AgeCommit message (Collapse)AuthorFilesLines
2017-01-24move variable declarations out of the 'for' loopHEADmasterHarald Welte2-2/+4
this fixes compilation with gcc-4.8
2017-01-24add missing gprs_l1.c fileHarald Welte1-0/+47
2017-01-24Add python script to generate value_string from file with #definesHarald Welte1-0/+39
2017-01-24python scripts: Generate file header with #include statementsHarald Welte2-11/+34
2017-01-24qxdm_db.py: Move value_string code generator to separate fileHarald Welte2-10/+30
2017-01-24Add python script to generate 'struct value_string' from enumHarald Welte1-0/+60
The idea is that you have an input file that defines an enum with symbolic names for numeric values. The script then generates a 'struct value_string' with corresponding strings.
2017-01-22add *.pyc to .gitignoreHarald Welte1-0/+1
2017-01-22Add python tools that (partially) parse the QXDM databaseHarald Welte2-0/+198
2017-01-22diag_msg: Mark dispatch table as 'const'Harald Welte1-1/+1
2017-01-21Add support for DIAG EVENT reportingHarald Welte5-10/+170
2017-01-21diag_msg: Don't replace second character of every msg with 'p'Harald Welte1-1/+1
We want to replace all %s with %p, but not modify other parts of the format string...
2017-01-21fix various compilation warnings all over the codeHarald Welte5-11/+12
2017-01-21decode L2 channel typeHarald Welte3-4/+31
2017-01-21decoder for RLC-EPDAN, MDSP-CMD, L2-STATE and L2-TRANSM-STATUSHarald Welte4-6/+227
2017-01-20Enable + Interpret more GPRS related loggingHarald Welte5-6/+227
Those logging groups are not present on all firmwares/devices. Best success for GPRS low-layer logging so far with an old Gobi2000 card from Lenovo.
2017-01-11gprs: handle RLC uplink/downlink ack/nack paramsHarald Welte2-12/+33
2017-01-09WIP: More WCDMA/RRC decodingHarald Welte2-1/+160
2017-01-09diag_log_gprs_rlc.h: More log message definitionsHarald Welte1-0/+39
2017-01-09Add (E)GPRS MAC ACK/NACK log supportHarald Welte1-0/+20
2017-01-09diag_log_gprs: Cosmetic change in re-ordering the log_dispatch_tblHarald Welte1-6/+11
2017-01-09add LLC XID information decodeHarald Welte2-0/+51
2017-01-09update .gitignoreHarald Welte1-1/+6
2017-01-09add missing diag_dpl.h to repoHarald Welte1-0/+11
2017-01-08WIP: SIM Card loggingHarald Welte3-1/+57
I'm getting related messages from the modem, but I somehow cannot decode the traces so far at all :/
2017-01-08fix diag_umts_rrc_chtype and add more diag_log_code_wcdmaHarald Welte1-2/+6
2017-01-08fix gen_log_config_set_mask()Harald Welte1-2/+12
We were missing the last byte in the log config mask, resulting in unintended behavior.
2017-01-08also log config request/response messages via GSMTAPHarald Welte1-0/+2
2017-01-08patch %s out of format strings received from phoneHarald Welte1-0/+6
%s can obviously not work, as we have no access to the targets memory space on the host PC. I guess the programmers adding those log lines have done so in error. Let's replace %s with %p in the format string.
2017-01-08separate diag_log_umts to diag_log_wcdmaHarald Welte6-16/+65
Qualcomm differentiates between WCDMA (the access stratum) and UMTS (the non access stratum). Let's reflect that here. As an added bonus, we get working NAS protocol traces.
2017-01-08Generate GSMTAP messages from raw received DIAG framesHarald Welte3-0/+17
This forwards the raw DIAG messages via GSMTAP, so the receiver (e.g. wireshark) will have to do a full DIAG protocol decode. I currently prefer this idea to that of converting only the protocol payload to "native" GSMTAP messages like GSMTAP_UM. One of the problems is that the LAPDm headers are alrady stripped, and we would have to re-add fake LAPDm headers to generate GSMTAP_UM. So let's rather forward all information we have and let wireshark deal with it. I'm not entirely sure if this is the best strategy, but we can always implement both modes and switch between them at runtime.
2017-01-08diag_read_msg(): Discard any messages shorter than 3 bytesHarald Welte1-0/+6
The 16bit CRC and the FLAG octet together are 3 bytes. If a message is shorter than that, something is going wrong, for sure.
2017-01-07WIP: Enable PCAP generation in modemHarald Welte5-10/+60
2017-01-07Genreate GSMTAP messages from raw received DIAG framesHarald Welte1-0/+8
2017-01-02replace old magic numbers with proper #definesHarald Welte2-2/+2
2017-01-02diag_io.c: remove left-over 'rc=%d' printHarald Welte1-1/+0
2017-01-02DPL: consequently use struct dpl_iid rather than uint32_tHarald Welte2-5/+6
2017-01-01DPL: add parsing of results from modemHarald Welte3-8/+49
we can now get a list of DPL interfaces and their links.
2017-01-01diag_push_subsys_hdr(): subsystem code is 16bitsHarald Welte2-2/+2
2017-01-01diag_cmd: Print error messages for bad {command, length, parameter}Harald Welte4-2/+12
2017-01-01add printing of DIAG_QSR_EXT_MSG_TERSE_FHarald Welte2-32/+65
Those are QSHRINK type messages where the format-string is not contained in the message itself but would have to be provided from an external source and looked up by a hash.
2017-01-01register diag_msg via the diag_cmd dispatch tableHarald Welte3-9/+10
2017-01-01don't use magic numbers if we have #definesHarald Welte1-4/+4
2017-01-01fixup dplHarald Welte1-1/+1
2017-01-01diag_msg: endian fixesHarald Welte1-5/+13
2017-01-01diag_msg: Include subsys_id in printed messageHarald Welte1-1/+1
2017-01-01diag_log: rough verification if response matches requestHarald Welte1-1/+3
2017-01-01remove Tx/Rx debugging from diag_io.cHarald Welte1-2/+0
2017-01-01diag_msgc: Parse responsesHarald Welte2-3/+19
2017-01-01add ability to hexdump all received and transmitted messagesHarald Welte2-0/+10
2017-01-01restructuring: Introduce _transceive_ functions for Tx-then-RxHarald Welte10-73/+208