) blocks = 127;
|
||||
if (blocks < 1) blocks = 4; //fix this at some point
|
||||
for(uint16_t i = 0; i < 12; i++)
|
||||
state->dmr_pdu_sf[slot][i+(blocks*12)] = dheader[i];
|
||||
state->data_block_counter[slot]++;
|
||||
}
|
||||
|
||||
else //if (p_sap != 1) //anything else
|
||||
{
|
||||
//sanity check to prevent segfault (this happened when the regular header was not received beforehand)
|
||||
if (state->data_header_blocks[slot] > 1)
|
||||
state->data_header_blocks[slot]--;
|
||||
}
|
||||
// else
|
||||
// {
|
||||
// //this method will just include the prop_pdu header as a data block and assemble the data blocks as a message, excluding duplicates
|
||||
// int blocks = state->data_header_blocks[slot]-1;
|
||||
// if (blocks > 127) blocks = 127;
|
||||
// if (blocks < 1) blocks = 4;
|
||||
// for(i = 0; i < 12; i++)
|
||||
// state->dmr_pdu_sf[slot][i+(blocks*12)] = dheader[i];
|
||||
// state->data_block_counter[slot]++;
|
||||
// }
|
||||
|
||||
//Start Setting DMR Data Packet Encryption Variables
|
||||
if (p_sap != 1 && p_mfid == 0x10) //p_sap 1 is reserved and doesn't seem to contain ENC header (may be extended address, or other signalling)
|
||||
if (p_sap != 1 && p_mfid == 0x10)
|
||||
{
|
||||
//check ENC bit, assuming this is an ENC bit, or SVC OPT like thing
|
||||
|
||||
//check ENC bit, assuming this is an ENC bit, or SVC OPT like thing (or could be an opcode for the rest of the extended header)
|
||||
if ((uint8_t)ConvertBitIntoBytes(&dheader_bits[20], 4) == 1)
|
||||
{
|
||||
//set to 0x100 so it won't trigger any weird flags, but still has a non-zero value to be checked later
|
||||
@@ -376,18 +381,13 @@ void dmr_dheader (dsd_opts * opts, dsd_state * state, uint8_t dheader[], uint8_t
|
||||
fprintf (stderr, " MI(32): %08X", (uint32_t)ConvertBitIntoBytes(&dheader_bits[48], 32));
|
||||
|
||||
}
|
||||
else if (p_sap == 1 && p_mfid == 0x10)
|
||||
{
|
||||
//This can be an LRRP packet first block (but slightly different configuration)
|
||||
fprintf (stderr, "\n Motorola Alternate Function Header; ");
|
||||
}
|
||||
else //if (p_mfid == 0x10)
|
||||
{
|
||||
//I wonder if this is like the extended addressing header used on P25p1
|
||||
//(either way, the CRC32 will still fail, maybe its RAS or a MAC?)
|
||||
/*Unknown Extended Header: 1F1002011127E70D20232018
|
||||
Slot 2 - Multi Block PDU Message CRC32 ERR
|
||||
Slot 2 - Multi Block PDU Message
|
||||
1F1002011127E70D20232018 <--fails CRC32 with or without this attached
|
||||
41341F9AAECDE6554616A43C
|
||||
0E2A6335040C872A0009006C
|
||||
007356B370000000458AB84D <--very clearly 32 bits (4 bytes) after padding
|
||||
*/ //End Example
|
||||
fprintf (stderr, "\n Unknown Extended Header: ");
|
||||
for (uint8_t i = 2; i < 10; i++)
|
||||
fprintf (stderr, "%02X", (uint8_t)ConvertBitIntoBytes(&dheader_bits[0+(i*8)], 8));
|
||||
@@ -836,6 +836,29 @@ void dmr_block_assembler (dsd_opts * opts, dsd_state * state, uint8_t block_byte
|
||||
uint16_t len = ((blocks+1)*block_len)-4; //total number of bytes in PDU minus 4 CRC32 bytes
|
||||
dmr_udp_comp_pdu (opts, state, len, state->dmr_pdu_sf[slot]);
|
||||
}
|
||||
else if (state->data_header_sap[slot] == 1 && state->dmr_pdu_sf[slot][1] == 0x10) //test SAP 1 MFID 10 (with -F) as a potential LRRP message (header inclusive)
|
||||
{
|
||||
uint16_t len = ((blocks+1)*block_len)-4;
|
||||
|
||||
//sanity check
|
||||
if (len > 150)
|
||||
len = 150;
|
||||
|
||||
uint8_t PDU[150]; memset (PDU, 0, sizeof(PDU)); //arbitrary size (should be okay)
|
||||
|
||||
//manipulate to match expected format alignment
|
||||
PDU[0] = 0; PDU[1] = state->dmr_pdu_sf[slot][0];
|
||||
|
||||
//speculation: the other 4 bytes are hashed values for dst and src, or something similar
|
||||
|
||||
uint16_t mprt = 0xFFFF;
|
||||
uint32_t msrc = state->dmr_lrrp_source[slot];
|
||||
uint32_t mdst = 0xFFFFFF; //TODO: Get the real DST or TG indicated in the header
|
||||
|
||||
//copy rest, starting at offset 5 of the reserved extended header, to offset 2 to match expected formatting
|
||||
memcpy(PDU+2, state->dmr_pdu_sf[slot]+5, (sizeof(PDU)-5));
|
||||
dmr_lrrp (opts, state, len-4, msrc, mdst, mprt, mprt, PDU);
|
||||
}
|
||||
}
|
||||
|
||||
if (CRCCorrect) ; //print nothing
|
||||
|
||||
Reference in New Issue
Block a user