aboutsummaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)AuthorFilesLines
2024-04-30ctl: When setting Sa bits, other bits must remain set to '1'tnt/workSylvain Munaut1-5/+8
Some of the other bits will be used 'as-is'. For instance the bit 0 (MSB) will be copied from this if operating in nocrc4 mode. Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: Id482c6a114e7bf051d157e81038f015e9d26fa99
2024-04-30iCE1usb: Transmit AIS events to clientsSylvain Munaut1-0/+12
Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I4a35a2ab2537cbc9ecec381ab5078c1b08067925
2024-04-30iCE1usb: Transmit LOS events to clientsSylvain Munaut1-1/+7
Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I8822d24c25f8ba0bb70e9f717ca04dbf0e656e71
2024-04-30Update ice1usb_proto.hSylvain Munaut1-0/+7
Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I8e041b3d94723eba9d3d30b9f03815d46dacf995
2024-04-30iCE1usb: Add support to configure CRC4 mode for TX/RXSylvain Munaut3-1/+72
Signed-off-by: Sylvain Munaut <tnt@246tNt.com> Change-Id: I9b627a9617af4ae9ed98214d672c7d0391f801eb
2024-03-23README.md: Improved markdown formatting; more linksHarald Welte1-6/+6
Change-Id: I7e4f1c8c527a63f13ef9eaa26a00e1b2efca82ae
2024-03-23Add funding link to github mirrorHarald Welte1-0/+1
see https://docs.github.com/en/repositories/managing-your-repositorys-settings-and-features/customizing-your-repository/displaying-a-sponsor-button-in-your-repository Change-Id: I4af5d8b94ef95f9187977a918d2ea9fe16360f94
2024-01-19Prevent memory leaks / double free of msgb in proto_clnt.cAndreas Eversberg1-3/+15
Change-Id: Ic0c4c06ba7c1a405ec3c73100fe9e8476f4f5512
2024-01-19Add command and client function to change transmitted Sa bitsAndreas Eversberg6-2/+78
Change-Id: I2608af7bbb8092fddd68d4f3bb36b10a1100ce0f
2024-01-19Indicate change in received Sa bitsAndreas Eversberg3-1/+15
Change-Id: I4ea2a11e221629a90ed227a0df33576c8607881b
2024-01-19Indicate remote alarm bit as remote alarm inidicationAndreas Eversberg2-4/+15
The flag that stores the current alarm is not cleared periodically. Instead it is cleared when the alarm ceases. Change-Id: Id6cd193c71330c350c27e02b3a692d2c7e0b3fbe
2024-01-19Indicate framing alignment error as loss of framing signalAndreas Eversberg1-0/+9
Change-Id: Iaad3adb6afb9e1d8a002fa4a04ba4ab8e51671a3
2024-01-19Add functions for events from server to clientAndreas Eversberg7-2/+61
The client may register a callback function to receive events. Because there is no relation between the connected client and the interface, all events are broadcasted to all clients that are connected to the server. Change-Id: I5ee3268f8349b611c3cf3fa0572dc5eab280ab2e
2024-01-19Add new protocol command and events for Sa bits and line statesAndreas Eversberg2-1/+42
Change-Id: I4c20ec19fa3d9cce82c8199b38d173aa272fb07d
2024-01-19Add option to automatically reset RIFO on underrun/overflowAndreas Eversberg11-13/+54
Whenever the RIFO buffer fill drifts away from its target, it can be automatically reset and filled to the initial prefill_frame_count value. The average fill is measured over several seconds. A given deviation in percent of the prefill_frame_count is used to trigger that reset. If the deviation is not set (0), this feature is deactivated. There are two reasons for this to happen: The GPS clock is missing, so the receiving interface is not in sync with the transmitting interface. The delay changes significantly, due to congestion on the path between both peers. (poor internet connection) Change-Id: Id7ccbfbdb288990c01f185dec79a1022a68b4748
2023-12-17usb: Deal with truncated ISO IN transfersHarald Welte3-0/+29
It seems that in some circumstances, an ISO IN transfer can be truncated by the bus / host. In such situation we'd currently pass a non-modulo-32 length to the mux_demux (deframer) code, and it ASSERTs on that. Let's try to handle this more gracefully by substituting random garbage and letting higher layers deal with massive bit errors. Related: OS#5490 Change-Id: Ic453325b93b0e12727625a1495a948d96df4b542
2023-12-17usb: Pass not just the actual transfer length but also the buffer sizeHarald Welte1-7/+9
Change-Id: I5088a5213d2945fdfaf3b242a63e35d0ed511914
2023-12-03[cosmetic] e1oip.h: Update TODO comment (we have RTT, frame loss stats)Harald Welte1-2/+1
Change-Id: I837676cb3905c0113307ebc400974c3d8cf3a3c7
2023-12-01Use uniform log format for default config filesAndreas Eversberg3-5/+14
Related: OS#6272 Change-Id: I4a8cb558816534ac942bc38ff0b178849d610457
2023-09-12Bump version: 0.5.0.11-b7c9-dirty → 0.6.00.6.0Pau Espin Pedrol6-16/+37
Change-Id: I8d5e3ffd928fa65bc3d3b43553fcd6774a7812dc
2023-08-14octoi: Add force-send-all-ts modeHarald Welte8-10/+45
This new mode (can be enabled per account) will force the E1OIP protocol to always send all timeslots, i.e. not do any of the suppression of timeslots that do not exhibit any change to the previous E1 frame. Change-Id: I6d17d3829b2c1c62e701a1d8c021d93d93593613
2023-07-18e1d-ts-pipe: Assert that osmo_fd_register() succeedsHarald Welte1-1/+1
Fixes: CID#307497: Unchecked return value (CHECKED_RETURN) Change-Id: Ie2abd580e37d7d58c507d12105ec2130fb533ed9
2023-07-18octoi: Fix missing check of osmo_fsm_register() return valueHarald Welte1-2/+2
Fixes: CID#307511: Unchecked return value (CHECKED_RETURN) Change-Id: Ia474a6c078f1aec3d04be494a888ffe87b199f5a
2023-07-18proto_srv: Fix file descriptor and memory leak in error pathsHarald Welte1-0/+3
Fixes: CID#307517 Change-Id: I44e349eee5b0b54843762db8cadd6b9b1758faa4
2023-07-18Check return code of fcntl() in several plaacesHarald Welte2-3/+23
Fixes: CID#307539, CID#307533 Change-Id: I46843174eb4699a59421dc3f3b900a3894c67081
2023-05-26systemd: depend on networking-online.targetOliver Smith1-0/+2
Related: SYS#6400 Change-Id: I29e547242b2ed1cfc4750c7d7e5f8636c2e8f3dc
2023-05-02proto_clnt: use osmo_fd_close to close the socket on connection lossPhilipp Maier1-2/+1
When the socket is closed due to a connection loss, we currently use close(). This is not enough since this will not remove the file descriptor from the select loop. Let's use osmo_fd_close. Related: OS#5983 Change-Id: I702b944baf2ebbcc84b6a211e245a4a41627bde6
2023-04-25debian: set compat level to 10Oliver Smith2-3/+3
Related: OS#5958 Change-Id: Id0f168a7d3c2ae6869121397e65ca1d0cfea30d3
2023-04-20proto_clnt: close osmo-e1d control socket on connection lossPhilipp Maier1-1/+5
When osmo-e1d is terminated the socket file descriptor on the client side will get permanent POLLHUP events. This means that the registered callback gets called with flags OSMO_FD_READ but the received data will be of length zero. We must detect this situations and close the file descriptor on connection loss. Otherwise we would get called over and over again in an endless loop, resulting in 100% CPU usage. Related: OS#5983 Change-Id: I3e1a29a9701a9432f58ef7cfedc32c916203017a
2023-04-20proto_clnt: fix function prototypePhilipp Maier1-1/+1
The function prototype for osmo_e1dp_client_destroy has a different parameter name in its signature in proto_clnt.h than in prto_clnt.c, let's rename it so that both are coherent. (srv -> clnt) Change-Id: I8bd4fbdf2bda332870da1b915a7898c396a85b0f
2023-02-20Run struct_endianness.pyOliver Smith1-0/+7
Ensure there is no diff to prepare to run this in CI. Related: OS#5884 Change-Id: If070a3204d47f71d47e01276fe16323f8af8cb4a
2023-02-07Bump version: 0.4.0.26-8aba-dirty → 0.5.00.5.0Pau Espin Pedrol7-18/+58
Change-Id: I178d34a67cba0359b16fe27fa30b2145703a2909
2023-02-01DAHDI trunkdev supportHarald Welte15-15/+490
DAHDI trunkdev is a newly-introduced 'virtual trunk' character device which is used instead of a real hardware driver. This means that an application (such as osmo-e1d) can implement a virtual E1 trunk and receive and transmit E1 frame data which is exposed to DAHDI users just like the data from a real physical E1 span. In order to build DAHDI trunkdev support into osmo-e1d, you will need a special fork of dahdi containing the required support, currently the laforge/trunkdev branch of the following repository: https://gitea.osmocom.org/retronetworking/dahdi-linux Change-Id: Ib15a7313fcd63e1ed9f2f5b349df967bc4335ec2
2023-01-26Increasing max FIFO size for feature request 5875, adapting testsChristoph Lauter3-117/+130
Change-Id: Iae76d87cf436c80c005a2f8fce5a30f74f7e4545
2022-12-19osmo-e1d: Transition to use of 'telnet_init_default'arehbein2-2/+2
Related: OS#5809 Change-Id: I68fafc519d7c7ca9788bfcad7be1fadbb208cebe
2022-12-08contrib/spec: add /etc/osmocom/osmo-e1d.cfgOliver Smith1-1/+2
Fix that the file only got installed to /usr/share as example, but not to /etc/osmocom in the rpm packaging. Related: OS#5817 Change-Id: I4ffa23f7ab26b7ed3cba04aed4f57eeaa3edca31
2022-12-08doc/examples: only install osmo-e1d.cfg to /etcOliver Smith1-3/+0
Adjust OSMOCONF_FILES, so only osmo-e1d.cfg of the examples gets installed to /etc/osmocom/. All other examples still get installed to /usr/share, as it is intended. Related: OS#5817 Change-Id: Ic449b7d38ed50add0164f056574d4da47530eb49
2022-11-10update .gitignore with files from manuals + doxygenHarald Welte1-0/+10
Change-Id: I952574e4b29db69ca767138b77b0aac4dbfd8973
2022-11-07doxygen API documentation for libosmo-e1dHarald Welte7-58/+1911
If we actually expect 3rd party applications to use libosmo-e1d to talk to osmo-e1d, we'd better add some basic documentation on how this API shall be used. Change-Id: Ib4a97045bca276fbd3892f801898a436de7dc39b
2022-11-07jenkins.sh: Support for building + publishing the manualHarald Welte1-2/+12
Change-Id: Ic399662e6ac5e5e7f999e6e7127df9a9580a0627
2022-11-07Initial user manual for osmo-e1dHarald Welte10-0/+488
Still quite incomplete, but significantly better than nothing. Change-Id: I42f8da1990092b5a3d8c63fde33e49978ad83281
2022-11-07gitreview: new fileOliver Smith1-0/+3
Change-Id: I32ef094163ad157c6174730fa4f0b51038280107
2022-11-07Add support of RPM package building via osmo-e1d.spec.inHarald Welte3-0/+141
Change-Id: If9d3d4b7b06aef156fd895415db3b30bdecd74b4
2022-11-05avoid segfault in 'show interface' of absent icE1usbHarald Welte1-0/+6
Don't segfault when processing a "show interfaces" for an interface that is configured via VTY but not actually attached/found in the system: usb.c:693:29: runtime error: member access within null pointer of type 'struct e1_usb_intf_data' Depends: https://gerrit.osmocom.org/c/osmo-gsm-manuals/+/30019 Depends: https://gerrit.osmocom.org/c/osmo-gsm-manuals/+/30020 Change-Id: I2ce990bb57f6ae4edb3a99a4b7bf26a49f362410
2022-11-03osmo-e1d: Add the usual '-V' / '--version' command line argumentHarald Welte1-1/+7
Change-Id: Ie2d79c44e580ffa6555f87b969206047ffb7eb6f
2022-11-02Differentiate between icE1usb and e1-tracer in log/vty outputHarald Welte2-5/+9
Now that we support the e1-tracer in addition to icE1usb, let's differentiate in the log output. Change-Id: Iea647b12d86c97cfacbc3df43010b9b94f0a97c7
2022-11-01usb.c: Bail out of e1-tracer doesn't accept set_configuration(2)Harald Welte1-2/+7
This likely means it's not an e1-tracer after all, or it's an old firmware that doesn't yet expose the e1d-compatible USB configuration. Related: OS#5734 Change-Id: If5a9bc20084d84885d5d97b4f982e94801612d24
2022-11-01osmo-e1d-pipe: Add command line option for --force open of timeslotHarald Welte1-3/+14
This exposes the existing capability of force-opening a timeslot via a command-line argument. Related: OS#5735 Change-Id: Ieefc89f2e48e9124ae744a587739ff3948110944
2022-11-01ctl: Prevent clients from opening TS0Harald Welte1-1/+1
This doesn't work, as the mux_demux.c code doesn't pass the TS0 bitstream to users anyway. So let's reject clients attempting this. Change-Id: Idb2d20da7de72dad38ae2fccdd7630677d0f0cc8
2022-11-01support for e1-tracer hardwareHarald Welte1-13/+45
Recent work on the e1-tracer firmware is introducing a set of USB descriptors (as configuration 2) that are mostly compatible to how osmo-e1d talks to icE1usb. The main difference is that there's only one ISO IN endpoint per USB interface, and no ISO OUT or ISU FB endpoints. So we introduce some minor adjustments here to accommodate that. Related: OS#5733 Closes: OS#5734 Change-Id: I855e18c0f229bd473123f96303e60ab2de90677f
generated by cgit v1.2.3 (git 2.25.1) at 2024-06-23 17:59:22 +0000