# Commit summaries (start not end) from release-0.28-default.lock.xml vs release-0.26-default.lock.xml ## asyncapi-spec-obj --------------------------- Commit: 9a969e5 --------------------------- ValidationError: Prevent error generator exhaustion This changeset fixes a bug which caused some ValidationError exceptions to be raised containing no validation errors. This happened because the generator of error messages being passed to the ValidationError constructor was being exhausted before the exception message could be created. This change prevents that from happening by storing a list of validation error messages in the ValidationError instance instead of the original generator object, allowing the list of messages to be iterated over multiple times. --------------------------- Commit: f451e08 --------------------------- .gitignore: ignore build and reports folders ## cfg-mgr --------------------------- Commit: a16b400 --------------------------- CSNMR-4816: Increase timeout value for entitlement checks due to increased traffic --------------------------- Commit: 2ba6e90 --------------------------- CSNMR-4816: Add startup check to cache role entitlement * This enables async updates to the role entitlements and prevents thread blocking caused timeout errors on entitlement checks --------------------------- Commit: caf2f23 --------------------------- CSNMR-4816: Allow license timeouts to trigger activation and expiration timeouts --------------------------- Commit: 66d169e --------------------------- CSNMR-4816: Remove hash check for license flag * This change makes it easier for users to add opt out flags via serial, as they do not need a specific file to trigger the flag, any file in the expected location will trigger the opt out --------------------------- Commit: 51b8ab3 --------------------------- CSNMR-4963: Add a timestamp to all outgoing event messages All agents that rely on the BaseCfgEvent and BaseEventEmitter classes will now get a timestamp field in the outgoing messages. If the agent already includes a timestamp field in the payload's top level, then that will be the one that is kept in the final published message since it is closer to the actual event. The asyncapi base_event_msg spec defines the timestamp as optional, so the transition to adding this timestamp will be seamless. Applications will work for older versions that don't have a timestamp, and this version will begin producing it. Unit tests are updated. ## judo-api-specs --------------------------- Commit: 8860f14 --------------------------- CSNMR-5490: add first pass of API conventions docs Added a note on conventions and link to syntax conventions doc. --------------------------- Commit: a936f19 --------------------------- CSNMR-5490: add first pass cfg/event API template Added first pass template-cfg-event.yml file. Using this file as a base for new APIs will help keep API additions in line with the syntax conventions. --------------------------- Commit: 7e91a20 --------------------------- fixup: remove unused tag from asyncapi.yml --------------------------- Commit: 6d6fcab --------------------------- CSNMR-5498: fix statusCode and casing in gnss_reset messages --------------------------- Commit: 7018613 --------------------------- CSNMR-5567: remove x-channel-name from asyncapi spec Removed x-channel-name extensions from all asyncapi definitions. This extension is deprecated and now unused. --------------------------- Commit: fba598f --------------------------- CSNMR-5302: Add signalBars to the wifi diagnostic spec In addition to the rssi signal strength, this adds signalBars for ease of displaying signal strength. The version is updated to 0.64.0. --------------------------- Commit: 80a7854 --------------------------- CSNMR-5523: Add minLength and maxLength to ssid field of 80211 spec The 80211 radios may have a SSID field 0 to 32 octets (bytes), so this change helps document the allowable size. --------------------------- Commit: ee6df44 --------------------------- CSNMR-5534: Change name of license_source to source to match spec conventions --------------------------- Commit: c920c83 --------------------------- CSNMR-5534: Add license source field to license status update --------------------------- Commit: f395c2b --------------------------- CSNMR-5363: Add system time message to the asyncapi spec A diagnostic system time event is added with the path {groupId}/event/diag/sys/time. This message uses the base_event_msg properties and the timestamp comes for free. There are no additional properties. The message will be produced once per minute. The version is updated to 0.62.0. --------------------------- Commit: 96f7528 --------------------------- CSNMR-4963: Add timestamp spec for all outgoing event messages By adding timestamp to the base_event_msg spec, all outgoing event messages may now have a timestamp with a defined format. It is optional so that the spec is compatible with older versions of the spec and code implementations. A corresponding change is expected to be made in the cfg-mgr base classes to automatically add a timestamp. Although agents can optionally add this timestamp to get a slightly more accurate time for when the event took place. The timestamp has date-time format (ISO 8601) like produced by the Python datetime.now().isoformat() function. --------------------------- Commit: 35dc80c --------------------------- CSNMR-5424: add soft_reset level to the sys cmd api Added soft_reset to the system command enum. This reset is not deleting data so was not included in the reset to defaults api enum. CTCT needs this to refresh the config agents without the overhead of a full reboot. --------------------------- Commit: 1625193 --------------------------- CSNMR-5424: add reply event for system commands topic Added event_device_sys_cmd channel so that system commands have a reply mechanism. --------------------------- Commit: 6cd1670 --------------------------- CSNMR-5427: add gnss reset API definition Added ctrl_gnss_reset and event_gnss_reset channels to the gnss manager APIs. These were asked for by CTCT to allow respins of the Titan engine when corrections sources change, and other levels of reset for troubleshooting. --------------------------- Commit: 7c08883 --------------------------- CSNMR-5424: move system commands into control.yml Moved the system command api into sys/control. Any control topics in sys/sys.yml are eventually destined for sys/control for smaller chunks of yaml to work on. --------------------------- Commit: b445cbe --------------------------- Add initial CODEOWNERS file Signed-off-by: Theodore A. Roth --------------------------- Commit: c5eab46 --------------------------- CSNMR-5362: Add temperature diagnostic API This temperature diagnostic API is intended to provide current temperatures from the /sys/devices/virtual/thermal folder. The message includes the sensor names and the temperature values in celsius as key-value pairs. It will be published on startup and at regular intervals. ## judo-auto-test --------------------------- Commit: 61d072d --------------------------- CSNMR-5622 Improve canbus tests and fix failures - Unskip test for CSNMR-5621: fix root cause of failure. A single shared buffer holds incoming MQTT messages; each new subscribe replaced it, so tests sometimes read from the previous subscribe’s (empty) buffer. Fix by aligning subscribe/listen usage so the correct buffer is used. - Improve canbus test flow and robustness. --------------------------- Commit: a4c4c31 --------------------------- CSNMR-5621 Skip canbus test --------------------------- Commit: ec85a83 --------------------------- CSNMR-5590 Remove duplicate ping tests - Multiple ping tests were imlemented in RF. eth_4we_1, eth_2we_1, eth_2we_2, and eth_bd992. Remove these tests from Slash as they are no longer needed there. --------------------------- Commit: 13445d0 --------------------------- CSNMR-5585 Unskip Reboot Initiates License Refresh RF testcase - CSNMR-5534 was implemented so unskip the reboot. test. After the reboot completes and the License Process State reaches complete. There is an additional check to verify that the source of the restored license is from the cloud. --------------------------- Commit: 8310ed5 --------------------------- CSNMR-5578 Fix can_bus.robot test - The CI fails the mp1010_mqtt_api/canbus.robot tests. Root cause is the enable parameter for the publish message is set to a RF ${True}/${False} and not true/false as required by in the payload. The fix is to set the values correctly --------------------------- Commit: 9fa319f --------------------------- CSNMR-5395 Add CANBus robot test CSNMR-5395 Version of canbus robot test CSNMR-5395-CANBus Robot test --------------------------- Commit: 19916a6 --------------------------- CSNMR-5393 - Implement CAN bus configuration handling keywords - Add keywords to enable/didable can interface and get can interface setting using MQTT --------------------------- Commit: 72f9b67 --------------------------- CSNMR-4816 Skip licensing test for CSNMR-5534 - Skipped the licensing->Reboot Initiates License Refresh test due to tests inablilty to detect if the license was updated via cloud. --------------------------- Commit: 3c92cef --------------------------- CSNMR-4816 Fix test Reboot Initiates License Refresh - A bug was introduced with CSNMR-4816 updates that caused a race condition in the test where after reboot the payload used to validate the entitlements could be stale and cause a false result. New keywords where added to the library to wait for certain conditions are met before doing the validation --------------------------- Commit: 848d8b7 --------------------------- CSNMR-4816 Add new keywords - Added 3 new keywords: - is_license_process_state - get_diag_license_process_state - wait_for_license_process_state --------------------------- Commit: 9b10ab0 --------------------------- CSNMR-5515 Remove depricated keywords - Removed the depricated keywords from the judo_mqttlibrary --------------------------- Commit: 03ba649 --------------------------- CSNMR-5515 Fix lint problems in RF tests - Fixed multiple lint problems in the RF test suites. --------------------------- Commit: b5097e5 --------------------------- CSNMR-5493 Add Install New Licence test cases - New test case that test the ability to obtain licenses from the license server if no licenses are currently applied to the device. --------------------------- Commit: 4183750 --------------------------- CSNMR-5493 Add Delete All Licenses keyword - Added a keyword called Delete All Licenses which publishes a remove_licenses topic to the Judo device --------------------------- Commit: 36b40c4 --------------------------- Fixed __init__.py SystemControlLib misspelling The spelling for SystemContrlLib in __init__.py did not match the class name trying to import the library as systemcntl_lib.SystemControlLib --------------------------- Commit: 5587434 --------------------------- CSNMR-5331 Add GNSS lease keep-alive to test cases - Added GNSS keep-alive to test cases to keep the GNSS lease from expiring prematurely before the HF GSOF check can be completed --------------------------- Commit: 38b2413 --------------------------- Skip GNSS UHF radio test cases - Skipping the test_feed_correction_data_from_uhf_to_gnss tests are failing Error: AssertionError: GNSS receiver fix mode is not RTKFixed after 60 seconds. See CSNMR-4669 --------------------------- Commit: d1bcf48 --------------------------- CSNMR-5068: Remove issue from tests/test_wifi.py:test_default_wifi_configuration skip marker This changeset removes CSNMR-5068 from the list of issues causing this test to be skipped. Note that CSNMR-5069 has not been completed, so this test will not run yet. --------------------------- Commit: ef07358 --------------------------- CSNMR-4238: Remove stray dependence on separate uut_mqtt_id fixture --------------------------- Commit: d0a908e --------------------------- CSNMR-4669: Remain in RTKFixed mode for non-zero time Removing the UHF correction stream from the GNSS receiver as soon as it reports RTKFixed position mode causes the receiver to get stuck in that mode even when the correction stream is removed. Having the test maintain the correction stream to the receiver for 20 seconds after it enters RTKFixed position mode fixes this behaviour. --------------------------- Commit: ad16040 --------------------------- CSNMR-4669: Ensure UUT is not in RTKFixed mode before test If this not done, the UHF radio could be enabled before checking that the BD992 is not in RTKFixed mode, making the test invalid (as the BD992 is not switching to RTKFixed mode as a result of receiving valid correction data). Ensuring that the UUT is not in RTKFixed mode is more correct than ensuring that it is in GNSS mode, as the UUT is not guaranteed to return to GNSS mode when the UHF correction data stream is lost, depending on how it is configured and which other correction sources are available. --------------------------- Commit: 3a62040 --------------------------- CSNMR-4669: Add useful error message to test_feed_correction_data_from_uhf_to_gnss --------------------------- Commit: 92f95f0 --------------------------- CSNMR-4669: Reinstate skipped test --------------------------- Commit: c31e770 --------------------------- CSNMR-5324: Reinstate skipped test --------------------------- Commit: b036731 --------------------------- CSNMR-5479 Increase timeout for SUB Event Diag Sys Resource test - Fix the intermittent timeout that occurs when sending 'event/diag/sys/resources' toping and waiting for a response. The was increased from 30s to 45s --------------------------- Commit: 06f9f1d --------------------------- CSNMR-5434 Refactor judmwttlibrary featurelibs for new keywords - Refactor all of the feature libs to use the new refactored single word keywords --------------------------- Commit: ec26b88 --------------------------- CSNMR-5434 Refactor judomqttlibrary single word keywords - Refactored single word keywords in mqttcommand.py and abstract_device_lib.py to be multi-word and more descriptive of what the keyword actually does. The keywords affected were connect(), disconnect(), publish(), and subscribe(). Additionally this resolves issues with linter warning that Robot Framework keywords are not to spec. - Retained origial single work keywords for backward compatibility and marked as depricated. Then from these each old keyword calls the new improved keyword name - Updated incorrect keyword examples --------------------------- Commit: 1f786bb --------------------------- CSNMR-5444 Add repeating pub gnss lease keyword - Added keywords gnss.resource file which contains keywords to support GNSS operations - Added Repeat Publish GNSS Lease keyword. This keyword basically allows GNSS Lease keep-alive by sending a gnss lease publish for some number times with a wait period between each send. --------------------------- Commit: aae40d2 --------------------------- CSNMR-5420: Switch WCO roof 450MHz UHF base station This changeset changes the base station used for 450MHz UHF reception tests from 464.55MHz to 469.55MHz, as the base station on 469.55MHz has a more stable configuration. --------------------------- Commit: a84c502 --------------------------- test_uhf.py: Remove test_correction_data_available_externally case This test case causes problems because its criteria are more strict than required for a listening GNSS receiver to maintain an RTK Fixed position, and it is more difficult to implement the criteria used by a GNSS receiver in this test than it is to simply _use_ a GNSS receiver to decide if the correction stream is acceptable or not. This sort of testing is already done by test_gnss.py:test_feed_correction_data_from_uhf_to_gnss, so there is no need for this test case to exist any more. ## judo-bd992-fw --------------------------- Commit: e86f0c1 --------------------------- CSNMR-5600: update BD images for stella Update BD images for stella to 6.60-0.27 Stella Checksum = 0x3EFE8B8D --------------------------- Commit: fc7599b --------------------------- CSNMR-5214: update BD images for stella Update BD images for stella to 6.60-0.24 Stella Checksum = 0x179D32E4 ## judo-cfg-agents --------------------------- Commit: b14dfc7 --------------------------- CSNMR-5428: Add support for GNSS Reset GNSS supports the following GNSS reset levels: * PVT: The PVT Engine (RTK/RTX) is restarted without restarting tracking or rebooting the GNSS subsystem. * SV_DATA: Clears stored ephemeris and almanac data and reboots the GNSS subsystem. * FULL: Restores the receiver to factory state. This will clear Ephemeris and Almanac data, clear all logged data, reformat storage, restore factory settings and reboot the GNSS subsystem. For each of the reset levels, the following steps are taken: 1. GNSS Manager Agent stops bd992-handler 2. Perform BD reset a. bd992-handler reset [serial_port] [reset_level] b. Where reset_level is PVT, SV_DATA, or FULL c. Wait for reset to complete 3. GNSS Manager Agent starts bd992-handler --------------------------- Commit: 1eea9e8 --------------------------- CSNMR-5564: Change firmware_ota_agent to check internet connectivity This changes the agent to only access the internet for identity certificates and firmware updates when there is internet connectivity. The reason for this is the cloud access functions can take a long time and might interfere with local firmware updates. Once internet is available, those cloud checks are usually fast and take a few seconds. In agent.py: It adds a new class to receive internet diagnostic messages to notify the manager when internet is available. The call to device_update_identity_cert() is removed where the environment is updated. Instead, the manager gets the cert when it needs to check for firmware so that both operations are controlled by the same logic. The class SetEnvCfgEvent is instatiated after the FirmwareOtaUpdateManager so that internal settings do not get overridden with defaults in the manager. In manager.py The function setDeviceParams is broken into two parts. It still sets the environment _device_params for stage or prod. But the cloud_params are extracted in the new _extract_base_url() after the certificate is downloaded. Finally, the check_update has two new additions. 1. It tests if the _internet_connected flag is set. If not it schedules another try for later. 2. It downloads the identity certificate if it is the first try or if the environment has changed. If that download fails, it schedules another try. --------------------------- Commit: cee59f6 --------------------------- CSNMR-5564: Fix local firmware update to not check for certificates This is the main fix for the local updates that were timing out because the call to update_env() was taking a long time. Now it does not check for a certificate when stop_cloud_update and install_update commands are sent. --------------------------- Commit: 837fdda --------------------------- CSNMR-5564: Check for firmware update after broker connection This change moves the initial firmware check to after the agent is connected to zenoh to enable it to check internet status. It also has the benefit of making the initial startup faster. --------------------------- Commit: a9b04bd --------------------------- CSNMR-5426: Add soft_reset to system_control_agent Add soft reset to system_control_agent using systemd stop and start commands from a soft_reset.sh script. system_control_agent sends the SUCCESS_COMMAND_EXECUTED event notification on accepted soft_reset command start. --------------------------- Commit: 6a8c581 --------------------------- CSNMR-5214: Add support for Stella firmware update Check for Stella or BD992 to determine static GNSS configuration file. Generate GNSS configuration adjusting the Stella firmware image and firmware checksum on a unit with Stella. Add GNSS Mgr check for cfg change. To avoid unnecessary GNSS restarts, GNSS configuration does the following: 1. Update and merge GNSS configurations 2. get current GNSS configuration 3. compare merged and current configuration 4. restart GNSS handlers if configuration has changed. --------------------------- Commit: 77ab641 --------------------------- CSNMR-5572: add MP1087 model support to system properties Added a duplication of the MP1086 component sets for MP1087. MP1087 hardware is essentially the same as MP1086, we just need to accept both models. Potential refactor of variant and suffix is a later question for Sumo development. --------------------------- Commit: dd1c520 --------------------------- CSNMR-4051: Add 80211/Wifi/Halow diagnostics info implementation Diagnostic code is implemented in the ieee80211_agent to produce the event_diag_80211_info message. The reason it is in the agent and not in the diags_agent is to keep this code close to the agent it monitors. The diagnostic class could be moved to the diags agent easily if needed. Messages are published for wifi and halow radios. At the moment, the halow message is mostly blank unless the halow radio gets turned on with "sudo modprobe nrc". This diagnostic code does a lot of command line queries to gather information with parsing steps. The unit tests cover all the conditionals so there are a lot of tests. --------------------------- Commit: c4357b0 --------------------------- CSNMR-5462: extend cell service stops in MFG API Added a restart of the ModemManager service just after the MFG API modem AT command handler takes down a bunch of cell services. Added celldiags_agent and NetworkManager to the stop list. In the last OS release to the CM the first issued modem AT command was failing 40 % of the time. Stopping celldiags and Network manager reduced the occurences. Restarting ModemManager after killing cell services fully resolved the issue. Serial port conflicts and half finishing ModemManager interactions were likely the cause. --------------------------- Commit: ec6e9cd --------------------------- CSNMR-5002: Add Logging to GNSS Mgr Add GNSS Logging support to GNSS Manager Agent to control BD Logging with "schedule" support of Disabled, Always, and Continuous. Add runtime generation of GNSS configuration files. The generated GNSS configuration file (.toml) will be located in '/cfg/sysconfig'. The generated configuration is a merge of the static base GNSS configuration files in '/etc/gnss' and the GNSS logging runtime configuration parameters The static GNSS base files are selected as before from the GNSS machine type cfg and GNSS lease cfg. The GNSS runtime cfg is generated from the GNSS logging cfg. The runtime configuration "wins" over the static GNSS cfg files if there are any commen keys. --------------------------- Commit: af5bc45 --------------------------- CSNMR-5534: Add tracked source of refresh request to license process state message --------------------------- Commit: 969efb1 --------------------------- CSNMR-4816: Check access token on connect before confirming remote connection --------------------------- Commit: 9401cae --------------------------- CSNMR-4816: Fix UHF agent licensing issue * UHF agent was not checking license validation on startup, which caused the check to not be properly cached * This caused subsequent misses checking UHF entitlement status --------------------------- Commit: b88e219 --------------------------- CSNMR-4816: Increase the license forced refresh period for internal testing purposes --------------------------- Commit: 5f1fd74 --------------------------- CSNMR-4816: Enable stage env by default for licensing * This is a temporary change to make the early phase of testing easier until the prod lms env is ready for use * This removes the requirement that users manually set the environment --------------------------- Commit: 5e5feff --------------------------- CSNMR-5478: Remove required id field from license xml schema --------------------------- Commit: ef63e7d --------------------------- CSNMR-5363: Add system time message emitter to the diags_agent This change produces the asyncapi message for event_diag_sys_time. It is a simple message with just a timestamp field and deviceId. The code publishes an empty dict and the base class adds the required fields. A unit test is included. --------------------------- Commit: a4d141a --------------------------- CSNMR-4963: Update unit tests to handle the new timestamp field Cfg-mgr now adds a timestamp field to all messages published with the BaseEventEmitter and BaseCfgEvent classes. The unit tests failed when comparing expected output with actual output payloads. This change is to fix the unit tests to drop the timestamp before doing the comparisons. --------------------------- Commit: fb0cbb4 --------------------------- CSNMR-4963: Update reference to cfg-mgr The cfg-mgr libraries are updated and this changes pyproject.toml to reference the updated hash. The cfg-mgr base BaseEventEmitter class now adds a timestamp to outgoing event messages. --------------------------- Commit: caf9ffb --------------------------- CSNMR-5407: Fix the snapshot tar to keep going on errors A case was found where the tar.add function could not open a listed file and cancelled the snapshot. Since we want to get the snapshot even if a file could not be read, it now continues after logging a warning with an error message. Unit tests are updated to test the new exception handler. Some redundant unit tests are removed. --------------------------- Commit: 1318ee2 --------------------------- CSNMR-5362: Add temperature reporting to the diags_agent This implements the api message for event_diag_temperatures by adding temperatures.py to the diags_agent. It reads the temperatures from /sys/devices/vertual/thermal/thermal_zone* folders. It publishes the message every 5 minutes or whenever a change greater than 2 deg C is detected at the regular check interval of 30 sec. The file diags_agent/agent.py is updated to import the new class and run it. Unit tests are added. ## judo-design-docs --------------------------- Commit: b4b328d --------------------------- CSNMR-5428: Update GNSS Manager Docs for GNSS Reset Update GNSS Manager design documentation for GNSS Reset. --------------------------- Commit: 5b71ecd --------------------------- CSNMR-5214 Update GNSS Design Documentation Updated the GNSS Manager design documentation to support dynamic runtime parameter changes. The system now generates a unified configuration by merging static base files, runtime-generated logging settings, and optional manual overrides using a "last-one-wins" priority hierarchy for matching keys. ## judo-radio-utils --------------------------- Commit: e9f54c3 --------------------------- CSNMR-5628: Remove functional test start and stop scripts The setup and teardown actions performed by these scripts are now performed by the MQTT test console. --------------------------- Commit: 3046c99 --------------------------- CSNMR-5628: Add Phoenix functional test service This changeset adds a service which runs the Phoenix handler, receiving data on the frequency used in the functional test and publishing that data on the port used for observable data output, so that the data can pass through the firewall and be received by the test monitor/controller. --------------------------- Commit: 703f44e --------------------------- CSNMR-5628: run-phoenix-suhf.py: Allow configuring output address The firewall allows outgoing traffic on port 12101 for observables data from the UHF radio. This changeset allows the output address to be configured for this port so that observable data can be received by the functional test controller, while retaining the default behaviour of sending observable data on 0.0.0.0:1234. --------------------------- Commit: d43e789 --------------------------- CSNMR-5628: Add send-observables systemd service Making send-observables a systemd service avoids the complexity of managing the process directly in tests. --------------------------- Commit: afc66c6 --------------------------- SUMO-553: Add read-5v-psu-amps.py and run-phoenix-suhf.py helpers With some minor modifications, these scripts have been copied from the attachments to SUMO-553. --------------------------- Commit: a8fcdd0 --------------------------- CSNMR-5538: Fix HaLow subnet The default HaLow subnet is 192.168.44.0/24, not 192.168.200.0/24. --------------------------- Commit: 4a096b7 --------------------------- CSNMR-5538: Remove HaLow AP/STA setup scripts Setting up the HaLow AP and STA can now be done using the MQTT API. --------------------------- Commit: 460c5cb --------------------------- CSNMR-4649 Functional Test Updates for Swift Modify halow-test-us-*-open.sh to add arguments for not running wpa_supplicant or hostapd. Create a systemd service file for managing these services. ## judo-rest-api --------------------------- Commit: 645a0b6 --------------------------- CSNMR-5448: Add file patch REST service This new patch service runs as a separate process from the Judo-rest-api. The reason is so it can be started and stopped on its own. Later, a license check may be done to enable the service. The new patch service takes a file and attempts to uncompress it. It returns error messages and logs activity to a file in /cfg/patch/mp_patch_history.log. Unit tests are added for basic functionality. ## judo-security --------------------------- Commit: 4488171 --------------------------- Replace jabil references with mfg and/or manufacturing ## judo-web-ui --------------------------- Commit: 378e08f --------------------------- Hide the API Debug section with a toggle button on the v0 page The Judo API Debug section is now hidden because it is rarely used. It can be shown with a toggle button. It is still useful for testing communication with the back end. --------------------------- Commit: a3eb190 --------------------------- CSNMR-5448: Add a File Patch Service section to the v0 web UI A new section is added to the v0 web UI to allow a patch zip file to be sent to the device. It shows a panel with patch log file contents. A test folder is created with a simple test zip file that can be used to test the service. --------------------------- Commit: c9af362 --------------------------- CSNMR-5002: Add v0 web UI dialog for gnss logging configuration The v0 web page has index.html and judo.js updated and gnss_log.js added to include a form for gnss logging configuration. It sends a message to the Judo that will be picked up the the gnss_manager agent. The agent will pass the information to the bd922-handler to configure the BD GNSS receiver logging settings. ## manifest-judo --------------------------- Commit: 026dc59 --------------------------- CSNMR-5428: Freeze manifest for release 0.28 This changeset freeze the manifest using the layer hashes from https://bamboo.trimble.tools/browse/CC-JUDO899-2/artifact --------------------------- Commit: 22e57f3 --------------------------- CSNMR-5483: Setup for release 0.28 --------------------------- Commit: d34f237 --------------------------- SUMO-552 commit message summaries for builds All of these scripts are designed for stand-alone use. They are **not** part of the build process (at least not yet). All of these scripts scrape the git history of all the Judo repositories, to list commits used for a build. The repositories must already checked out or updated (using the Android repo tool). - buildinfo_summaries.py uses the `/etc/os-release` and `etc/buildinfo` from an live Judo/Timo/Sumo, to get the messages. This script will grab the necessary files from the DUT, via ssh. - xml_lock_summaries.py uses a default.lock.xml from the Bamboo artifacts, to get the messages. The default.lock.xml must be manually downloaded. - xml_lock_summary_diff.py uses two default.lock.xml from the Bamboo artifacts, to get the messages for all commits called out by hashes in the two default.lock.xml files. The default.lock.xml files must be manually downloaded. - scripts/bamboo_fw_download.py: download Bamboo firmware artifacts (.raucb, optional .wic.xz) via REST API and PAT; supports multiple plan keys (Judo-HS-Prod, RPi, Sumo), label expansion from HTML browse pages, pre-signed S3 URLs, --force, -v/--verbose. - scripts/*.md: requirements for the script (purpose, inputs, core behavior, CLI, constants, dependencies, etc). Note: All the scripts in this commit were created with AI, which is why the scripts/*.md files have been included, even though they are not directly used. --------------------------- Commit: b4b9ea3 --------------------------- CSNMR-5203: Propagate build errors up to bamboo We are not properly propagating yocto build errors up to bamboo from the docker containers. As such, bamboo always thinks the build succeeded and then tried to run the regression tests when it should not and the tests fail. This is confussing since it looks like the tests failed, when that is not actually where the failure was. Signed-off-by: Theodore A. Roth --------------------------- Commit: 12b5296 --------------------------- CSNMR-5413: OS snapshot capture after test jobs Add pull-dut-os-snapshot.py to capture the DUT OS snapshot via MQTT after each test run. The script subscribes to system_properties for the device ID, triggers a snapshot, waits for completion, and downloads the resulting archive to the Bamboo workspace. Each test wrapper (firmware, radio/GNSS, cell) now calls the snapshot script inside the TEST_CORE_IMAGE block. The snapshot exit code is ignored; a required Bamboo artifact enforces that the snapshot was produced, failing the job at artifact collection if missing. Signed-off-by: Brady Hill --------------------------- Commit: 7625499 --------------------------- scripts: Remove judo-build hash server from configuration The hash server on the judo-build server no longer exists, and attempting to contact it generates warning messages in the build log. This change fixes the following warnings in the build output (and other similar warnings in specific recipes): WARNING: Error contacting Hash Equivalence Server judo-build.ap.trimblecorp.net:8686: [Errno -2] Name or service not known --------------------------- Commit: 2d54883 --------------------------- New kernel lock on builds to help with ec2 scaling Signed-off-by: Brady Hill --------------------------- Commit: 4d1783c --------------------------- SUMO-483: Update notes on enabling compliance feature --------------------------- Commit: 0a7c2fc --------------------------- BEDROCK-194 yocto mirrors used in manifest-judo Signed-off-by: Brady Hill --------------------------- Commit: 5fe14e4 --------------------------- Remove sumo logic from bamboo-build There was previous logic that pinned sumo to only build a couple images instead of what was passed into the script arguments. That logic has been removed and it will build anything passed in as an argument Signed-off-by: Brady Hill --------------------------- Commit: cefa5d6 --------------------------- CI: bugfix: artifact Sumo full bundles --------------------------- Commit: 9a2ad42 --------------------------- Switch to github mirrors for upstream The Yocto project repositories are running extremely slowly which is causing numerous build failures. To be reverted if/when the Yocto upstream issues are resolved. --------------------------- Commit: 39b46dc --------------------------- SUMO-539: CI: Artifact OPTEE TA public signing key ## meta-bedrock --------------------------- Commit: dee7577 --------------------------- BEDROCK-191: certmgmt: Update recipe to support 0.0.8 and 0.1.2 Between 0.0.8 and 0.1.1 of device-id-cert-tools, there was a major refactoring of the code base to add support for IAM as the future of profiles and bootstrap. Instead of just updating the recipe from 0.0.8 to 0.1.2, we need to continue to support the 0.0.8 code base as Judo should not consume the refactored code since it has not been tested on devices yet. We need to integrate the new code base into bedrock first to facilitate testing on RPI devices. For now, the default behavior will see builds getting the 0.0.8 code base unless the following is set in `local.conf` or `user.conf`: PREFERRED_VERSION_device-id-cert-tools = "0.1.2%" This will force using the `device-id-cert-tools_git.bb` version of the recipe. We will be doing that for bedrock builds so that we can test the new code base on real devices. Signed-off-by: Theodore A. Roth --------------------------- Commit: 435e3fb --------------------------- BEDROCK-191: certmgmt: Fix adding hsm-init-yubikey.py When setting the FILES variable for the device package, need to append instead of assign in the .bbappend so as not to clobber the value set in the main recipe. This was causing the files defined in the main recipe to not be included in the device sub-package. Signed-off-by: Theodore A. Roth --------------------------- Commit: 77281fb --------------------------- BEDROCK-183: Force setting adminuser password on first login Signed-off-by: Theodore A. Roth --------------------------- Commit: dc50134 --------------------------- BEDROCK-183: openssh: Remove debug-tweaks Upstream has removed `debug-tweaks` as allowed value of `IMAGE_FEATURES`. Signed-off-by: Theodore A. Roth --------------------------- Commit: 7843d41 --------------------------- packagegroup-bedrock-core: Remove openssl-ossl-module-legacy Upstream has disabled legacy-openssl by default, so the need to pull in the openssl-ossl-module-legacy-module for the python3-cryptography package is no longer needed. The upstream changes were breaking the bedrock builds. Signed-off-by: Theodore A. Roth --------------------------- Commit: d25d879 --------------------------- Add systemd-analyze to os images This enables the ability to analyze the boot sequence and timing on a device. Signed-off-by: Theodore A. Roth --------------------------- Commit: f3a0aa0 --------------------------- BEDROCK-195: hostapd: Fix to avoid invalid long SSIDs The hostapd daemon will fail to start if the SSID is set to a value longer than 32 characters. Example failure: Feb 11 13:47:30 bedrock-f60d58706fb34fc3-x1 hostapd[635]: Line 1: invalid SSID 'bedrock-f60d58706fb34fc3-x1-7c7749' Feb 11 13:47:30 bedrock-f60d58706fb34fc3-x1 hostapd[635]: 1 errors found in configuration file '/tmp/hostapd.conf' If the SSID is too long, it will be truncated to 32 characters. Signed-off-by: Theodore A. Roth --------------------------- Commit: 4ce9cb2 --------------------------- crun: Fix QA error which is breaking the builds Quells the following QA error: ERROR: crun-v1.24.0+git-r0 do_package_qa: QA Issue: crun: recipe doesn't inherit features_check [unhandled-features-check] This needs to be submitted to upstream meta-virtualization. Signed-off-by: Theodore A. Roth --------------------------- Commit: 24fb9e3 --------------------------- BEDROCK-190: Use nginx instead of lighttpd The nginx http server supports openssl engines and providers for pkcs#11 storage/access of privates keys for https. This also introduces a basic bedrock-web-ui package to install static web pages for the server to host. The web ui package uses a virtual webserver config dependency to decouple it from a specific web server, so in theory, end users can select which web server to use via setting `PREFERRED_RPROVIDER_virtual-webserver-config`. Signed-off-by: Theodore A. Roth ## meta-judo --------------------------- Commit: 559b07b --------------------------- CSNMR-5483: Set version number for release 0.28 --------------------------- Commit: 41dab7e --------------------------- CSNMR-4696: recovery: Fix DHCP server race on boot When booting into recovery mode, there is a race condition with starting the dhcp server and the mounting of the `/var/lib` path. The `/var/lib` must be read-write before the DHCP server can be started. There are multiple services that rely on `/var/lib` being read/write, so improving the startup service dependencies allows the system to boot up into recovery and provide DHCP services to connected devices. Signed-off-by: Theodore A. Roth --------------------------- Commit: 34694d1 --------------------------- CSNMR-5600: Add GNSS stella firmware update Update stella firmware GNSS override configuration to 6.60-0.27 --------------------------- Commit: 7c99eca --------------------------- CSNMR-5538: Add halow-uhf-rf-path-ctl utility to Sumo image The halow-uhf-rf-path-ctl utility configures the RF path shared by the HaLow and UHF radios on Sumo hardware. --------------------------- Commit: d266da2 --------------------------- CSNMR-5538: Enable load-nrc service on boot Running the load-nrc service to load the HaLow driver on boot seems to produce reliable results, and is simpler than running the service manually. --------------------------- Commit: 98fe292 --------------------------- CSNMR-5588: Update the Judo web UI to version 0.24.10 The recipes are renamed with the new version and the Artifactory checksums are changed to match the new files. --------------------------- Commit: f8eab82 --------------------------- Fix judo-web-ui.conf to not cache v0 pages There can be times when the v0 web UI displays cached versions of files. This change adds the necessary header to v0 responses to avoid caching. It may still be necessary to hit Ctrl-F5 to do a hard reload if an older version was installed that did not have this caching fix. --------------------------- Commit: 150049c --------------------------- CSNMR-5214: Add GNSS auto stella firmware update Add GNSS automatic firmware update for stella. If unit has a stella, then the GNSS Manager will read a stella configuration override file to generate a GNSS configuration with the stella firmware image file and firmware checksum. If there is a stella firmware checksum mismatch, the bd992-handler will automatically update the stella firmware. --------------------------- Commit: 6b5b9a9 --------------------------- CSNMR-5567: update SRCREV judo-api-specs --------------------------- Commit: e4af1da --------------------------- CSNMR-5302: Update SRCREV for judo-api-specs --------------------------- Commit: b195607 --------------------------- CSNMR-5448: Add Judo Patch Service files and update SRCREV This change brings in a service file named judo-patch.service to start the judo-patch service. Even though it is part of the judo-rest-api, the patch service runs independently because later there may be a license check before the service will start. The patch service allows .tar.gz patch files to be installed on the device. The judo-web-ui.conf file modifies lighttpd to send the api call to the patch service instead of the standard judo-rest-api. It also provides an alias for the mp_patch_history.log file so that the web UI can load the log file. The judo-web-ui and pythong-judo-rest-api SRCREV is updated to get the corresponding change for the v0 web UI. --------------------------- Commit: 05328a1 --------------------------- CSNMR-5514: Update Judo web UI to version 0.24.7 The recipes for the web UI and the recovery web UI are renamed with the new version and the Artifictory check sums are updated. --------------------------- Commit: 85bb68d --------------------------- BEDROCK-191: device-id-cert-tools: Set preferred version Judo should not yet use the latest version of device-id-cert-tools (`0.1.2+git`), so we need to explicitly select the older version. Before we switch to the latest version (which we need for IAM support), more testing of it is needed. Signed-off-by: Theodore A. Roth --------------------------- Commit: b1788c7 --------------------------- CSNMR-5523: Update SRCREV for judo-api-specs --------------------------- Commit: 202d645 --------------------------- CSNMR-5002 Move judo-web-ui SRCREV to an include file Both judo-web-ui and judo-web-ui-recovery rely on the same SRCREV, so it is now moved to an include file so that it is updated once. The judo-api-specs SRCREV is also updated. --------------------------- Commit: c191789 --------------------------- CSNMR-5534: Update SRCREV --------------------------- Commit: 1b13fa3 --------------------------- CSNMR-5534: Update SRCREV --------------------------- Commit: 71f969b --------------------------- Bugfix: Use correct nrc7394 halow module FW file The `nrc7394` module fails to initialise with the default FW file, so install/deploy `nrc7394_cspi_eeprom.bin` and update the driver to load that file by default. --------------------------- Commit: 5ae3694 --------------------------- CSNMR-4649: Add openssh-misc package to CI image This package provides the ssh-agent and ssh-add utilities which allow setting up SSH connections to the UUT and secondary unit for tests which are not yet supported by the MQTT API. --------------------------- Commit: bbf9402 --------------------------- CSNMR-4649: Add Sumo test setup RPi to hostname lookup --------------------------- Commit: 9474a6b --------------------------- Bugfix: Store container data in /data partition Was previously trying to store in the user home, but that only worked for non-root users as the root users home is now `/root` and there is no R/W overlay for it. --------------------------- Commit: 1bab4a6 --------------------------- SUMO-483: Add bluetooth CW recipe when compliance feature present Convert compliance feature from a MACHINE feature to a DISTRO feature. --------------------------- Commit: 68d26bb --------------------------- CSNMR-5363: Update SRCREV for judo-api-specs --------------------------- Commit: bc0ece9 --------------------------- CSNMR-5340: Set version of main to 0.27 --------------------------- Commit: ab0f466 --------------------------- SUMO-279: ms9xx-bridge -> rust-gnss refactor Handle the renamed packages in the packagegroups, and only include the command-line tools in -devel images. --------------------------- Commit: 6dcf9b4 --------------------------- CSNMR-4963: Update SRCREV for judo-api-specs --------------------------- Commit: 4f9f09c --------------------------- CSNMR-5424: update SRCREV --------------------------- Commit: 8d37995 --------------------------- CSNMR-5208: put mfg partition write/clear script in prod Put recipe mfgfs-create into the base judo packagegroup so that write-clear-mfg-partition.sh is available in devel and prod builds. The system control reset API depends on this script being present for the manufacturing level reset to work. We want the API to work in prod, so the script is being made available there. --------------------------- Commit: a9e8595 --------------------------- SUMO-481: minicom: Fix setting of serial port baud rates Required for glibc 2.42+ support which Yocto has recently updated to. --------------------------- Commit: fdb51a1 --------------------------- Bugfix: Define device variant for Sumo --------------------------- Commit: 9d23dd3 --------------------------- SUMO-276: Add Bluetooth FW updater recipe to Sumo --------------------------- Commit: ec2d454 --------------------------- CSNMR-5198: Add model suffix to systemd environment Adding the model suffix (500, 510, etc) to the systemd environment allows services to be started based on this value. For example, the enable-uhf service which controls the power supply for the Satel radio should only be enabled on -510 Judo units. --------------------------- Commit: e87054e --------------------------- CSNMR-5362: Update SRCREV for judo-api-specs ## meta-judo-bsp --------------------------- Commit: c31c1c9 --------------------------- CSNMR-5628: judo-radio-utils: Update SRCREV This changeset also adds systemd services for running the secondary Judo unit's Mikimoto UHF transmitter and the UUT's Phoenix UHF receiver. These are used during functional testing of the UUT's Phoenix UHF receiver. --------------------------- Commit: cbd8df9 --------------------------- CSNMR-5628: gpioset-services: Enable Sumo UHF LNA on boot --------------------------- Commit: ea481c9 --------------------------- SUMO-553: judo-radio-utils: update SRCREV This changeset also adds the python3-smbus2 dependency to the judo-radio-utils recipe, as it is required by the read-5v-psu-amps.py script. --------------------------- Commit: 7ae5eac --------------------------- CSNMR-5538: Configure Sumo HaLow/UHF shared RF path on boot This changeset adds a service which configures the shared HaLow/UHF RF path on Sumo when the device boots. This service is a temporary solution to configuring this part of the hardware, and this commit should be reverted once the RF path is configured via the MQTT API. --------------------------- Commit: 67197f2 --------------------------- CSNMR-5538: Add Sumo HaLow/UHF shared RF path config utility --------------------------- Commit: f210b3b --------------------------- CSNMR-5538: judo-radio-utils: update SRCREV This changeset also removes the HaLow AP and STA services used by the functional test, as HaLow is now configured using the MQTT API. --------------------------- Commit: 9e25513 --------------------------- CSNMR-5454: program-uhf-fpga: Update FPGA image --------------------------- Commit: ab82d76 --------------------------- CSNMR-4649: judo-radio-utils: Package systemd service directories The systemd service directories need to be packaged by this recipe because they now contain services for starting the HaLow AP and STA for functional testing. --------------------------- Commit: 18a12f7 --------------------------- CSNMR-4649: judo-radio-utils: Update SRCREV --------------------------- Commit: aab6bc5 --------------------------- Bugfix: Set radio regulator to a fixed voltage Defining a 'fixed' regulator with differing min and max voltages is not supported and was preventing the radio and all derived regulators (Wi-Fi regulators in particular) from being initialized properly, this does not affect the HW in any way as the regulators are not SW controlled, just cleaning up a harmless warning. --------------------------- Commit: 83f4dc9 --------------------------- DeviceTree: remove non-removable property from Wi-Fi With a GPIO-controlled power supply (`WF_BT_PWR_EN` GPIO) the Wi-Fi module is effectively removable and removing the `non-removable` property ensures the kernel will detect and initialize the Wi-Fi interface whenever it is powered up. --------------------------- Commit: 4248776 --------------------------- Wifi Silex SDMAC: configurable IO voltages Ongoing issues are preventing the Wi-Fi module from operating at 1v8 IO voltages, but there is still a need to be able to test fixes. Allow Wi-Fi IO voltages to be configured via the presence of a `enable_wifi_1v8` file in the boot partition, (i.e `/boot/enable_wifi_1v8`) that results in an overlay being applied to the kernel DeviceTree by U-boot. --------------------------- Commit: 8a67899 --------------------------- SUMO-483: Convert ath10k_compliance feature to compliance the `compliance` feature now enables compliance FW and scripts for more than just the Wi-Fi module, currently adds support for running BT in compliance modes too. --------------------------- Commit: bc33235 --------------------------- program-uhf-fpga: Remove references to Jenkins --------------------------- Commit: cf28d45 --------------------------- CSNMR-5436: program-uhf-fpga: Update to latest nightly FPGA build --------------------------- Commit: 70286cb --------------------------- CSNMR-5454: suhf-sword: reset FPGA on driver startup This changeset updates the SRCREV for Scabbard to integrate control of the FPGA_RESET line into suhf_driver, and adds the libgpiod dependency to the recipe. --------------------------- Commit: 0f13d92 --------------------------- CSNMR-5454: Mark FPGA_RESET pin as active high The FPGA_RESET pin is actually active high, so the 'n' at the start of the pin name is not correct. --------------------------- Commit: d66b514 --------------------------- SUMO-491: GNSS: use internal TCXO Stella receivers have an upgraded internal TCXO and do not require the external one (that is being de-popped) to meet our accuracy requirements. Can force all devices to use the internal TCXO (including devices with BD992 receivers) as all MP10XX devices are going to be upgraded to Stella receivers before release to customers. Legacy/development devices with BD992 receivers will still function fine for development purposes, but will suffer performance issues in extreme conditions (mainly temperature extremes?). --------------------------- Commit: 9cfc28e --------------------------- SUMO-491: Refactor GNSS power management service Properly sequence GNSS power/reset and bootstrap option GPIOs on device startup. Ensure the receiver is powered up before starting the GNSS 'handler' service and ensure restarts of this service propagate out to the handler service. --------------------------- Commit: 1d66513 --------------------------- CSNMR-5250: suhf-sword: update Sword SRCREV --------------------------- Commit: 36bab3c --------------------------- CSNMR-5250: suhf-sword: Recover suhf_driver service on failure Previously, the suhf_driver service would stop when the suhf_driver binary failed for any reason. Because the driver should not stop unless it is ordered to by the UHF configuration agent, this changeset makes systemd restart the suhf_driver service automatically when it fails. --------------------------- Commit: 4cdb3aa --------------------------- SUMO-276: Add instructions for building bt_host_uart_dfu --------------------------- Commit: 6c2c221 --------------------------- SUMO-276: Bluetooth: Firmware update Add firmware and programmer for Lyra24 Bluetooth module. `bt_host_uart_dfu` requires the Silicon Labs IDE and tools to build so the binary is checked in. `write_bt_firmware.sh` wraps the programmer binary and sequences the device-specific GPIOs. --------------------------- Commit: b25f479 --------------------------- CSNMR-5198: Enable Satel radio power only on 510 variant Power to the Satel radio is enabled by the enable-uhf service, which is now only started on the -510 variant of Judo as this is the only variant which has the Satel radio installed. --------------------------- Commit: 6899700 --------------------------- SUMO-539: Improve OPTEE signing reliability Make the kernel fitimage recipe depend on the OPTEE recipe, rather than relying on inter-task-dependencies alone. In the `optee-os` recipe, deploy the public key that can be used to check/verify the OPTEE TA signatures - using the script from: https://github.com/OP-TEE/optee_os/blob/master/scripts/sign_encrypt.py Note: neither the public key nor the above script are included in the OS image - must be copied around manually when testing TA signatures. TA signing key is deployed to a fixed filename for ease of artifacting in CI/Bamboo. ## meta-judo-proprietary --------------------------- Commit: 702f8c4 --------------------------- CSNMR-5428: Update SRCREV Update SRCREV for judo-cfg-agents --------------------------- Commit: 4f72db8 --------------------------- CSNMR-5428: Update SRCREV for rust-gnss Update rust-gnss SRCREV to add GNSS reset support to 4c0485a14818bd315b6043e059ab6c4340d97642 --------------------------- Commit: 874ba56 --------------------------- CSNMR-5628: python3-mqtt-api: Manage run-phoenix-suhf in functional tests using systemd service --------------------------- Commit: 4cd307c --------------------------- CSNMR-5628: python3-mqtt-api: Manage send-observables in functional tests using systemd service --------------------------- Commit: 72579a7 --------------------------- CSNMR-5628: python3-mqtt-api: Use enable-uhf service to sequence TR49 power in functional tests --------------------------- Commit: 6ac2412 --------------------------- CSNMR-5564: Update SRCREV for judo-cfg-agents --------------------------- Commit: d442f89 --------------------------- CSNMR-5600: update SRCREV Update SRCREV for bd992-fw-update repo which holds stella 6.60-0.27 image. --------------------------- Commit: a444478 --------------------------- CSNMR-5600: update Stella firmware from 6.60-0.24 to 6.60-0.27 Install Stella firmware image 6.60-0.27 --------------------------- Commit: bab1362 --------------------------- CSNMR-5426: Update SRCREV soft_reset for system_control Update SRCREV for soft_reset support in system_control_agent. Add bash script that needs to run as a separate process to perform a system control soft_reset. Services being restarted are all cfg-agent agents except for licensing. Do not restart licensing agent as it needs to start and be ready before the other agents start. --------------------------- Commit: 5d3d493 --------------------------- CSNMR-5538: Minimise restarting of UUT MQTT brokers This changeset adds code to start-tmux-monitor-controller which avoids restarting the MQTT brokers on the UUT and secondary unit if the associated bridge is already configured. --------------------------- Commit: 7a1a406 --------------------------- CSNMR-5538: Fix HaLow subnet The default HaLow subnet is 192.168.44.0/24, not 192.168.200.0/24. --------------------------- Commit: 4d6d4d1 --------------------------- CSNMR-5538: Initialize HaLow in emc-test-monitoring This changeset reinstates the HaLow initialization in the emc-test-monitoring script and removes the associated code from start-tmux-monitor-controller.sh. It also tears down the HaLow AP and STA when the test console is shut down. --------------------------- Commit: 617f1d8 --------------------------- CSNMR-5214: Add SRCREV for Stella auto fw update GNSS Manager agent checks for Stella or BD992 to determine generate GNSS configuration overriding the Stella firmware image and firmware checksum on a units with Stella. --------------------------- Commit: 0104484 --------------------------- CSNMR-5214: Add Stella firmware image to Judo OS Add Stella 6.60_0.24 firmware image to Judo OS --------------------------- Commit: ef76cd6 --------------------------- CSNMR-5214 Update to latest rust-gnss Update rust-gnss to SRCREV_rust-gnss = "bd4b506b1a9ced7cbba19c33ff44736dbf1e94dc" which includes: bd4b506b1 Add HL_nGPIO_CTL to phoenix-handler 8d433fcc0 Try harder to get fw checksum to avoid unnecessary installs --------------------------- Commit: 404ce47 --------------------------- CSNMR-5572: update SRCREV --------------------------- Commit: a4e402a --------------------------- CSNMR-4051: Update SRCREV for judo-cfg-agents --------------------------- Commit: 80fcc22 --------------------------- CSNMR-5537: Initialize Wi-Fi in emc-test-monitoring This changeset reinstates the Wi-Fi initialization in the emc-test-monitoring script and removes the associated code from start-tmux-monitor-controller.sh. It also tears down the Wi-Fi AP and STA when the test console is shut down. --------------------------- Commit: 70d69a0 --------------------------- CSNMR-5462: update SRCREV --------------------------- Commit: 896d158 --------------------------- CSNMR-5002: Add GNSS Logging support to GNSS Mgr Add GNSS Logging support to GNSS Manager Agent. --------------------------- Commit: 2a0f2d2 --------------------------- CSNMR-5534: Update SRCREV --------------------------- Commit: 3532a4b --------------------------- CSNMR-4649: Bridge UUT and secondary unit MQTT brokers to RPi broker This change is necessary because the test controller and monitor expect to receive messages from both units on the RPi broker and send commands to both units using the same broker. --------------------------- Commit: 5d463af --------------------------- CSNMR-4649: Tweak keyfile name in ssh-add helper message --------------------------- Commit: 0c262ed --------------------------- CSNMR-4649: python3-mqtt-api: update SRCREV --------------------------- Commit: 56aef79 --------------------------- CSNMR-4649 Functional Test Updates for Swift - emc-test-monitoring: - update IP addresses to match compliance scripts (we will want to revert the ip address changes long term) - change the cell ping interval to 3, to cut down on bandwidth usage - start-tmux-monitor-controller: - Configure UHF, Halow, and WiFi - Check for ssh agent to be running, which is recommended for ssh config of UHF, Halow and WiFi. - Don't run mqtt-select-device-id.py, if uut_id.env file exists - tweak the tmux window split ratios --------------------------- Commit: 13a689a --------------------------- CSNMR-4816: Update SRCREV for judo-cfg-agent and cfg-mgr --------------------------- Commit: c3ca1e7 --------------------------- CSNMR-4816: Fix list formatting in judo-cfg-agent --------------------------- Commit: 62ceaaa --------------------------- Bugfix: Install Wi-Fi certification scripts to root home Install into where-ever the `ROOT_HOME` variable points rather than a hard-coded `/home/root` path (as the root home is now `/root` by default). --------------------------- Commit: baeef89 --------------------------- SUMO-483: BT compliance tools and script Leverages work done for ORCA/Porpoise and modifies it to work on MP10XX devices. --------------------------- Commit: 5aeace2 --------------------------- CSNMR-5265: Update rust-gnss for Stella and Sumo Update Judo rust-gnss SRCREV from: ef2f9dff19cef99d63e19c1a5fbbded18bb46296 to: d9360226a12947b7dd8035f2d9b83ee929a0540a CSNMR-5265: Add support for timo/sumo/kendo RxIDs SUMO-493 : Add sumo board ref to pinmux hardware check --------------------------- Commit: 3e1cc76 --------------------------- CSNMR-5478: Update SRCREV --------------------------- Commit: 4fc9bfc --------------------------- CSNMR-5363: Update SRCREV for judo-cfg-agents --------------------------- Commit: da5089d --------------------------- SUMO-279: Disable UHF agent on Sumo Support for configuring Phoenix radio to be added in SUMO-540. --------------------------- Commit: f6862e6 --------------------------- SUMO-279: ms9xx-handler -> rust-gnss refactor Renames/replaces the existing `ms9xx-bridge` recipe as that builds many more binaries than just the MS9xx bridge, including `phoenix-handler`. Split artifacts into individual packages so they can be individually selected/installed. Remove the unused patch that was fixed in the latest `rust-gnss` update, see meta-judo-proprietary commit `a5cc084602f80c72f21a0d97f1efd5d43de0dc31` --------------------------- Commit: fede53b --------------------------- CSNMR-4963: Update SRCREV for cfg-mgr and judo-cfg-agents --------------------------- Commit: a002332 --------------------------- CSNMR-5407: Update SRCREV for judo-cfg-agents --------------------------- Commit: 265e8b2 --------------------------- CSNMR-5208: add mfgfs-create dependency to cfg agents Added mfgfs-create runtime dependency to cfg agents recipe. The system reset agent depends on the write/clear mfg partition script that mfgfs-create installs. --------------------------- Commit: f0f1049 --------------------------- CSNMR-5410: update SRCREV --------------------------- Commit: de6c1ba --------------------------- CSNMR-5362: Update SRCREV for judo-cfg-agents ## mqtt-api --------------------------- Commit: 55244de --------------------------- CSNMR-5538: Set up and tear down HaLow interface in controller scripts --------------------------- Commit: 95b7aa1 --------------------------- CSNMR-5537: Add functional test teardown script This script currently only disables the Wi-Fi interfaces on the UUT and secondary unit. It will be extended to disabling the HaLow interfaces on these units as part of CSNMR-5538, and can be used for any other teardown that may be required in future. --------------------------- Commit: 7507c7c --------------------------- CSNMR-5537: Add wait_for_net_interface_state to controller console This is used to ensure that wireless IP connections (Wi-Fi and HaLow) are disestablished before exiting the test console following a test session, by waiting for the interface to go down. --------------------------- Commit: 3f9ec82 --------------------------- CSNMR-5537: Add wait_for_dhcp_address_in_range to controller console This is used to ensure that wireless IP connections (Wi-Fi and HaLow) are established during setup before starting the test session, by waiting for the access point to issue an IP address to the station via DHCP. --------------------------- Commit: 3401cff --------------------------- CSNMR-5537: Set up secondary unit as Wi-Fi AP in EMC tests --------------------------- Commit: 1808b88 --------------------------- CSNMR-4649: mqtt-api-controller.py: Enable functional test service on UUT at startup --------------------------- Commit: 3e6fdfb --------------------------- CSNMR-4649: mqtt-select-device-id.py: Allow selecting Sumo device IDs --------------------------- Commit: 88dde3e --------------------------- CSNMR-4649: mqtt-api-monitor: Use event_diag_gnss_data for GNSS tests --------------------------- Commit: 015cdc4 --------------------------- CSNMR-4649: mqtt-api-controller: Allow exiting session without stopping existing requests --------------------------- Commit: 26d382a --------------------------- CSNMR-4649: scripts: Add scripts for setting up HaLow and Wi-Fi STAs --------------------------- Commit: 5e1aeec --------------------------- CSNMR-4649: scripts: Exit after setting up Wi-Fi/HaLow AP This change makes the controller shell scripts for setting up Wi-Fi and HaLow APs exit after the AP is set up. --------------------------- Commit: 8f7bc09 --------------------------- fixup: remove mfg api hard coded addresses --------------------------- Commit: 3ad734f --------------------------- CSNMR-4928: swap MFG API wifi config for systemctl/pkill Swapped DISABLED agent config call for systemctl disable and pkill hostapd. The DISABLED command started intermittently failing at Jabil and is now non functional. This method continues to work for getting rid of hostapd and halow-ap processes. ## nrc7394_sw_pkg --------------------------- Commit: 7068a25 --------------------------- Bugfix: Fix default FW file The default non-EEPROM FW fails to initialise the Halow module with the below error (has a HW revision 0f 0x0000). ``` Feb 17 23:14:38 mp1010 kernel: nrc80211 spi3.0: size of bd file is 8464 Feb 17 23:14:38 mp1010 kernel: nrc80211 spi3.0: Major 01 Minor 04 Total len 2100 Num_Data_Groups 0024 Checksum B3FF Feb 17 23:14:38 mp1010 kernel: nrc80211 spi3.0: target version is not matched(0 : 2) Feb 17 23:14:38 mp1010 kernel: nrc80211 spi3.0: target version is not matched(0 : 3) Feb 17 23:14:38 mp1010 kernel: nrc80211 spi3.0: target version is not matched(0 : 1) Feb 17 23:14:38 mp1010 kernel: nrc80211 spi3.0: target version is not matched(0 : 4) Feb 17 23:14:38 mp1010 kernel: nrc80211 spi3.0: BD file is invalid! Stop loading FW ``` ## scabbard-judo --------------------------- Commit: 1b859e3 --------------------------- CSNMR-5454: Toggle FPGA_RESET line on suhf_driver startup ## sword --------------------------- Commit: 134072a --------------------------- CSNMR-5250: Report hardware error if FIFO cannot be flushed On the unit which fails to start suhf_driver, suhf_driver gets into an infinite loop flushing the FPGA's FIFO buffer because the FIFO never becomes empty, probably because there is an issue with the FPGA hardware. This changeset sets a hard limit of 4096 bytes that may be flushed out of the FIFO before the driver stops with an error code. TO-DO: The value of 4096 was chosen because that is the value used for Judo's version of Scabbard, but Sword needs to get this value from its host's version of Scabbard. --------------------------- Commit: 6dfe5ed --------------------------- CSNMR-5250: Reduce duplication of FIFO empty detection code ## yocto-docs --------------------------- Commit: f2811603d --------------------------- set_versions.py: fix codestyle/formatting issues Fix issues reported by pycodestyle (long lines, spurious or missing whitespaces, underindented lines), and add the missing description of the get_abbrev_hash() function. Signed-off-by: Antonin Godard Signed-off-by: Richard Purdie --------------------------- Commit: 6c8961015 --------------------------- set_versions.py: move the poky_mapping dict where it is used The poky_mapping dict isn't used until after the replacements dict is created. Move it's definitions to where it's used. Also fix a codestyle issue while we're at it (whitespace before ":"). Signed-off-by: Antonin Godard Signed-off-by: Richard Purdie --------------------------- Commit: 98815cee1 --------------------------- set_versions.py: use bitbake mappings from remote releases.json After [1] we can now use the bitbake version mappings maintained in the remote releases.json file, allowing to reduce maintenance effort here. [1]: Signed-off-by: Antonin Godard Signed-off-by: Richard Purdie --------------------------- Commit: 5d711350a --------------------------- ref-manual/system-requirements: add fedora 43 as supported distro Fedora 43 is now properly tested for Poky[1], so let's add it to the list of supported distros. [1] https://git.yoctoproject.org/meta-yocto/commit/?id=4c156ef673f8ff7c967b2b34dd1d366e1b74d41c Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: b92cffd3e --------------------------- ref-manual/system-requirements: update list of EoL distros we still test Fedora 41 is EoL since end of 2025[1] and Ubuntu 25.04 since mid January 2026[2], so let's add them to the list of distros we're testing but don't receive updates for. [1] https://docs.fedoraproject.org/en-US/releases/eol/#_unsupported_fedora_linux_releases [2] https://documentation.ubuntu.com/project/release-team/list-of-releases/#end-of-life Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 5afd21a88 --------------------------- overview-manual/concepts: list other possible class directories Classes can be located in three different directories since Langdale (4.1), so let's update which directories you can find in a distribution layer. Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 0d05dedd6 --------------------------- ref-manual: delete claim PACKAGE_EXCLUDE does not support .deb The current explanation of "PACKAGE_EXCLUDE" claims that this variable does not support .deb packaging; this is incorrect. Signed-off-by: Robert P. J. Day Tested-by: Antonin Godard Signed-off-by: Antonin Godard --------------------------- Commit: 73b3beb99 --------------------------- overview-manual: escape wildcard in inline markup This change escapes the wildcard in 'recipes-*' to properly italicize the string Signed-off-by: Kristiyan Chakarov Signed-off-by: Antonin Godard --------------------------- Commit: 5bebe38a8 --------------------------- set_versions.py: build with versions from releases.json Build using the releases.json file obtained from https://dashboard.yoctoproject.org/releases.json. If this file is already downloaded, don't fetch it again. If we fail fetching the file, use default values (and print a warning). Signed-off-by: Antonin Godard Signed-off-by: Richard Purdie --------------------------- Commit: d97cfeaa5 --------------------------- tools/containerfiles: fix docker warning The current Containerfiles produce the following warning with docker: 1 warning found (use docker --debug to expand): - UndefinedVar: Usage of undefined variable '$PWD' (line 12) Remove $PWD and set the WORKDIR so that our pip3_docs.sh script finds the virtual env. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: ba351b354 --------------------------- convert SVGs to PDF and PNG using sphinxcontrib.rsvgconverter plugin The sphinxcontrib.rsvgconverter plugin allows to generate a PDF or PNG from an SVG. This is what we already do manually via Make targets but it isn't good enough. Sphinx generates a cache upon first parsing and stores it in a doctrees directory. Sphinx claims that it can be shared between all builders[1]. For glob patterns in image or figure directives, Sphinx will look for all possible matches in the source tree and store those in a "candidates" map for each file, along with the associated mimetype. When building, Sphinx will then look in this map to try to find a match in the current builder's supported_image_types. If none is found, the build will fail. The latexpdf (using the LaTeXBuilder) target does not support SVGs by default[2]. We used Make to generate PDFs from them before generating the doc PDF though (see PDFs variable and %.pdf in Makefile) and that type is supported by default[2]. The epub (via the Epub3Builder) target does support SVGs by default[3] but we disabled their support in commit ff3876ca4910 ("conf.py: use PNG first in EPUB output"). We used Make to generate PNGs from them before generating the doc epub though (see PNGs variable and %.png in Makefile) and that type is supported (c.f. Epub3Builder.supported_image_types in our conf.py). The issue is that this is done transparently from Sphinx. When we generate the PDFs or PNGs variants of the SVGs, we put them in-tree directly along their source file. Then, when caching, Sphinx will find both the source file and the appropriate variant. However, the cache isn't updated if there are new files in the tree and the source rST files aren't modified. So, the cache will not have its map updated and we won't be able to find the new variant when building for a non-SVG-compatible builder. Take the following scenario: - start from a clean source file (fresh clone or git clean -ffdx) - build the html target (which supports SVGs by default[4]) - sphinx will find all the files in the source tree matching the glob pattern in ".. image:: test.*", in our case only an SVG since the PDFs and PNGs are only generated for the latexpdf and epub targets respectively. The cache will only store the path to the SVG file because it is the only source file that matches the glob, - attempt to build the epub target (which doesn't support SVGs, only PNGs) - Sphinx checks for the file to include for '.. image:: test.*' and finds an SVG in the cache map and then check the list of supported image types for the Epub3Builder and find that it doesn't support that. It cannot find anything satisfying the dependency and thus fails to build. This scenario can easily be reproduced by running the `make all` command since the html builder will be used first, then epub and finally latexpdf. The `make publish` target works by chance, because the epub builder is built first and will cache SVG + PNG for each glob, then the latexpdf builder is built and supports PNGs so that's fine and then html, which supports SVG as well. To fix this issue, we could simply always generate PDFs and PNGs of all SVGs in the source tree, but this isn't ideal. Instead, let's use an ImageConverter from a Sphinx plugin. This allows to map a plugin as being able to generate a file of type Y from a file of type X. When Sphinx wants to build an image, it'll try to find the image with the type the current builder supports in the cache. If it cannot, it's going to try to find an ImageConverter plugin that is able to convert one of the image types in cache with one of the image types the current builder supports. Then Sphinx will call this plugin to generate the file and put it into the build directory (not in the source!). This allows to simplify the Makefile as well and is a much cleaner approach. The epub target is removed as the catch-all target contains the same instructions as the epub target we remove in this commit. sphinxcontrib.rsvgconverter is a plugin from sphinxcontrib-svg2pdfconverter python module. SVGs to PNGs is only supported since 2.0.0. [1] https://www.sphinx-doc.org/en/master/man/sphinx-build.html#cmdoption-sphinx-build-d [2] https://www.sphinx-doc.org/en/master/usage/builders/index.html#sphinx.builders.latex.LaTeXBuilder.supported_image_types [3] https://www.sphinx-doc.org/en/master/usage/builders/index.html#sphinx.builders.epub3.Epub3Builder.supported_image_types [4] https://www.sphinx-doc.org/en/master/usage/builders/index.html#sphinx.builders.html.StandaloneHTMLBuilder Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 4fbae0147 --------------------------- consistently use pip packages for all distros Until now, Debian and Ubuntu had special instructions for installing Python packages to build the docs. Specifically, the packages that all other distros are installing via pip were installed from the package feed (apt). In the next commit, a new Python module will be required, which isn't available in the package feed. We thus have two options, have a pip3 Debian/Ubuntu-specific set of instructions to install this one package, or migrate packages currently installed through the package manager to use pip like other distros. To lower maintenance burden and the possibility of missing to update both "generic" and "Debian/Ubuntu" variants of pip instructions, the latter was chosen. Suggested-by: Antonin Godard Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 16599c4f6 --------------------------- use a venv for installing packages with pip In a few commits, we'll migrate Debian/Ubuntu to use pip packages. However, Debian rejects pip commands outside of virtual environments. It is generally best practice to use virtual environments with pip, so let's migrate distros that currently use pip to use virtual environments as well. Because BitBake is a Python project and requires a few Python packages that we currently install with distro package managers, it would be best to also include system packages via --system-site-packages when creating the virtual environment such that the user would either be able to build the docs or run BitBake, from the same venv. Unfortunately, one of the packages[1] we'll include in an upcoming commit breaks sphinxcontrib namespace package and thus we cannot have a mix of sphinxcontrib-namespaced packages in and outside the venv. Let's create a completely empty venv for now until the offending package is fixed (see pending merge request[2]). Let's instruct the user to install the docs venv in a relative directory (likely the root of the yocto-docs git repo) instead of guessing a safe location (/tmp may or may not be an actual volatile directory depending on the distro). [1] https://github.com/missinglinkelectronics/sphinxcontrib-svg2pdfconverter/ [2] https://github.com/missinglinkelectronics/sphinxcontrib-svg2pdfconverter/pull/34 Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 8354f3b5b --------------------------- Makefile: use python3 -m sphinx instead of sphinx-build sphinx-build is a "script"[1][2] which is then made available on the system as a binary, upon installing the sphinx package. You can have multiple instances of sphinx installed, e.g. system-wide via the traditional distro package manager or in a venv. In the case of a venv, if you have sphinx already installed in your host system, source the venv activate script and then install sphinx via pip inside the venv, sphinx-build command will still point at the sphinx from the host system. The work-around is either to run the venv activate script again, run `hash -r` or `rehash` in your current shell, or call python3 -m sphinx instead of the Python script. Let's do the latter as it's the easiest/least-user-facing solution. This works because python3 interpreter will be the one from the venv, and it sets the appropriate paths for loading modules (use only packages from the venv by default, or prefer them if --system-site-packages is passed). [1] https://github.com/sphinx-doc/sphinx/blob/v9.1.0/pyproject.toml#L99 [2] https://packaging.python.org/en/latest/guides/writing-pyproject-toml/#creating-executable-scripts Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 8cc5a1302 --------------------------- qemu.rst: parsing of the code-block was not done Because the '::' at the previous line starts a literal block. Signed-off-by: Jan Vermaete Signed-off-by: Antonin Godard --------------------------- Commit: ff3cfb0ed --------------------------- ref-manual: add SPDX_INCLUDE_KERNEL_CONFIG and SPDX_INCLUDE_PACKAGECONFIG Adding documentation for the new variables to export the Linux kernel configuration and recipe PACKAGECONFIG features into the SPDX document. Signed-off-by: Kamel Bouhara (Schneider Electric) [AG: fix variable ordering https://lore.kernel.org/r/DGO3PZFM5WDO.3C92IW83D5RXA@bootlin.com] Signed-off-by: Antonin Godard --------------------------- Commit: 6ab93e2eb --------------------------- dev-manual: delete references to "tar" package format There is no "tar" packaging format anymore, so delete these references. Signed-off-by: Robert P. J. Day Signed-off-by: Antonin Godard --------------------------- Commit: 3f4ed1d41 --------------------------- Makefile: pass -silent to latexmk This should drastically decrease the number of lines in the docs build logs, which currently can go up to 1700000 lines on the Autobuilder[1]. [1]: https://autobuilder.yoctoproject.org/valkyrie/#/builders/34/builds/400/steps/6/logs/stdio Signed-off-by: Antonin Godard --------------------------- Commit: 538b55418 --------------------------- what-i-wish-id-known.rst: replace figure by the new SVG We have a new SVG figure for the Yocto Project workflow figure, which was updated with d2aaf54bee49 ("overview-manual: convert YP-flow-diagram.png to SVG"). The one in the "What I wish I’d known about Yocto Project" is the same. Remove the PNG and make a reference to the new one. Signed-off-by: Antonin Godard --------------------------- Commit: 2817a9077 --------------------------- tools/host_packages_scripts/tlmgr_docs_pdf.sh: install perl Fix an issue reported by Quentin on IRC: The LaTeX files are in _build/latex. Run 'make' in that directory to run these through (pdf)latex (use `make latexpdf' here to do that automatically). make[1]: Entering directory '/docs/documentation/_build/latex' latexmk -pdf -dvi- -ps- 'theyoctoproject.tex' Can't locate Time/HiRes.pm in @INC (you may need to install the Time::HiRes module) (@INC contains: /usr/local/lib64/perl5 /usr/local/share/perl5 /usr/lib64/perl5/vendor_perl /usr/share/perl5/vendor_perl /usr/lib64/perl5 /usr/share/perl5) at /opt/docs-build-tex-tools/tl/bin/x86_64-linux/latexmk line 97. BEGIN failed--compilation aborted at /opt/docs-build-tex-tools/tl/bin/x86_64-linux/latexmk line 97. make[1]: *** [Makefile:28: theyoctoproject.pdf] Error 2 make[1]: Leaving directory '/docs/documentation/_build/latex' make: *** [Makefile:74: latexpdf] Error 2 make: Leaving directory '/docs/documentation' We require perl for the latexpdf target to complete successfully. Add it to the tlmgr_docs_pdf.sh script. Tested-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 3f052ae60 --------------------------- docs-wide: remove Bazaar fetcher references Remove the Bazaar fetcher references as it was deprecated in [1]. This fixes the following build issue on master: .../ref-manual/variables.rst:1506: WARNING: term not in glossary: 'bitbake:BZRDIR' [ref.term] [1]: https://git.openembedded.org/bitbake/commit/?id=8e057d54f09ca0a1fb64c716da6b66f0ce4555b0 Signed-off-by: Antonin Godard --------------------------- Commit: c9f83bb2e --------------------------- ref-manual/classes.rst: fix broken links to U-Boot documentation In the chapter describing the "uboot-sign" class Signed-off-by: Michael Opdenacker Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 7570d210e --------------------------- _templates/footer.html: show a dynamically created id of the docs Since we have switched to branch tips for all versions of the active releases, having the information of which commit id of yocto-docs was built is useful. Show it in the footer as a link to git.yoctoproject.org/yocto-docs. When a tag is built, create a link to the tag instead. When building on a floating commit, this will create an invalid link, but this never happens when publishing the documentation, only for local development. Signed-off-by: Antonin Godard Signed-off-by: Richard Purdie --------------------------- Commit: 065e70a62 --------------------------- set_versions.py: set a more accurate DISTRO value The changes from the previous commit make the current value inaccurate in some situations, especially since this variable is used in some commands. Take these situations into account and compute a more accurate value for DISTRO. Signed-off-by: Antonin Godard Signed-off-by: Richard Purdie --------------------------- Commit: a8749670e --------------------------- switchers.js: rework the logic and improve it Rework the generation of the switcher menu, and also fix it. It's not obvious how this could be split for easier review, as many parts depend on one another, so this is what this patch changes and fixes: - The menu only contains branch tips versions of the documentation for supported releases now - even though the latest tag is displayed in the menu. This approach will fix the lack of release notes for the latest version of an active branch, as release notes are always published after the yocto-docs repository is tagged. - Show a commit id when building on a separate commit that is not a branch tip or not a tag. This should only ever happen when building locally (not for the published version). - All stable/LTS versions are displayed in the menu UNLESS the current version for which we are viewing is part of them, in which case it's shown last in the menu. - Rework the logic to show outdated or obsolete versions to conform with the above changes, but overall it behaves the same: - granted that Kirkstone is an active release, and 4.0.32 is the latest tag for this release, the "outdated" banner is displayed if we browse any version of Kirkstone inferior to 4.0.32. - browsing any version not part of the active releases (all LTS versions + the stable branch) will show the "obsolete" banner. Signed-off-by: Antonin Godard Signed-off-by: Richard Purdie --------------------------- Commit: b0a9cf917 --------------------------- set_versions.py: remove duplicate and unneeded logic The following commit will make it unneeded to pass the version to the script directly, as it will be automatically deduced no matter which branch we're on. We can therefore the ourversion assignment. We still want the "getlatest" special argument, so read it from sys.argv directly. Add a comment mentioning that the order here is important for getlatest to work. Signed-off-by: Antonin Godard Signed-off-by: Richard Purdie --------------------------- Commit: 98840d241 --------------------------- migration-guides/migration-5.3.rst: add note on *FLAGS behavior change Add a note on how existing *FLAGS += assignments can lead to unexpected behavior in the 5.3 release notes. Cc: Richard Purdie Signed-off-by: Antonin Godard --------------------------- Commit: f40dc0b42 --------------------------- set_versions.py: check that we build from the git repository Check that we are building from a Git clone of yocto-docs, not a bare directory, which is not supported as we need to tags to be fetched. Inform the user how to clone the documentation too. Fixes [YOCTO #15834] Signed-off-by: Antonin Godard --------------------------- Commit: 095981c08 --------------------------- migration-guides: add release notes for 4.0.32 Signed-off-by: Lee Chee Yang Signed-off-by: Antonin Godard --------------------------- Commit: e95e60ef8 --------------------------- migration-guides: add release notes for 5.3.1 Signed-off-by: Lee Chee Yang [AG: oecore -> openembedded-core https://lore.kernel.org/r/DGAAISXQEM8B.1JQP7X9NMQ5HA@bootlin.com] Signed-off-by: Antonin Godard --------------------------- Commit: 5fd9d1602 --------------------------- document the OPENSSH_HOST_KEY_DIR and OPENSSH_HOST_KEY_DIR_READONLY_CONFIG variable Added by commit addd80ddfd89 ("openssh: add variable for key path") in OE-Core. Signed-off-by: Patrick Vogelaar Signed-off-by: Antonin Godard --------------------------- Commit: 749b79b1a --------------------------- ref-manual/variables.rst: document new FIT image variables - FIT_MKIMAGE_EXTRA_OPTS: new variable to pass extra mkimage options - FIT_CONF_MAPPINGS for flexible DTB mappings Signed-off-by: Adrian Freihofer Signed-off-by: Antonin Godard --------------------------- Commit: 37667bcc6 --------------------------- brief-yoctoprojectqs/index.rst: remove extra word from text The sentence in brief-yoctoprojectqs/index.rst contained a duplicated word ("dependencies it"). This change removes the extra word to improve clarity for new users following the quick start guide. Signed-off-by: Mohammad Rahimi Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 411122812 --------------------------- dev-manual/packages.rst: fix example recipe version The example recipe taken above is hello-world on version 1.0 (because PV equals "1.0+git". Fix this issue. Signed-off-by: Antonin Godard --------------------------- Commit: 7a0324b6a --------------------------- dev-manual/packages.rst: rename r0.0 to r0 when PR server is not enabled When we don't have a PR server enabled, we don't have leading ".0" to the PKGR variable, as this is added by the PR server. Signed-off-by: Antonin Godard --------------------------- Commit: 09f0430bc --------------------------- dev-manual/packages.rst: pr server: fix and explain why r0.X increments on SRCREV change The current example of the SRCREV change triggering a gitX bump is wrong, as both gitX and r0.X get incremented. Why this is happening is explained in bug 15729, which I copy here: > +gitX+ is indeed related to changes in the source code. > > r0.X is bumped each time the checksum of the do_package task of the > simple-hello-world-git recipe changes. This happens here: > https://git.openembedded.org/openembedded-core/tree/meta/classes-global/package.bbclass?id=235e6d49e5888ad04416219e10b6df91a738661a#n306 > > This line sets the value of PRAUTO and represents the number X found in > r0.X. It will in the end make it into EXTENDPRAUTO, which itself makes > to PKGR == r0.X. > > This line calls getPR(version, pkgarch, checksum). Between test case 5 > and 6, only the checksum changes. This checksum is the checksum of the > do_package task (gotten from get_do_package_hash() above). > > Now, let's dump what changed with regards to this task between two > consecutive runs, using the sigdata file in build/tmp/stamps/: > > ``` > [...] > Variable fetcher_hashes_dummyfunc value changed from '2650ad6714c3f3248abfe9d3daf1196f307ed494' to '4af682a50174f5deb0397847da97d7cdba4ad067' > ``` > > The last line shows that the value of fetcher_hashes_dummyfunc changed > from '2650ad6714c3f3248abfe9d3daf1196f307ed494' to > '4af682a50174f5deb0397847da97d7cdba4ad067'. Those are the commit hashes > in the git history of the simple-hello-world-git repository. > > Now you can see why this 0.X gets bumped, is because of the SRCREV change. Fix the example, and detail what gets changed and why. [YOCTO #15729] Cc: Robert Berger Signed-off-by: Antonin Godard --------------------------- Commit: d3f9b0a5a --------------------------- ref-manual: Document updated UBOOT_CONFIG flow The UBOOT_CONFIG flow was updated in oe-core [1]. Document all of the new variables and preferred method of specifying the various controls for each config. [1] https://git.openembedded.org/openembedded-core/commit/?id=cd9e7304481b24b27df61c03ad73496d18e4d47c Signed-off-by: Ryan Eatmon [AG: Typo fixes https://lore.kernel.org/r/DG1YKKEBYK3R.1RDFPILOIHIOP@bootlin.com] Signed-off-by: Antonin Godard --------------------------- Commit: 3ec23e9fb --------------------------- what-i-wish-id-known.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: b0eed805b --------------------------- test-manual/understand-autobuilder.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: dfe9140b6 --------------------------- test-manual/reproducible-builds.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 783aa4d14 --------------------------- ref-manual/yocto-project-supported-features.rst: remove obsolete poky repo reference Poky -> meta-yocto link replacement. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 805b666f0 --------------------------- ref-manual/system-requirements.rst: fix wrong path to install-buildtools Reported-by: Quentin Schulz Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: ca6a30277 --------------------------- ref-manual/system-requirements.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 66fb4d577 --------------------------- ref-manual/release-process.rst: remove repeated "in the" Reported-by: Quentin Schulz Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: e7231e74a --------------------------- ref-manual/release-process.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: a2e67529c --------------------------- ref-manual/features.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 173edb82b --------------------------- ref-manual/faq.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Rework the difference between Poky and OpenEmbedded. Copy the note from terms.rst regarding the now obsolete Poky repository. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 1923550cd --------------------------- ref-manual/devtool-reference.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 91605ea5c --------------------------- dev-manual/vulnerabilities.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Mention that only vulnerabilities in OpenEmbedded-Core are tracked as that's where the packages are. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 0bcfcf8c7 --------------------------- dev-manual/upgrading-recipes.rst: move meta-openembedded to bitbake-builds/layers It makes more sense as we expect the layout to be that way. Suggested-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 67ae0525f --------------------------- dev-manual/upgrading-recipes.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: f0c6ba3f4 --------------------------- dev-manual/qemu.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: faf322d8d --------------------------- dev-manual/python-development-shell.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: d2cddd8c5 --------------------------- dev-manual/packages.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 1820232eb --------------------------- dev-manual/licenses.rst: instruct to use git-archive instead of removing .git Removing .git directories is destructive. Instruct to use git archive instead. Suggested-by: Quentin Schulz Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: f0b3659e6 --------------------------- dev-manual/licenses.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 6ec9ce47f --------------------------- dev-manual/init-manager.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 04af92638 --------------------------- dev-manual/custom-template-configuration-directory.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Adapt paths to match to encouraged Source Directory layout. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 421ec4742 --------------------------- dev-manual/custom-distribution.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 6e440c9ae --------------------------- dev-manual/build-quality.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 6c8d49202 --------------------------- contributor-guide/identify-component.rst: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 0cb6e102f --------------------------- README: replace obsolete substitution variables Replace obsolete substitution variables by current ones from poky.yaml.in. Suggested-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: d95db54e4 --------------------------- README: remove obsolete poky repo references Refresh the document now that the Poky repository is obsolete. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 29bd24aee --------------------------- test-manual/runtime-testing.rst: remove obsolete poky repo references Remove reference to the Poky repository, replace by OpenEmbedded-Core. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: d26aab08b --------------------------- sdk-manual/intro.rst: remove obsolete poky repo references Remove the mention of the Poky git repository. Refer to the existing setup instructions. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: d7ce0ad51 --------------------------- sdk-manual/appendix-obtain.rst: remove obsolete poky repo references Rename the source directory to project, as specified in its new definition in terms.rst. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 4c0e97afe --------------------------- ref-manual/variables.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core in most cases. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 89818234a --------------------------- ref-manual/tasks.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 83d88fa2c --------------------------- ref-manual/structure.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core in most cases. Add a introductory directory layout at the top that is reflected in the following section. This document no longer uses Poky as its source directory, so all paths are explicitly mentioned, to highlight the separation between core components. This lead to some re-organization of some sections. Signed-off-by: Antonin Godard --------------------------- Commit: 63d0fc075 --------------------------- ref-manual/images.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 864a7ec1e --------------------------- ref-manual/classes.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 178d31924 --------------------------- overview-manual/yp-intro.rst: remove obsolete poky repo references Remove references to the Poky repository. This document had a lot of reference to it, so a lot was removed. The "Reference Distribution (Poky)" section was removed as there already was a "Reference Embedded Distribution (Poky)" below. This section was simplified as now Poky is just a distro, not a repo. The historical section was also updated to mention the obsolete Poky repo. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 345003a75 --------------------------- overview-manual/development-environment.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Replace the poky examples for the Git section by yocto-docs. Signed-off-by: Antonin Godard --------------------------- Commit: 91e96f5c0 --------------------------- overview-manual/concepts.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Signed-off-by: Antonin Godard --------------------------- Commit: 2a2de4da9 --------------------------- migration-guides/migration-1.4.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: bedbbb69a --------------------------- kernel-dev/common.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: e8b03ddb0 --------------------------- dev-manual/wic.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: de6d0a16d --------------------------- dev-manual/wayland.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 71f9a3f78 --------------------------- dev-manual/temporary-source-code.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: b184acb48 --------------------------- dev-manual/securing-images.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 0dceb6dd9 --------------------------- dev-manual/new-recipe.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 3da2e7bb3 --------------------------- dev-manual/libraries.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: ff8595ec8 --------------------------- dev-manual/layers.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Signed-off-by: Antonin Godard --------------------------- Commit: 40af24177 --------------------------- dev-manual/layers.rst: re-organize the document The first section of the layers document shows how to manually create a layer step-by-step, and references bitbake-layers create-layer at the end. Change the approach of this section to show how to use the bitbake-layers create-layer command to create a layer, and then explain how a layer is organized and configured. This avoids repetition, and we should encourage users to use pre-built tools when they can do things in a less error-prone way. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: d96a93b1c --------------------------- dev-manual/external-toolchain.rst: remove obsolete poky repo references Remove the single poky repository reference in this document, after the poky repository was obsoleted. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 8a8a4d77e --------------------------- dev-manual/error-reporting-tool.rst: remove obsolete poky repo references Remove the single poky repository reference in this document, after the poky repository was obsoleted. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 27cfed1fc --------------------------- dev-manual/debugging.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 2c45968fd --------------------------- dev-manual/building.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core in most cases. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 472a5c353 --------------------------- contributor-guide/submit-changes.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core in most cases. Signed-off-by: Antonin Godard --------------------------- Commit: b47987aba --------------------------- kernel-dev/advanced.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core in most cases, and refer to bitbake-setup or the manual setup in the intro. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 05a506686 --------------------------- bsp-guide/bsp.rst: remove obsolete poky repo references Remove references to the Poky repository, replace by OpenEmbedded-Core in most cases, and refer to bitbake-setup or the manual setup in the intro. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 68586e603 --------------------------- dev-manual/start.rst: remove obsolete poky repo references The Poky repository has stopped being updated in favor of bitbake-setup or manual clones of the required repositories. Simplify the document by making references to these guides as well as removing sections that are covered in the references documents. Keep a "Initializing the Build Environment" as this task is common and referenced in different places in the documentation. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: a7c169229 --------------------------- dev-manual/poky-manual-setup.rst: remove obsolete poky repo references The commit "dev-manual/start.rst: simplify the document after poky obsolescence" removed some of the information related to source archives published on downloads.yoctoproject.org. Mention it in this document as it makes sense to use those when manually setting up a Poky build. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 0b909e6c8 --------------------------- transitioning-to-a-custom-environment.rst: remove obsolete poky repo references - Use :term: for Poky. - Mention use of fragments instead of local.conf. - Remove Yocto < 2.4 reference. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: d7d5ddeda --------------------------- tree-wide: figures: remove title PNGs We used to have banners at the top of each main section before we migrated to Sphinx and those aren't used anymore, so let's remove them from the tree. Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: fdc14a205 --------------------------- ref-manual/terms.rst: refresh the OpenEmbedded build system definition Remove the mention of Poky containing BitBake, just mention BitBake. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: b42c033bf --------------------------- dev-manual/poky-manual-setup.rst: mention the Source Directory After switching to the new definition of "Source Directory", i.e. the base directory for your project (which used to be poky), mention it in this document on manually setting up the layers for a project. As stated in terms.rst, use "project" for the directory name. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 9ad8a0aa7 --------------------------- ref-manual/terms.rst: simplify the Source Directory definition Now that we no longer have Poky around, simplify the Source Directory definition: - Make references to existing documents that show how to create this Source Directory. - This directory contains sources, the layout in which they are presented does not matter so much, as it can be different and/or freeform depending on whether we use bitbake-setup or not. The definition chooses "bitbake-builds" as the name of the source directory and is the one that should be used when showing examples in the documentation. A typical Source Directory would look like the following: bitbake-builds/ ├── build/ └── layers/ Other parts of the documentation referencing to this as the Poky repository still have to be updated. This will also mean that this definition will be used a lot less across the documentation: in most cases, when the "Source Directory" is referenced as the Poky repository, it can be replaced by OpenEmbedded-Core, Bitbake, the documentation, or meta-yocto. We'll keep this definition around regardless, as it can be useful to denote this group of directories in a more general sense. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: dd328b79a --------------------------- ref-manual/terms.rst: refresh the OpenEmbedded-Core definition After the Poky repository obsolescence, refresh the OpenEmbedded-Core definition. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 476780d77 --------------------------- ref-manual/terms.rst: refresh the Build Directory definition After the Poky repository obsolescence, refresh the build directory definition. Fix the formatting and indenting issues while at it. Reviewed-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 6256a73d7 --------------------------- migration-guides: add release notes for 5.0.15 Signed-off-by: Lee Chee Yang Signed-off-by: Antonin Godard --------------------------- Commit: 52ec12522 --------------------------- sdk-manual: appendix-obtain: fix default path for eSDK installer script The eSDK installer script installs in poky_sdk directory but under the user's home directory so let's make that explicit. Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: ede411d70 --------------------------- sdk-manual: using: fix SDK filename example I'm pretty sure we meant to use the DISTRO poky.yaml.in placeholder which can only be replaced when surrounding by & and ; characters. Therefore let's fix this oversight. Also, the example isn't actually pointing at the snapshot, but the actual release, so let's reword that as well so it matches the example. I decided to do that instead of adding +snapshot to the filename because the example in the next section is actually using that same filename so for consistency it seemed more appropriate. Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 41ee642c3 --------------------------- sdk-manual: appendix-obtain: use parsed-literal block for naming convention of the installer scripts It's easier on the eye to use a parsed-literal block for highlighting things that must be changed than simply stating the word, so this makes use of a parsed-literal block instead of a simple inline code-block and also highlight in italic which words are placeholders. Also make the explanation below the literal-block use the same highlighting (italic). Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: ffaea8391 --------------------------- sdk-manual: delete sdk-title PNG It hasn't been used since the migration to Sphinx so let's remove it. Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 39fb14879 --------------------------- sdk-manual: replace sdk-environment PNG with SVG This replaces the PNG figure with an SVG. Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 77b24b4a8 --------------------------- sdk-manual: appendix-obtain: replace eSDK directory structure PNG with a parsed-literal block This PNG is totally unnecessary and makes accessibility much worse, so let's simply replace the PNG with a parsed-literal block showing the same directory layout with the same highlighted placeholders. While at it, update the layout to match the more recent one (5.0.15) and have it alphabetically sorted (mixed files and directories). The placeholders are highlighted the same way inside and outside the parsed-literal block for consistency. Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 0c68497ec --------------------------- sdk-manual: appendix-obtain: replace directory structure PNG with a parsed-literal block This PNG is totally unnecessary and makes accessibility much worse, so let's simply replace the PNG with a parsed-literal block showing the same directory layout with the same highlighted placeholders. While at it, update the layout to match the more recent one (5.0.15) and have it alphabetically sorted (mixed files and directories). The placeholders are highlighted the same way inside and outside the parsed-literal block for consistency. Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: d682bce63 --------------------------- sdk-manual: appending-customizing: use none lexer for BitBake code blocks No lexer exists for BitBake right now, so let's simply disable the lexer entirely by using the "none" lexer. I'm using "none" instead of "text" to be able to "easily" replace none code blocks the day a BitBake lexer exists. I "reserve" "text" for actual text content (e.g. a filename, or the output of a command line without the command line). Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 10d40604c --------------------------- sdk-manual: working-projects: properly highlight code code-blocks There are some blocks containing C code so highlight them as C with the "c" lexer. For autotool'ed files, there's no lexer available so "none" will do just fine. Finally, there's one Makefile code block so highlight it with the "makefile" lexer. Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 04c883216 --------------------------- sdk-manual: fix improper indent of general form of tarball installer scripts poky-glibc-host_system-image_type-arch-toolchain-ext-release_version.sh is in a code-block and thus requires to be indented (it is). Sphinx knows a block has ended when the indentation of the next line is smaller than in the block, otherwise it's all in the same block. The explanation of the replaceable parts in the tarball installer script filename should be text and not part of the block, so let's do that. While at it, reformat by putting each replaceable part in a bullet list, highlight the replaceable part or what it can replaced with as well as putting the replacement on the same line instead of making it a quote by having them indented in the bullet list. Finally, replace the literal block (::) with a parsed-literal block so we can highlight the placeholders appropriately. Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: c6039ce11 --------------------------- sdk-manual: fix incorrect highlight language for console code-blocks When unspecified in conf.py via the highlight_language variable (and highlight variable for code-blocks), the lexer used for literal blocks is "default" which tries to highlight the block as Python code. These blocks aren't Python but simple command lines prefixed by a prompt so let's use the "console" lexer to properly highlight. Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard --------------------------- Commit: 72354c270 --------------------------- sdk-manual: fix incorrect highlight language for text code-blocks When unspecified in conf.py via the highlight_language variable (and highlight variable for code-blocks), the lexer used for literal blocks is "default" which tries to highlight the block as Python code. These blocks aren't Python but either simply the console output or a file name (which we should probably just double tick-quote instead), so let's render those blocks with the "text" lexer. Signed-off-by: Quentin Schulz Signed-off-by: Antonin Godard