Commit Graph

11 Commits

Author SHA1 Message Date
Harald Welte 6e2150ee0c sw/e1tracer: Rename 'main' to 'e1-tracer-record'
Change-Id: If70429e6a0504c657f44f9d0301a128d69ce7c18
2022-10-31 08:38:04 +01:00
Harald Welte 1d062b82c5 sw/e1-tracer: split recorder + analysis tools in sub-directories
this way it's easier to see which part of the code is used where

Change-Id: I7d727263383ec1d6d61dd31fcfebf4a4b1a04765
2022-10-31 08:37:26 +01:00
Harald Welte f5e7264a27 import latest e1-tracer host software
This was originally in https://github.com/smunaut/ice40-playground
but has meanwhile been removed, only a fork at
https://github.com/laf0rge/ice40-playground remained.

Let's import the code here.  The history is not all that exciting
so I'm saving myself the effort of git filter-branch or the like
to import the actual old commits.

The commit log is here for reference:

commit bd36d40cc30acbae68978df79eff03f0539a9a21
Author: Harald Welte <laforge@osmocom.org>
Date:   Fri Sep 4 10:21:51 2020 +0200

    e1-recorder main.c: Don't exist just because one ISO transfer failed

    I get LIBUSB_TRANSFER_ERROR every so often (once every few hours) on my
    laptop.  Sure, it likely means there was some data lost, but the trace
    can continue juts fine after this error.  So make it non-fatal and
    continue running.

commit 9b02b3e87400ba3d27bff05e2632fbd5a0698e8c
Author: Harald Welte <laforge@osmocom.org>
Date:   Mon Aug 31 10:01:20 2020 +0200

    Add 'replay' to re-play 2Mbps stream with realistic speed (to stdout)

commit 23103d3a8f9cc8b3f0ba602dbb4240040b2d008a
Author: Harald Welte <laforge@osmocom.org>
Date:   Sun Aug 30 15:41:12 2020 +0200

    dump: Use osmo_e1f to decode; send HLDC over GSMTAP

commit 7acd063a1245a34835d4488be3caad592f3c19d5
Author: Harald Welte <laforge@osmocom.org>
Date:   Thu Aug 20 16:48:15 2020 +0200

    dump.c: Add error plotting support

commit 90b90d04fff494b70ad220bd82ac48d43f3096c9
Author: Harald Welte <laforge@osmocom.org>
Date:   Mon Jan 13 18:20:39 2020 +0100

    dump: Add support for writing LAPD PCAP of E1 superchannel

commit f3a0cc34f0901c2cf257367351dddfddc6f7301e
Author: Harald Welte <laforge@osmocom.org>
Date:   Mon Jan 13 16:37:23 2020 +0100

    dump.c: Move handling of frames to separate function

commit 8deffb5907979e43e3a4236ea7bb8d54513513ac
Author: Harald Welte <laforge@osmocom.org>
Date:   Mon Jan 13 16:34:39 2020 +0100

    dump.c: remove local hexdump.[ch] and use libosmocore proper

commit 80d81e78737237ba3b6bae1deb7d5107d7f613b6
Author: Harald Welte <laforge@osmocom.org>
Date:   Tue Sep 1 13:52:52 2020 +0200

    tracer: use 0x6151 as PID

    See 5e68831b26

commit b219af27b34bbd6deb2419aa38852d8d227b9fc0
Author: Harald Welte <laforge@osmocom.org>
Date:   Mon Jan 6 13:47:28 2020 +0100

    e1-recorder: Don't silently ignore write failures

    The disk may be full, for example. We should report that to the user
    and terminate the program.

commit ead99fdfae6f1db54c79a91ef407a86c67352d87
Author: Harald Welte <laforge@osmocom.org>
Date:   Mon Jan 6 12:41:23 2020 +0100

    e1-recorder: Exit process on transfer + resubmit failure

    This is particularly importnat in case the USB device for some reason
    gets disconnected and re-connects.  We want the process to fail fast
    and have systemd respawn us.

commit a521e026c76056a60531c490fa2ef84e8e633c9a
Author: Harald Welte <laforge@osmocom.org>
Date:   Mon Jan 6 12:35:20 2020 +0100

    e1-recorder: use libusb_strerror() to decode error messages

commit 43a8fbdfdc35f5ba7564cfe5c92f43b124abf710
Author: Harald Welte <laforge@osmocom.org>
Date:   Sun Jan 5 19:49:58 2020 +0100

    e1-recorder: Acytually make '-r' (SCHED_RR) work

commit 84e28225d9ba1de9fb09660a96bd3b0b798c88d9
Author: Harald Welte <laforge@osmocom.org>
Date:   Sun Jan 5 17:53:11 2020 +0100

    e1-recorder: Add small 'dump' program

    This program will print one line for each E1 frame, where each line
    consists of 32 hex bytes: one for each timeslot

commit 0c697b40620b5d18480e626991dc639daadf58a6
Author: Sylvain Munaut <tnt@246tNt.com>
Date:   Sun Dec 29 13:13:17 2019 +0100

    projects/riscv_usb: Host software for the dual channel sniffer

    Signed-off-by: Sylvain Munaut <tnt@246tNt.com>

Closes: OS#5673
Change-Id: I9319c1eb8f822830307c1a181d357c58ce43efba
2022-10-31 08:28:43 +01:00
Harald Welte 9e7a21ae58 osmo_e1f: Allow to transmit AIS
AIS is an all-1 bit-pattern which is used by intermediate equipment
to indicate that the link somewhere upstream is lost.

Change-Id: Icc5a16ee44ecebe802c4cca0b6cc763a88b6cc05
2021-01-05 23:25:38 +01:00
Harald Welte 6360c07d91 osmo_e1f: Fix inverted logic of E bits
A value of '1' means _no_ CRC4 error observed. We got that right
in the receive side, but not in the transmit side so far

Change-Id: I5ef9ef3b9781da2f53f54253d31cdbf81957ea67
2020-12-19 12:34:40 +01:00
Harald Welte d9a8f9a07c osmo_e1f.c: Fix CRC4 insertion
When encoding the CRC bit for Frame number 0 and 8 in the multiframe,
we must first move the CRC4 into those of the last SMF.

Change-Id: I088741fc4528f33b3b989ada0e21957456deedb6
2020-12-19 12:08:17 +01:00
Harald Welte 2e7284b5be osmo_e1f: Fix use of uninitized variable
Change-Id: I42024856a58f0bd4fcc1c767688de5334994a8a8
2020-12-19 12:08:17 +01:00
Harald Welte 61a23eaf5d osmo_e1f.c: sa4..sa8 bits should be all-1 by default
From ITU-T G.704 Table 5A Note 4:

> Bits S a4 to S a8 (where these are not used) should be set to 1 on
> links crossing an international border.

Change-Id: Id55599a7152a90f85076de810db2a03ac6f3aac1
2020-12-19 12:08:17 +01:00
Harald Welte 7d4ba34552 add udev.rules files for icE1usb and e1-tracer
Change-Id: Ie1be28359def6fc181f88558834ac52a001312b7
2020-12-15 08:48:26 +01:00
Harald Welte 2aa6d028a8 software/obsloete: Add sockbuf_test.c from my local tree 2020-09-04 14:57:52 +02:00
Harald Welte 6719ac2019 move old software snippets and hacks to software/obsolete 2020-09-04 14:56:45 +02:00