Changes to bluetooth mesh for k_work API #33782
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
github-actions
bot
added
area: API
18 tasks
jhedberg
reviewed
Mar 29, 2021
subsys/bluetooth/mesh/cfg_srv.c
Outdated
| /* This looks really wrong: the variable says the value is in | ||
| * ms, but the assigned value is ms / 100. | ||
| */ | ||
| timeout_ms = k_delayed_work_remaining_get(&frnd->timer) / 100; |
There was a problem hiding this comment.
Based on the Mesh spec "The PollTimeout timer value in units of 100 milliseconds", i.e. the variable name is wrong here but the code does the right thing.
pabigot
force-pushed
the
nordic/20210329a
branch
from
b2595d3 to
ebd9e4a
Compare
|
Mistakenly taken out of draft, as this is not intended to be merged as-is. It's intended to be adapted by the maintainers of bluetooth who should verify and address the documented concerns. |
trond-snekvik
added a commit
to trond-snekvik/zephyr
that referenced
this pull request
Apr 15, 2021
Changes lpn_timeout_get behavior in the config server to report the configured LPN timeout, instead of the currently remaining timeout time. According to the Bluetooth Mesh Profile specification, section 4.2.21, the PollTimeout list is a list of the PollTimeout timer values, and according to table 4.32 in this section, values 1-9 are prohibited. Although this is not explicitly stated, this indicates that the PollTimeout value is the configured poll timeout time - not the time remaining until the timeout value expires. This patch changes the implementation to reflect this. Split out from zephyrproject-rtos#33782. Signed-off-by: Trond Einar Snekvik <[email protected]>
trond-snekvik
added a commit
to trond-snekvik/zephyr
that referenced
this pull request
Apr 15, 2021
Periodic publication would previously build and send the first publication inside the bt_mesh_model_pub() function, before cancelling and rescheduling the next publication. The timer handler would only handle retransmissions, and would abandon the rest of the publication event if one of the packets failed to send. This design has three issues: - If the initial timer cancel fails, the publication would interfer with the periodic publication management, which might skip an event or send too many packets. - If any of the messages fail to publish, the full publication event would be abandoned. This is not predictable or expected from the API. - bt_mesh_model_pub() required 384 bytes of stack to build the message, which has to be factored into all calling threads. This patch moves all transmission into the publication timer by replacing k_work_cancel with a single k_work_reschedule(K_NO_WAIT). It also changes the error recovery behavior to attempt to finish the full publication event even if some of the transmissions fail. Split out from zephyrproject-rtos#33782. Signed-off-by: Trond Einar Snekvik <[email protected]>
trond-snekvik
added a commit
to trond-snekvik/zephyr
that referenced
this pull request
Apr 15, 2021
Moves the beacon_enabled check in the beacon work handler to check the beacon flag before sending anything, in case a cancel call fails. Split out from zephyrproject-rtos#33782. Signed-off-by: Trond Einar Snekvik <[email protected]>
trond-snekvik
force-pushed
the
nordic/20210329a
branch
from
fa53ed6 to
3cace1b
Compare
Friend structure allocaction logic is implemented over and over throughout the friend module. Move it into a static utility function for readability. Signed-off-by: Trond Einar Snekvik <[email protected]>
Switch to the new API in friend, net and main. Signed-off-by: Peter Bigot <[email protected]> Signed-off-by: Trond Einar Snekvik <[email protected]>
Switch to the new API in Mesh's extended advertising handler. Signed-off-by: Peter Bigot <[email protected]> Signed-off-by: Trond Einar Snekvik <[email protected]>
Switch to the new API. Consolidates reliable sending logic for the first transmission and the retransmit into one. Adds check for link active in protocol timeout. Signed-off-by: Peter Bigot <[email protected]> Signed-off-by: Trond Einar Snekvik <[email protected]>
Switch to the new API. Adds a link check to the protocol timeout to ensure the link is still active. Signed-off-by: Peter Bigot <[email protected]> Signed-off-by: Trond Einar Snekvik <[email protected]>
Switch to the new API. Adds check for a pending buffer in the SAR timeout handler. Signed-off-by: Peter Bigot <[email protected]> Signed-off-by: Trond Einar Snekvik <[email protected]>
Switch to the new API. Adds early exits for the ack and retransmit timers, and replaces a remaining_time() + submit() call with schedule(). Signed-off-by: Peter Bigot <[email protected]> Signed-off-by: Trond Einar Snekvik <[email protected]>
trond-snekvik
force-pushed
the
nordic/20210329a
branch
from
3cace1b to
5131355
Compare
carlescufi
changed the title
trond-snekvik
approved these changes
Apr 22, 2021
jhedberg
approved these changes
Apr 28, 2021
| continue; | ||
| } | ||
|
|
||
| BT_DBG("%u bytes: %s", buf->len, bt_hex(buf->data, buf->len)); |
There was a problem hiding this comment.
This should not be here will cause failed message should be part of the function below
| bt_mesh_adv_send(buf, &buf_sent_cb, NULL); | ||
| } | ||
| } | ||
| send_reliable(); |
There was a problem hiding this comment.
Needs to be removed.
| } else { | ||
| bt_mesh_adv_send(buf, &buf_sent_cb, NULL); | ||
| } | ||
| } |
There was a problem hiding this comment.
Needs to be reinstated due to not closing link completely. Most notably when using the mesh provisioning functionality.
The first node works correctly then the prov_link flag remains "LINK_ACTIVE"
|
@WilliamGFish Note that this PR has already been merged; if your comments are still relevant you need to an open an issue for them. |
Shall do when I get a chance... |
|
@WilliamGFish I'm able to reproduce these issues, and will submit a patch tomorrow, along with a test that covers it. Thanks for reporting. |
trond-snekvik
added a commit
to trond-snekvik/zephyr
that referenced
this pull request
May 18, 2021
The send_reliable function was reused in multiple places as part of the k_delayed_work changes for Bluetooth Mesh in zephyrproject-rtos#33782. This function contains a line that resets the start timer, causing prov_retransmit to continously move the goal post for when to give up sending. Extract this line out of the send_reliable function, and put it along with the other link.tx initialization in bearer_ctl_send and prov_send_adv. Signed-off-by: Trond Einar Snekvik <[email protected]>
jhedberg
pushed a commit
that referenced
this pull request
May 19, 2021
The send_reliable function was reused in multiple places as part of the k_delayed_work changes for Bluetooth Mesh in #33782. This function contains a line that resets the start timer, causing prov_retransmit to continously move the goal post for when to give up sending. Extract this line out of the send_reliable function, and put it along with the other link.tx initialization in bearer_ctl_send and prov_send_adv. Signed-off-by: Trond Einar Snekvik <[email protected]>
ioannisg
pushed a commit
to ioannisg/zephyr
that referenced
this pull request
Jun 7, 2021
...retransmit The send_reliable function was reused in multiple places as part of the k_delayed_work changes for Bluetooth Mesh in zephyrproject-rtos#33782. This function contains a line that resets the start timer, causing prov_retransmit to continously move the goal post for when to give up sending. Extract this line out of the send_reliable function, and put it along with the other link.tx initialization in bearer_ctl_send and prov_send_adv. (cherry picked from commit 85ad497) Signed-off-by: Trond Einar Snekvik <[email protected]>
jori-nordic
pushed a commit
to jori-nordic/zephyr
that referenced
this pull request
Jun 15, 2021
...retransmit The send_reliable function was reused in multiple places as part of the k_delayed_work changes for Bluetooth Mesh in zephyrproject-rtos#33782. This function contains a line that resets the start timer, causing prov_retransmit to continously move the goal post for when to give up sending. Extract this line out of the send_reliable function, and put it along with the other link.tx initialization in bearer_ctl_send and prov_send_adv. (cherry picked from commit 85ad497) Signed-off-by: Trond Einar Snekvik <[email protected]> (cherry picked from commit 861adbe)
trond-snekvik
added a commit
to trond-snekvik/zephyr
that referenced
this pull request
Jul 13, 2021
Adds a BabbleSim test for provisioning multiple devices in a row. This scenario had a regression in zephyrproject-rtos#33782, which is fixed in zephyrproject-rtos#35405. Signed-off-by: Trond Einar Snekvik <[email protected]>
jhedberg
pushed a commit
that referenced
this pull request
Jul 13, 2021
Adds a BabbleSim test for provisioning multiple devices in a row. This scenario had a regression in #33782, which is fixed in #35405. Signed-off-by: Trond Einar Snekvik <[email protected]>
DineshDK03
added a commit
to DineshDK03/zephyr
that referenced
this pull request
Sep 3, 2021
Signed-off-by: Dinesh Kumar K <[email protected]> drivers: display: test to clear display Signed-off-by: Dinesh Kumar K <[email protected]> drivers: display: EK79652 display clear alone working Signed-off-by: Dinesh Kumar K <[email protected]> common: Add headers to support C++ style includes When compiling existing libraries that are difficult to change, these headers simplify the library's integration. This specifically was the agreed upon fix for trying to compile Android's CHRE as a subsystem. Making changes to the Android repo to use the C style includes would be very difficult and would likely take a very long time. Signed-off-by: Yuval Peress <[email protected]> tests: net: coap: Convert unit tests to ztest API Convert CoAP unit tests to ztest API. Additionally, replace dynamic memory allocation with a static one, to prevent memory leaks with ztest. Signed-off-by: Robert Lubos <[email protected]> Bluetooth: host: Move bt_data_parse to hci_core.c Move the function from scan.c to hci_core.c. When in scan.c, the function is only available if CONFIG_BT_OBSERVER was enabled. Since the function can be used in other scenarioes where we need to parse LTV data, it has been moved to a more generic place. hci_core.c might not be the ideal place, but it is where most other common bluetooth functions are located. Signed-off-by: Emil Gydesen <[email protected]> Bluetooth: Iso: Do not remove iso data path on disconnect We previously removed the iso data path when the iso channel disconnected. Since the iso data path is unique for a given iso channel (by handle), it does not make sense to remove it for a disconnected channel, as the channel is, in a sense, not existing anymore. This update is to better comply with the bluetooth core spec, and to avoid getting errors from the controller on disconnect. Rather than removing the implementation of being able to remove the data path, the function was made non-static and moved to the internal header file, in case we ever want to use it. This should not affect compile size. Signed-off-by: Emil Gydesen <[email protected]> ARC: MWDT: don't use compact C library -Hcl option instructs linker to use MetaWare C Compactlib. According to MWDT documentation "Compactlib is not thread-safe" So, let's use MetaWare C library instead which is thread-safe. Signed-off-by: Evgeniy Paltsev <[email protected]> dts/arm: STM32: Adding DMA support for stm32l5 in dtsi This commit adds dma1, dma2, dmamux support in dtsi for stm32l5 series. This commit also fixes a compilation warning due to the usage of upper case in "reg = <0x4000F400 0x400>". Signed-off-by: Krishna Mohan Dani <[email protected]> boards/arm: stm32l562e_dk: Updating the documentation This commit updates the documentation on DMA for stm32l562e_dk platform. Signed-off-by: Krishna Mohan Dani <[email protected]> boards/arm: nucleo_l552ze_q: Updating the documentation This commit updates the documentation on DMA for nucleo_l552ze_q platform. Signed-off-by: Krishna Mohan Dani <[email protected]> boards/arm: stm32l562e_dk: Enabling usart3 This commit enables usart3 channel on stm32l562e_dk platform. This would also be used to test usart in dma mode with tests/drivers/uart/uart_async_api. Signed-off-by: Krishna Mohan Dani <[email protected]> boards/arm: nucleo_l552ze_q: Enabling usart3 This commit enables usart3 channel on nucleo_l552ze_q platform. This would also be used to test usart in dma mode with tests/drivers/uart/uart_async_api. Signed-off-by: Krishna Mohan Dani <[email protected]> boards/arm: nucleo_l552ze_q: Enabling spi1 This commit enables spi1 in dtsi and updates the documentation for nucleo_l552ze_q platform. Signed-off-by: Krishna Mohan Dani <[email protected]> drivers/dma: STM32: Initialising DMA_InitStruct with 0 for stm32l5xx In stm32l5xx soc, the LL_DMA_InitTypeDef has 2 more fields (DoubleBufferMode and TargetMemInDoubleBufferMode) that must be initialised with 0 else the configuration is wrong and gives wrong values to the LL_DMA_Init function. Due to this the test tests/drivers/dma/loop_transfer too would fail. Signed-off-by: Krishna Mohan Dani <[email protected]> tests/drivers: stm32l562e_dk: Configures DMA for chan_blen_transfer test This commit configures the required DMA parameters to run tests/drivers/dma/chan_blen_transfer test on stm32l562e_dk platform. This has been tested and is working as expected. Signed-off-by: Krishna Mohan Dani <[email protected]> tests/drivers: stm32l562e_dk: Configures DMA to run loop_transfer test. This commit configures the required DMA parameters to run tests/drivers/dma/loop_transfer test on stm32l562e_dk platform. This has been tested and is working as expected. Signed-off-by: Krishna Mohan Dani <[email protected]> tests/drivers: nucleo_l552ze_q: Configures DMA for chan_blen_transfer test This commit configures the DMA to run tests/drivers/dma/chan_blen_transfer test on nucleo_l552ze_q platform. Signed-off-by: Krishna Mohan Dani <[email protected]> tests/drivers: nucleo_l552ze_q: Configures DMA to run loop_transfer test This commit configures DMA to run tests/drivers/dma/loop_transfer test on nucleo_l552ze_q platform. Signed-off-by: Krishna Mohan Dani <[email protected]> tests/drivers: nucleo_l552ze_q: Configuration for spi_loopback test This commit configures spi, dma & dmamux to run spi_loopback test on nucleo_l552ze_q platform. The tx & rx pin of spi1 should be shorted before running spi_loopback. Signed-off-by: Krishna Mohan Dani <[email protected]> tests/drivers: stm32l562e_dk: uart_async_api test This commit configures usart3 to use dma in uart_async_api test on stm32l562e_dk platform. Short pin2(PC10) & pin 3 (PC11) of usart3 in CN4 connector on stm32l562e_dk platform. Signed-off-by: Krishna Mohan Dani <[email protected]> tests/drivers: nucleo_l552ze_q: uart_async_api test This commit configures usart3 to use dma in uart_async_api test on nucleo_l552ze_q platform. Signed-off-by: Krishna Mohan Dani <[email protected]> boards: arm: bl5340_dvk: Rename NS target This commit update the NS board variant from `bl5340_dvk_cpuappns` to `bl5340_dvk_cpuapp_ns` to maintain consistency across zephyr. Signed-off-by: Kevin Townsend <[email protected]> boards: arm: mps2_an521: Rename NS target This commit update the NS board variant from `mps2_an521_nonsecure` to `mps2_an521_ns` to maintain consistency across zephyr. Signed-off-by: Kevin Townsend <[email protected]> boards: arm: v2m_musca_b1: Rename NS target This commit update the NS board variant from `v2m_musca_b1_nonsecure` to `v2m_musca_b1_ns` to maintain consistency across zephyr. Signed-off-by: Kevin Townsend <[email protected]> boards: arm: v2m_musca_s1: Rename NS target This commit update the NS board variant from `v2m_musca_s1_nonsecure` to `v2m_musca_s1_ns` to maintain consistency across zephyr. Signed-off-by: Kevin Townsend <[email protected]> boards: arm: nrf5340dk_nrf5340: Rename NS target This commit updates the NS board variant from `nrf5340dk_nrf5340_cpuappns` to `nrf5340dk_nrf5340_cpuapp_ns` to maintain consistency across zephyr. Signed-off-by: Kevin Townsend <[email protected]> boards: arm: nrf9160dk_nrf9160: Rename NS target This commit updates the NS board variant from `nrf9160dk_nrf9160ns` to `nrf9160dk_nrf9160_ns` to maintain consistency across zephyr. Signed-off-by: Kevin Townsend <[email protected]> Bluetooth: controller: Fix PKT_AC_US macro Advertising channel packets do not have MIC, there's no need to have extra parameter which always has to be set to 0 anyway. Signed-off-by: Andrzej Kaczmarek <[email protected]> Bluetooth: controller: Add helper to allocate new pdu_adv This adds helper to always allocate advertising PDU either from memory pool or pdu_free queue and does not reuse existing PDU in adv_pdu. Signed-off-by: Andrzej Kaczmarek <[email protected]> Bluetooth: controller: Optimize lll_adv_pdu_and_extra_data_alloc This function is the same as lll_adv_pdu_alloc except it also allocates extra data at the end - it can just use lll_adv_pdu_alloc call to avoid extensive c&p. Signed-off-by: Andrzej Kaczmarek <[email protected]> Bluetooth: controller: Add support for linked adv PDUs This adds support to allow advertising PDUs to be linked which is required to send advertising trains, i.e. AUX_CHAIN_IND. PDUs are linked with a simple single-linked list, the pointer to next PDU is stored at the very end of PDU buffer. This prevents it from being overwritten if PDU is modified and allows for build time offset calculation. There are few helpers added to make handling easier, e.g.: - get next linked PDU - get last linked PDU - link one PDU to another - link PDU at the end - release all linked PDUs (except for 1st) Signed-off-by: Andrzej Kaczmarek <[email protected]> Bluetooth: controller: Enable back-to-back chaining for periodic adv This enables chaining ota for periodic advertising. AUX_CHAIN_IND PDUs will be sent automatically if AuxPtr is detected in preceding PDU. AuxPtr offset is always set to achieve minimal required frame spacing, i.e. 300us (T_mafs). AuxPtr in all PDUs in advertising train are updated on enqueue since PDU spacing is already known at that time so we do not need to waste time in LLL. Signed-off-by: Andrzej Kaczmarek <[email protected]> Bluetooth: controller: Add helper to initialize adv PDU from scratch This adds adv_sync_pdu_init helper which initializes pdu_adv buffer with contents of AUX_SYNC_IND/AUX_CHAIN_IND PDU. Extended header flags can be specified to reserve required space for corresponding fields if necessary. Signed-off-by: Andrzej Kaczmarek <[email protected]> Bluetooth: controller: Add initial support to update AD in chain This adds some initial support to update AD in chain. We still only support placing AD in 1st PDU, but this will properly copy any linked PDUs that may be added due to e.g. CTEInfo present. Signed-off-by: Andrzej Kaczmarek <[email protected]> Bluetooth: controller: Add helper to update CTEInfo This adds simple helper to update CTEInfo. It assumes proper periodic adv PDU as input to simplify code. Signed-off-by: Andrzej Kaczmarek <[email protected]> Bluetooth: controller: Remove adv_sync_hdr_set_clear adv_sync_hdr_set_clear was just wrapped by ull_adv_sync_pdu_set_clear so we can merge both into single function. Signed-off-by: Andrzej Kaczmarek <[email protected]> flash_page_layout: refactor flash_page_get_info refactor flash_page_get_info to simplify and to avoid using mixing the usage of an off_t (offs) and an uint32_t (page_index). Signed-off-by: Laczen JMS <[email protected]> Bluetooth: Controller: Warn when building experimental features Add CMake warning message when building experimental features like Advertising Extensions. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: controller: Add/Remove ACAD on create/terminate BIG Added implementation to add/remove ACAD field in the common extended header format of the periodic advertising PDU on create/terminate BIG. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: controller: Use set/clear to add/remove ACAD Use the set/clear function to modify the common extended header format in the PDU to add/remove ACAD. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: controller: Refactor sync_info population Based on review comments, refactor out sync_info population to be performed by the caller of the function that prepares the extended advertising PDU. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> tests: Bluetooth: bsim: ISO test with ACAD in periodic advertising Updated ISO test to demonstrate ACAD field in periodic advertising PDUs. Here, test changing the periodic advertising data. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> tests: Bluetooth: bsim: name adv_handle for advertising handle Use the name adv_handle instead of index to store advertising handle. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Minor fix debug message with channel map mask Minor fix to channel map mask used in debug message to print sync info fields. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Use macros for SCA and Channel Map access Use macros to access SCA and Channel Map fields in the Sync Info structure in advertising PDUs. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> boards: mxrt685: Update to documentation Move the I2S_DATAIN pin description so it is grouped with I2S pins Signed-off-by: Mahesh Mahadevan <[email protected]> dts: bindings: jedec,jesd216: remove deprecated 'has-be32k' prop This property has been marked as deprecated in 2.5.0 and was not actually used for even longer time. Signed-off-by: Marcin Niestroj <[email protected]> drivers: pwm: nrf_sw: drop deprecated 'timer-instance' DT prop This property has been marked as deprecated in 2.5.0 and was replaced by 'generator' property. Signed-off-by: Marcin Niestroj <[email protected]> Bluetooth: Controller: Fix Periodic Sync setup with invalid channels Fix Periodic Synchronization setup when handling invalid number of channel count in Periodic Advertising's Sync Info structure. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> soc: npcx: Add soc log register NPCX power.c use LOG_MODULE_DECLARE(soc), but NPCX chip doesn't register soc log module. This CL register soc log in soc.c to fix NPCX build error for power management & log system. Signed-off-by: Wealian Liao <[email protected]> driver: clock: npcx: add a option to generate LFCLK via XTSOC This commit adds a new Kconfig option CLOCK_CONTROL_NPCX_EXTERNAL_SRC. With this option enabled, the internal 32.768 KHz clock (LFCLK) is generated by the on-chip Crystal Oscillator (XTOSC). Otherwise, the LFCLK is generated by the Low-Frequency Clock Generator (LFCG). Signed-off-by: Jun Lin <[email protected]> doc, cpp: Update CXX support to show exception support Updates CXX support documentation to reflect exception support added in fixes for #32448 and #35772. Signed-off-by: Lauren Murphy <[email protected]> drivers: pinmux: stm32: fix name clashes with G4 series There is a name clash when using G4 series LL TIM driver depending on the inclusion order of the LL TIM and pinmux headers. If the LL headers are included after pinmux is included, AF1 and AF2 definitions used by pinmux clash with the AF1 and AF2 TIMx register names. In order to solve this problem with minimum impact, the following has been done: 1. Prefix the AFx and ANALOG definitions with STM32 2. In order to avoid changing all *-pinctrl.dtsi files, the STM32_PINMUX macro contatenates STM32_ with the provided mode. Signed-off-by: Gerard Marull-Paretas <[email protected]> driver: flexcan: flexcan support for the rt1170 add the flexcan3 support for the rt1170 Signed-off-by: Crist Xu <[email protected]> boards: fixing typo from MX24R64 to MX25R64 Comments in .dts files with MX25R64 and QSPI contained a typo Signed-off-by: Akash Patel <[email protected]> ci: reduce disk usage for daily build Avoid some cases of running out of disk space in the daily build. Add setting -M option to remove artifacts as we build. Signed-off-by: Kumar Gala <[email protected]> kernel: native_posix: Don't run global C++ constructors On the native_posix board global object constructors are run by the underlying OS runtime init prior to Zephyr kernel init. Thus Zephyr should not run global object constructors a second time. Doing so breaks application behavior that relies on global constructors doing work that must be done only once. See bug #36858 for more information. Signed-off-by: David Palchak <[email protected]> drivers: edac: Fix PCIe Kconfig dependency It's not possible to build the IBECC driver without PCIe support. Signed-off-by: Johan Hedberg <[email protected]> cmake: extended zephyr_library_amend description with an extra example Add some more example to the description of zephyr_library_amend(). This should help users to get input of the extra possibilities that this function provides. See: #35770 Signed-off-by: Torsten Rasmussen <[email protected]> Signed-off-by: Andrzej Głąbek <[email protected]> net: ip: Fix assertion failure when tcp_send_data() When tcp_send_data() is called to resend data, but there is no data to resend, zero length packet is allocated and NULL net_buf is passed to net_buf_frag_insert() in which assertion fails. Fixes #36578 Signed-off-by: Chih Hung Yu <[email protected]> net: lib: sockets: Fix assertion failure when zsock_close() When zsock_close() is called, socket is freed before the mutex for the socket is unlocked. If the freed socket is given to another thread immediately, the mutex for the socket will be initialized by the new socket owner, while the mutex is still locked by the thread calling zosck_close(). Fixes #36568 Signed-off-by: Chih Hung Yu <[email protected]> Bluetooth: Mesh: Tests of BLE Mesh beacon flags management Tests check transmission, receiving beacons in situations of IV update (with test mode on) and key refresh procedures. Signed-off-by: Aleksandr Khromykh <[email protected]> subsys/mgmt/mcumgr: Add SMP Zephyr specific storage erase command The commit adds support for Zephyr basic mgmt group to mcumgr. The first command added to the group is storage erase command. Authored-by: Sigvart Hovland <[email protected]> Signed-off-by: Dominik Ermel <[email protected]> Bluetooth: SDP: judge the buf->len before sys_get_be16 There may be less than 2 bytes in buf before calling sys_get_be16. Signed-off-by: Mark Wang <[email protected]> tests: test_build: Exclude some NS platforms from debug builds The nrf9160dk_nrf9160_ns and nrf5340dk_nrf5340_cpuapp_ns don't have enough space for debug builds as configured so excluded them from this specific test. Signed-off-by: Kumar Gala <[email protected]> boards: arm64: fvp_baser_aemv8r: select CACHE_MANAGEMENT fvp_baser_aemv8r supports CACHE_MANAGEMENT too and CACHE_MANAGEMENT should also be selected. Signed-off-by: Huifeng Zhang <[email protected]> boards: arm64: fvp_baser_aemv8r: modifing dts for enabling SMP Add psci and more cpu nodes into fvp_baser_aemv8r.dts. The purpose of it is perparing to support SMP. Signed-off-by: Huifeng Zhang <[email protected]> soc: arm64: arm: fvp_aemv8r: Init VMPIDR_EL2 in el2 plat init. Add strong definition z_arm64_el2_plat_init() and it is controlled by CONFIG_SOC_FVP_AEMV8R_EL2_INIT. VMPIDR_EL2 must be set manually on EL2. The purpose of VMPIDR_EL2 is that holds the value of the Virtualization Multiprocessor ID and This is the value returned by EL1 reads of MPIDR_EL1 Signed-off-by: Huifeng Zhang <[email protected]> arch: arm64: Unify the initialization of MMU and MPU Because MMU and MPU should not be enabled together and they provide the same functionalities. Signed-off-by: Huifeng Zhang <[email protected]> driver: pm_cpu_ops: change PM_CPU_OPS_PSCI's dependency Armv8-A and Armv8-R both support PSCI. So PM_CPU_OPS_PSCI's dependency should be "ARM64" rather than "ARMV8-A". Signed-off-by: Huifeng Zhang <[email protected]> soc: arm64: arm: fvp_base_r: define a strong pm_cpu_on() function With this patch, zephyr can enable SMP directly. Otherwise zephyr needs TB-R to provide psci function. Signed-off-by: Huifeng Zhang <[email protected]> boards: arm64: fvp_baser_aemv8_r: change FVP startup parameters Eliminate GICv3 warning messages printed by FVP_BaseR_AEMv8R. And "cluster0.NUM_CORES" depends on CONFIG_MP_NUM_CPUS now. Signed-off-by: Huifeng Zhang <[email protected]> boards: arm64: add fvp_baser_aemv8r_smp Support SMP for fvp_baser_aemv8r. Signed-off-by: Huifeng Zhang <[email protected]> arch: arm64: select SCHED_IPI_SUPPORTED for Armv8_R Armv8_R supports IPI Signed-off-by: Huifeng Zhang <[email protected]> soc: boards: arm: Refactor i.MX RT600 zephyr,flash usage Refactors the i.MX RT600 SoC series to be more consistent with the i.MX RT10xx SoC series by choosing a child node (external flash device) of the FlexSPI bus for zephyr,flash. Signed-off-by: Maureen Helm <[email protected]> boards: arm: Add mcuboot flash partitions to mimxrt685_evk Adds flash partitions and chosen nodes to the mimxrt685_evk device tree to support mcuboot on the external octal SPI flash. This flash is rated for 100K minimum program-erase cycles per sector, therefore this partition configuration supports approximately 100K / (24576/8128) = 33073 upgrades. Tested with samples/subsys/mgmt/mcumgr/smp_svr. The image swap takes about a minute and a half to complete. Signed-off-by: Maureen Helm <[email protected]> soc: boards: Disable i.MX RT6xx boot header in chainloaded applications The bootloader itself contains the i.MX RT6xx boot header, so we don't need to duplicate it when building chainloaded applications. Signed-off-by: Maureen Helm <[email protected]> boards: imxrt685: Update pinmux setting to remove the const keyword This would save some space by using a local variable Signed-off-by: Mahesh Mahadevan <[email protected]> pm: device: remove runtime PM callback The callback used by the device runtime PM can be easily replaced by a simple state set after calling the state set/get calls. Broadcast logic is simplified too, leading to the same previous behavior. Since this is the only place where this callback was used, it can now be removed from all devices and so pm_control callback signature simplified. Signed-off-by: Gerard Marull-Paretas <[email protected]> pm: remove redundant callback usage the device PM callback is not used anymore by the device PM subsystem, so remove it from all drivers/tests using it. Signed-off-by: Gerard Marull-Paretas <[email protected]> pm: remove callback from control function The callback is not used anymore, so just delete it from the pm_control callback signature. Signed-off-by: Gerard Marull-Paretas <[email protected]> dts: stm32wl: add definitions for further stm32wl socs. This commit adds dt support for stm32wle4, and stm32wle5 single core, as well as stm32wl54 dual core socs. Signed-off-by: Thomas Stranger <[email protected]> soc/arm: add support for additonal stm32wl socs This commit adds support for stm32wle4xx, stm32wle5xx single core socs, as well as stm32wl54 dual core soc. Signed-off-by: Thomas Stranger <[email protected]> drivers/clock_control: stm32wl set cpu2 prescaler only for STM32WL5X To support single core stm32wlex series, cpu2 prescaler is set only on dual core soc variants. Signed-off-by: Thomas Stranger <[email protected]> dts: arm: introduce seeed lora-e5 module This commit adds the dts definitons for the seeed lora-e5 module. Additionally I add myself as codeowner for the new dts/arm/seeed directory. This module packages a stm32wle5jc Sub-GHz Wireless Soc, together with a 32MHz TCXO, a 32.768KHz crystal oscillator, and power and RF circuitry. With the introduction of lora support definitions for the radio will be added in a future commit. Signed-off-by: Thomas Stranger <[email protected]> boards: add seeed studio LoRa-E5 Dev board This commit adds support for the seeed studio LoRa-E5 Dev board, which is powered by a module based on stm32wle5jc soc. Signed-off-by: Thomas Stranger <[email protected]> drivers: entropy: add support for stm32wl and stm32g0 This commit adds entropy support for stm32wl and stm32g0. Pll is used as clock source and has to be enabled, other clock sources are not supported at the moment. Signed-off-by: Thomas Stranger <[email protected]> dts: stm32g0: add definitions for socs with rng (and aes) support. This commit adds support for stm32g0 socs with integrated rng and hw aes acceleratior, which are stm32g041, stm32g061, stm32g081, and stm32g0c1. It also adds the definitions for the rng peripheral and sets it as chosen zephyr,entropy source. Signed-off-by: Thomas Stranger <[email protected]> soc/arm: add support for stm32g0 socs with (hw aes and) rng support This commit introduces g041, g061, g081, and g0c1 socs in kconfig. Signed-off-by: Thomas Stranger <[email protected]> dts: stm32wl: add definitions for rng peripheral This commit adds the dt node for rng to the stm32wl series and sets it as chosen zephyr,entropy source. Signed-off-by: Thomas Stranger <[email protected]> boards: nucleo_wl55jc enable rng This commit enables entropy support for nucleo_wl55jc. Additionally it sets the PLL Q divider to 2, which was not set in the board dts before. Signed-off-by: Thomas Stranger <[email protected]> logging: log_backend_net: Add support for logging v2 Added support for logging v2 backend API in net backend. Signed-off-by: Krzysztof Chruscinski <[email protected]> subsys/cpp: update tag for C++2a to C++20 C++20 was tagged and released in December of 2020. It is no longer forth coming. Signed-off-by: Dan Kalowsky <[email protected]> subsys/cpp: add support for C++2b Add the ability to enable support for C++2B (aka 23) to the build system. Signed-off-by: Dan Kalowsky <[email protected]> drivers: uart: Add "Interrupt driven" to R-Car This commit add the support of "Interrupt driven UART" to the R-Car UART driver and enable it in the related Kconfig. The driver is supporting nearly all the methods that are described in the "Interrupt driven UART" part of the uart_driver_api. This new version of the driver has been tested on H3ULCB board by running "uart_basic_api" test suite. Signed-off-by: Aymeric Aillet <[email protected]> boards: rcar_h3ulcb: Enable Interrupt driven UART This commit enables "Interrupt driven UART" for the renesas R-Car H3ULCB board. Signed-off-by: Aymeric Aillet <[email protected]> driver: gpio: add initial support for synopsys creg gpio Add single-register MMIO GPIO driver for complex cases where only several fields in register belong to GPIO lines and each GPIO line owns a field with different length and on/off value. Such CREG GPIOs are used in Synopsys em_starterkit and HSDK boards. Signed-off-by: Watson Zeng <[email protected]> boards: arc: hsdk: add creg_gpio driver support Add creg_gpio driver support for hsdk board. Signed-off-by: Watson Zeng <[email protected]> boards: arc: em_starterkit: add creg_gpio driver support Add creg_gpio driver support for em_starterkit board. Signed-off-by: Watson Zeng <[email protected]> west: update west.yml to get updated hal_stm32 modules This commit changes the entry referencing the hal_stm32 module into west.yml. This brings the fixes to build issues when stm32 asserts are enabled. Signed-off-by: Krishna Mohan Dani <[email protected]> boards: acrn_ehl_crb: Add PCIe to defconfig A lot of basic functionality depends on PCIe, so enable it in the defconfig. Signed-off-by: Johan Hedberg <[email protected]> boards: acrn_ehl_crb: Add ibecc DTS node We need an ibecc DTS node so that the EDAC driver can be used. Signed-off-by: Johan Hedberg <[email protected]> soc: it8xxx2: introduce SOC_IT8XXX2_PLL_FLASH_48M option Enable SOC_IT8XXX2_PLL_FLASH_48M at default to reduce latency of fetching code from flash. Signed-off-by: Dino Li <[email protected]> logging: Shell commands available when shell acts as log backend Some commands shell not be present when shell is not acting as a log backend. Use of them lead to crash in that case. Signed-off-by: Krzysztof Chruscinski <[email protected]> tests: add min_ram to test applications add min_ram to some test applications as we found below platforms have size issues TWR_KE18F and FRDM_KL25Z Signed-off-by: Hake Huang <[email protected]> docs: Added documentation for native posix mode Added documentation for native posix mode. Signed-off-by: Lingao Meng <[email protected]> iii Signed-off-by: Lingao Meng <[email protected]> Bluetooth: Mesh: Align capitalization for BT mesh Align the capitalization of the term "Bluetooth Mesh" to Bluetooth mesh" in the documentation. This is done to to match the new updated naming convention done in Bluetooth SIG. In the upcoming spec versions, it its used "Bluetooth mesh" with the lower case convention. Signed-off-by: Ingar Kulbrandstad <[email protected]> doc: extensions: doxyrunner: process Doxygen output This change will process Doxygen output and will map it to the Sphinx logger. Things like errors and warnings will be mapped to actual Sphinx logger error and warnings. In practice this means that when Doxygen throws a warning and Sphinx is run in "-W" (warning as error) mode, the build will fail. It also has some other advantages such as the possibility of filtering issues using the warnings_filter extension. It is also expected that CI errors not being displayed issue is fixed with this change. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: extensions: doxyrunner: handle quiet mode QUIET flag is now overriden according to the `doxyrunner_silent` configuration value. Signed-off-by: Gerard Marull-Paretas <[email protected]> Bluetooth: Mesh: Test provisioning multiple devices Adds a BabbleSim test for provisioning multiple devices in a row. This scenario had a regression in #33782, which is fixed in #35405. Signed-off-by: Trond Einar Snekvik <[email protected]> Doc: Getting Started: Image fix for PDF generation The `pdftex` utility that is used for PDF documentation generation does not support GIF images. This fix replaces the animated `ReelBoard-Blinky.gif` with an animated PNG (APNG) image `ReelBoard-Blinky.png`. Note that APNGs are backwards compatible with "normal" PNGs. Modern web browsers will display APNGs in the same way as animated GIFs, while any application that can read PNG files can read APNG files successfully. Signed-off-by: Chris Pearson <[email protected]> arch: give the choice "Cache type" a name Give the choice a name so that the soc/board developers can change the default selection in their Kconfig.*. For example: choice CACHE_TYPE default HAS_EXTERNAL_CACHE endchoice There was a similar issue had beed discussed: https://github.com/zephyrproject-rtos/zephyr/issues/6948 Signed-off-by: Dylan Hung <[email protected]> Change-Id: I07c3e78a5243b30912f8e44fa3181fa163016318 cpp: Restore C++2A Kconfig Restore the legacy C++2a Kconfig for compilers that do not support C++20. Signed-off-by: Alexander Wachter <[email protected]> drivers: usb: device: Rename usb_dc_sam to usb_dc_sam_usbhs The SoC driver name is 'USB High-Speed Interface (USBHS)'. This rename from usb_dc_sam to usb_dc_sam_usbhs allowing add others SoC drivers like 'USB Device Port (UDP)' that is found at SAM4S/E variations. Signed-off-by: Gerson Fernando Budke <[email protected]> drivers: ethernet: stm32: add support for disabling auto-negotiation drivers: support disabling auto-negotiation for stm32 eth drivers Signed-off-by: Rup Gajurel <[email protected]> ITE: drivers/i2c: create pinmux phandle to the I2C driver node Create the pinmux phandle to the I2C driver node in the devicetree. When the pinmux_pin_set function in i2c_it8xxx2_init can refer to the setting of this phandle. It is more flexible to use. Signed-off-by: Tim Lin <[email protected]> ITE: driver/i2c: add I2C recovery function Use GPIO output high and low to simulate I2C start and stop conditions to restore i2c to normal. Signed-off-by: Tim Lin <[email protected]> tests: net: socket: select: Relax timing requirements The test ocassionally fails on the mps2_an385 platform in the CI, due to strict timing requirements of the test. Relax the timeouts and acceptable fuzz time a bit, to prevent the failures in the future. Signed-off-by: Robert Lubos <[email protected]> Bluetooth: samples: Add ISO connected benchmark sample Add a sample that measures packet loss in a connected ISO setup. Signed-off-by: Emil Gydesen <[email protected]> tests: kernel: device: Exclude beaglev_starlight_jh7100 We excluded the beaglev_starlight_jh7100 from this test but only did the kernel.device.pm test. We should have excluded the platform from both tests. The beaglev_starlight_jh7100 uses a full 64-bit devicetree map which uses #{address/size}-cells = 2. The device test expects that #{address/size}-cells = 1 so exclude beaglev_starlight_jh7100 from the test. Signed-off-by: Kumar Gala <[email protected]> doc: cmake: remove west check west is currently optional, however, when not found its docs will not be available and the zephyr_module CMake utility will not be able to find modules (unless ZEPHYR_MODULES is used). Other Python dependencies, i.e. the ones listed in the requirements file, are not checked, so do the same for west. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: improve dependencies section Group dependencies all together. Some have been simplified, e.g. Sphinx and others improved, e.g. LaTeX. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: group options Group options all together and make them part of cache. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: remove unused variables Remove some unused variables. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: refactor html targets Refactor HTML targets. The 'html' target is kept as is, 'sphinx-html' equivalent is now 'html-nodeps'. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: cleanup doxygen section Cleanup doxygen section and simplify target. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: refactor pdf build Refactor the pdf build targets. The Sphinx latex build is now invoked by the 'latex' target. The 'pdf' target can be used to build the resulting LaTeX files. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: move environment, paths and functions Move environment, paths and functions before they are used. Helps on readability. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: remove unused version include version.cmake is not being used. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: remove RST_OUT redundant variable DOCS_SRC_DIR is used instead. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: refactor kconfig Refactor the Kconfig target. Mostly formatting in this case. The Kconfig script is also added as a dependency of the CMake configuration step, so that if it is changed, the Kconfig generation step will be re-run. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: refactor devicetree Simplify the devicetree generation. Most of the extra options are not used, so they have been removed to make things easier to maintain. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: add support for clean target Indicate which files have to be cleaned for each target. This feature is only supported for CMake >= 3.15 and is ignored on older CMake versions. Build will still succeed, but the `clean` target will not clean the additional build files. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: move pristine to a separate section Move 'pristine' target to 'others' section. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: make: adjust Makefile Adjust Makefile according to latest CMake changes. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: update build instructions Update build instructions according to latest changed. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: add target comments Add target comments to make build process more user friendly. Signed-off-by: Gerard Marull-Paretas <[email protected]> logging: log_backend_xtensa_sim: Add support for logging v2 Added support for logging v2 backend API in xtensa_sim. Signed-off-by: Krzysztof Chruscinski <[email protected]> logging: log_backend_spinel: Add support for logging v2 Added support for logging v2 backend API in spinel backend. Signed-off-by: Krzysztof Chruscinski <[email protected]> drivers: serial: add a dummy driver for vnd,serial This commit adds a serial dummy driver compatible to vnd,serial. This is needed that devices can access the uart device in tests like tests/drivers/build_all/... . Add myself as codeowner to avoid complicance check failure. Signed-off-by: Thomas Stranger <[email protected]> tests: drivers: build_all: sensor: enable serial driver Enable serial driver such that sensors commuicating via uart can be added to this test(using the vnd,serial driver). Signed-off-by: Thomas Stranger <[email protected]> modules: hal_nxp: replace Kconfig option USB with USB_DEVICE_DRIVER Replace Kconfig option USB with USB_DEVICE_DRIVER. Signed-off-by: Johann Fischer <[email protected]> samples: jesd216: Exclude hifive1 board from sample By default the hifive1 board doesn't enable the SPI controller that the flash is on. As such this test will not build on that platform due to lack of a missing SPI bus controller device. Signed-off-by: Kumar Gala <[email protected]> net: openthread: handle NONE level logs `OT_LOG_LEVEL_NONE` has some uses within OpenThread but it is not hanled in the Zephyr's platform implementation. This commit makes use of those logs as `LOG_LEVEL_ERR` level. Signed-off-by: Eduardo Montoya <[email protected]> soc: boards: Enable Segger RTT/SystemView on mimxrt1170_evk_cm7 Adds a devicetree chosen node to the mimxrt1170_evk_cm7 board to link Segger RTT and SystemView sections in DTCM by default. Enables the AHB clock while the CM7 is sleeping to allow debug access to the TCM. Note that automatic RTT control block detection may not search the DTCM address region, therefore you may need to manually set the RTT control block address or search range in the Segger host tools (SystemView or RTT Viewer). For example, $ JLinkRTTViewer -ra 0x20000000 Tested with: - samples/subsys/shell/shell_module/ - samples/subsys/tracing/ Signed-off-by: Maureen Helm <[email protected]> drivers: usb_dc_mcux_ehci: fix style Fix style, two lines are indented too far. Signed-off-by: Johann Fischer <[email protected]> drivers: usb_dc_mcux_ehci: use K_NO_WAIT in k_heap_alloc() This is called in ISR context and timeout must be set to K_NO_WAIT. Reported-by: Pieter De Gendt <[email protected]> Signed-off-by: Johann Fischer <[email protected]> drivers: display: ili9xxx: use gpio_dt_spec Use the recently introduced struct gpio_dt_spec to store GPIO information and operate with them. Signed-off-by: Gerard Marull-Paretas <[email protected]> drivers: display: ili9xxx: use spi_dt_spec Use the recently introduced struct spi_dt_spec to store SPI information and operate with it. Signed-off-by: Gerard Marull-Paretas <[email protected]> drivers: display: ili9xxx: remove unnecessary casts The config/data casts are not strictly necessary. Furthermore, config was being casted to non-const. Signed-off-by: Gerard Marull-Paretas <[email protected]> dts: arm: nxp: kinetis: ke1xf: add cpu nodelabel Add nodelabel for cpu0 to allow referencing it in the board devicetree. Signed-off-by: Henrik Brix Andersen <[email protected]> soc: arm: nxp: kinetis: ke1xf: determine hw clock cycles/sec from dts Determine the default CONFIG_SYS_CLOCK_HW_CYCLES_PER_SEC from devicetree when using the Arm SysTick hardware timer. When the NXP KE1xF SoC series is using the Arm SysTick as hardware timer, the cycles/second will always be equal to the CPU core clock frequency. Signed-off-by: Henrik Brix Andersen <[email protected]> boards: arm: twr_ke18f: do not enable Arm SysTick driver by default Do not enable the Arm SysTick driver by default. Instead, rely on the default Kconfig settings for the Arm SysTick driver. Signed-off-by: Henrik Brix Andersen <[email protected]> soc: arm: nxp: kinetis: ke1xf: keep the sirc running in low-power mode Keep the Slow Internal Reference Clock (SIRC) running in low-power mode. This allows peripherals that needs to remain operative in low-power mode to use the SIRC as clock source. Signed-off-by: Henrik Brix Andersen <[email protected]> drivers: timer: add NXP LPTMR timer driver Add NXP Kinetis Low Power Timer (LPTMR) OS timer driver shim. Since the LPTMR does not support asynchronous changes to the timer period, only non-tickless mode is supported. Signed-off-by: Henrik Brix Andersen <[email protected]> soc: arm: nxp: kinetis: ke1xf: enable lptmr as timer Enable the NXP Kinetis Low Power Timer (LPTMR) OS timer driver when power management is enabled as the Arm SysTick timer cannot wake up the KE1xF from deep sleep. Signed-off-by: Henrik Brix Andersen <[email protected]> soc: arm: nxp: kinetis: ke1xf: add support for power management Add power management support (runtime-idle and suspend-to-idle) support for the NXP Kinetis KE1xf SoC series. Signed-off-by: Henrik Brix Andersen <[email protected]> dts: arm: nxp: ke1xf: add supported cpu power states Add supported CPU power states for idle, stop, partial stop 1, and partial stop 2. Signed-off-by: Henrik Brix Andersen <[email protected]> boards: arm: twr_ke18f: enable cpu power-states Enable CPU idle and stop power-states on the NXP TWR-KE18F development board. Signed-off-by: Henrik Brix Andersen <[email protected]> tests: pm: power_mgmt_soc: enable testing on twr_ke18f Enable the SoC power management test suite on the NXP TWR-KE18F development board. Signed-off-by: Henrik Brix Andersen <[email protected]> devicetree: Add _VARGS variants to _FOREACH_ marcos `_FOREACH_` macros do not allow the caller to pass additional arguments to the `fn`. A series of `_VARGS` variants have been added that allow the caller to pass arbitrary number of arguments to the `fn`: ``` DT_FOREACH_CHILD_VARGS DT_FOREACH_CHILD_STATUS_OKAY_VARGS DT_FOREACH_PROP_ELEM_VARGS DT_INST_FOREACH_CHILD_VARGS DT_INST_FOREACH_STATUS_OKAY_VARGS DT_INST_FOREACH_PROP_ELEM_VARGS ``` Signed-off-by: Arvin Farahmand <[email protected]> dtlib: add force DT kwarg Modeled after dtc's --force option, the idea is this will try harder and harder over time to produce an object despite malformed input. Signed-off-by: Martí Bolívar <[email protected]> dtlib: allow dangling aliases with DT(..., force=True) As a first step towards being more forgiving on invalid inputs, allow string-valued aliases properties that do not point to valid nodes when the user requests permissiveness. Signed-off-by: Martí Bolívar <[email protected]> tests: lib: mpsc_pbuf: Add test case for max packet allocation Add test for validating max packet allocation. Signed-off-by: Krzysztof Chruscinski <[email protected]> dts: bindings: i2c: add Renesas R-Car I2C Third generation R-Car series products have up to seven I2C bus interfaces conformant with the Philips Semiconductors (now NXP Semiconductors) I2C bus (Inter-IC bus) specification. Signed-off-by: Aymeric Aillet <[email protected]> drivers: i2c: add Renesas R-Car driver This patch add support for I2C on the Renesas R-Car. This I2C hardware block can be found on various Renesas R-Car SoC series. It allows to perfom read and write on I2C buses in an interrupt based way on R-Car Gen3 H3ULCB board. Signed-off-by: Aymeric Aillet <[email protected]> dts: arm: rcar_gen3_cr7: Add I2C support to DTS Add I2C bus initial support to Renesas R-Car SoC series. Both I2C2 & I2C4 buses are supported on R-Car H3 board. Signed-off-by: Aymeric Aillet <[email protected]> boards: arm: rcar_h3ulcb: Enable I2C Enable I2C driver for Renesas R-Car H3 boards. Signed-off-by: Aymeric Aillet <[email protected]> drivers: modem: RSSI measurement for PPP link This PR addresses radio signal stength measurement during and before PPP session. The PR provides provides facility of readout for both +CSQ and +CESQ versions depending upon the modems. This PR follows the idea of rssi readout of PR#35496. Additionally, reliable Cell info update is also ensured. Signed-off-by: Tahir Akram <[email protected]> module: mbedtls: add shell module Add mbedTLS specific shell module, which allows (for now) to show heap allocation statistics. Signed-off-by: Marcin Niestroj <[email protected]> drivers/clock_control: stm32: Fixes around LSE clock Fix macro used in g4 file to enable LSE clock. Then, to avoid no-op configurations, generate an error when MSI Hardware auto calibration is selected but LSE clock is not enabled. Signed-off-by: Erwan Gouriou <[email protected]> soc: arm: stm32l0: set voltage scaling to range1 Default Voltage scaling range selection (range2) doesn't allow to configure Max frequency switch to range1 to match any frequency Signed-off-by: Alexandre Bourdiol <[email protected]> soc: arm: stm32l1: set voltage scaling to range1 Default Voltage scaling range selection (range2) doesn't allow to configure Max frequency switch to range1 to match any frequency Signed-off-by: Alexandre Bourdiol <[email protected]> Bluetooth: Mesh: Add return value for opcode callback ``` 3.7.3.4 Message error procedure When receiving a message that is not understood by an element, it shall ignore the message. Note: A message can be falsely identified as a valid message, passing the NetMIC and TransMIC authentication using a known network key and application key even though that message was sent using different keys. The decryption of that message using the wrong keys would result in a message that is not understood by the element. The probability of such a situation occurring is small but not insignificant. A message that is not understood includes messages that have one or more of the following conditions: • The application opcode is unknown by the receiving element. • The access message size for the application opcode is incorrect. • The application parameters contain values that are currently Prohibited. Note: An element that sends an acknowledged message that is not understood by a peer node will not receive any response message. ``` Signed-off-by: Michał Narajowski <[email protected]> Bluetooth: Mesh: Add msg length check for Cfg and Health models According to spec we should ignore messages with incorrect msg size. This patch adds a check to every opcode handler. Signed-off-by: Michał Narajowski <[email protected]> samples: Update Mesh opcode handlers Add return values to opcode handlers and update message length definitions. Signed-off-by: Michał Narajowski <[email protected]> doc: release-notes-2.7: Return value for opcode callback Add release notes for Stable API Change in Mesh. Signed-off-by: Michał Narajowski <[email protected]> Bluetooth: controller: df: Fixes truncated transmission of CTE There was an issue with wrong length of CTE send in connectionless mode, with periodic advertising PDUs. Radio peripheral was not configured to send CTE with correct length while PDU had CTEInfo field informing receiver that CTE is attached to the PDU. Source of the problem was in ll_df_set_cl_cte_tx_enable function. Order of parameters in ull_adv_sync_pdu_alloc was wrong. ULL_ADV_PDU_HDR_FIELD_CTE_INFO was speficed as hdr_rem_fields. Because of that extra_data, memory used to provide CTE configuration to LLL, was not allocated. PDU content is prepared in ULL, so CTEInfo field included correct information, while Radio was never configured by LLL to send CTE. ull_adv_sync_extra_data_set_clear received a pointer to hdr_data, instead of a direct pointer to df_cfg structure. When extra_data was allocated correclty, wrong CTE configuration was provided copied there and LLL received invalid CTE length. It was different than the length in PDUs CTEInfo field. Signed-off-by: Piotr Pryga <[email protected]> drivers: spi: Use timeout for transfer completion Instead of waiting forever for the SPI transfer to complete, let's use a timeout value and bail out if elapsed. The timeout value logic is, xfer_len/frequency + tolerance Tolerance value can be modified using a Kconfig symbol, CONFIG_SPI_COMPLETION_TIMEOUT_TOLERANCE. It defaults to 200ms. Fixes: #33192 Signed-off-by: Manivannan Sadhasivam <[email protected]> drivers: sensor: shtcx: add driver supporting shtc1 and shtc3 Add driver for sensirion consumer humidity sensor line. Supports shtc1 and shtc3, but only shtc3 is tested. Signed-off-by: Thomas Stranger <[email protected]> drivers: sensor: sht3xd: use sys functions for crc and byteorder replace custom crc8 with sys/crc8 use sys_put/sys_get helpers for byteorder specific operations Signed-off-by: Thomas Stranger <[email protected]> tests: drivers: enable shtc3 build Enable shtc3 sensor for build_all test Signed-off-by: Thomas Stranger <[email protected]> dts/arm: stm32: Adding rng (Random Number Generator) This commit adds rng in stm32f2 dtsi. Signed-off-by: Krishna Mohan Dani <[email protected]> boards/arm: nucleo_f207zg: Enabling true rng (Random Number Generator) This commit enables true rng for nucleo_f207zg platform. This has been tested with tests/drivers/entropy/api and is working as expected. Signed-off-by: Krishna Mohan Dani <[email protected]> drivers: flash: mcux flexspi nor: Fix write on arbitrary offset If a write offset isn't a multiple of the nor page size, and the length is too large to fit within a single page, it could wrap around in that page. Tested on i.MX RT1064 internal flash using NVS settings Signed-off-by: Pieter De Gendt <[email protected]> net: pkt: introduce net_pkt_remove_tail() Introduce a helper function for being able to remove any arbitrary length from tail of packet. This is handy in cases when removing unneeded data, like CRC once it was verified. Signed-off-by: Marcin Niestroj <[email protected]> drivers: net: ppp: fix removing CRC16 from packet CRC16 was removed by simply decreasing length of the last fragment by 2. This worked as long as last fragment was longer than 1 byte. If not, then last fragment was corrupted (its length ended up being 65535), leading to undefined behavior. Fix CRC16 removal by utilizing recently introduced net_pkt_remove_tail(), that properly handles multiple fragments. Reported-by: Jim Paris <[email protected]> Signed-off-by: Marcin Niestroj <[email protected]> doc: extensions: doxyrunner: fix HTML output handling The extension was not evaluating the GENERATE_HTML option correctly. The get_doxygen_option returns a `List[str]`, not a `str`. This effectively means that the Zephyr apidoc has not been updated for a while as the extension was not moving the output to the final destination folder. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: use Sphinx generated make files for PDF build Sphinx automatically generates a Makefile/make.bat that allows to build the LaTeX output. This converts the CMake `pdf` target into a shim to this Makefile, making the solution more future-proof if Sphinx decides to change something. Signed-off-by: Gerard Marull-Paretas <[email protected]> doc: cmake: allow to override LATEXMKOPTS This is useful if the user wants to set specific latexmk options, e.g. -quiet for CI. Signed-off-by: Gerard Marull-Paretas <[email protected]> drivers: ieee802154: cc13xx_cc26xx: add subghz power saving When stopping the interface, also abort the RX routine and enable RF power saving. Will re-start RX on iface start. Also fixed a bug with `cmd_fs` crashing at RF wakeup because `rf_mode` was allocated on stack. Signed-off-by: Stancu Florin <[email protected]> doc: cleanup for PDF Cleanup docs for PDFs hiding text that is relevant only for online docs. Signed-off-by: Anas Nashif <[email protected]> doc: move kconfig to indices and table Right now the kconfig index is randomly placed in the guides. This is a table similar to other indeces we have and should be considered as part of an appendix rather than place it in the middle of the documentation. Signed-off-by: Anas Nashif <[email protected]> doc: print tilte of Kconfig on top Add the symbol name on top and in the page title. This will enhance search results and adds a missing page header that is needed. Signed-off-by: Anas Nashif <[email protected]> doc: start include README after 'getting started' Otherwise we end up with a self-reference to the same page we are in. Signed-off-by: Anas Nashif <[email protected]> doc: CI badge should point to main instead of master Point to the new branch to get the correct status. Signed-off-by: Anas Nashif <[email protected]> doc: create an index for PDF files Most of the content in the default index is HTML only, so create an index just for PDF and keep it clean. Signed-off-by: Anas Nashif <[email protected]> doc: add a comment about copyright notices Add a link to best practices about copyright notices and when to and when not to change them or add them. Fixes #6777 Signed-off-by: Anas Nashif <[email protected]> Bluetooth: Add identifiers for Bluetooth 5.3 The Bluetooth 5.3 specification was recently released, and has a new version identifier (12) assigned to it in the Bluetooth Assigned Numbers. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> lpspi: the lpspi driver for rt1170 enable the lpspi1 for the rt1170 evk board Signed-off-by: Crist Xu <[email protected]> net: openthread: add Kconfig options for periodic parent search The openthread has enhanced features for periodic parent search, this commit adds kconfig options to enable and configure these. Signed-off-by: Pieter De Gendt <[email protected]> device: add DEVICE_DT_GET_ONE macro Add DEVICE_DT_GET_ONE() macro. This macro is similar to DEVICE_DT_GET_ANY() but instead of returning NULL if no enabled, compatible device instance is found, it throws a compilation error. This is useful for compile-time checks in samples. Signed-off-by: Henrik Brix Andersen <[email protected]> samples: shields: lmp90100_evb: convert to using DEVICE_DT_GET_ONE() Convert from using device_get_binding() to DEVICE_DT_GET_ONE(). Signed-off-by: Henrik Brix Andersen <[email protected]> tests: drivers: build_all: add fake serial device for modem tests Serial device is needed for building drivers that use DEVICE_DT_GET(). None of the currently used modem drivers use that right now, but this is about to change. Signed-off-by: Marcin Niestroj <[email protected]> drivers: modem: operate on device pointers instead of names So far modem API used UART device names / labels. Change API to operate on device pointers instead, so that we stop using device_get_binding() in modem core and in some DT compatible modem drivers. Signed-off-by: Marcin Niestroj <[email protected]> drivers/sensor: lis2mdl: Fix errors when CONFIG_PM_DEVICE=y Fix a couple of issues related to Power Management: 1. A build error because 'dev' was used even if not declared in the caller routine scope 2. The lis2mdl power management init specific routine was not declared in in the device instance definition Signed-off-by: Armando Visconti <[email protected]> dts: gpio: it8xxx2: Remove unused property Property of port-is-output is not used anymore, so let's remove it. Signed-off-by: Dino Li <[email protected]> drivers: gpio: it8xxx2: add support for GPIO_VOLTAGE_ flags Support GPIO_VOLTAGE_1P8, GPIO_VOLTAGE_3P3 flags on IT8xxx2 chips. Signed-off-by: Dino Li <[email protected]> arch/Kconfig: Remove stray tab from USERSPACE help Commit removes stray tab from help. Signed-off-by: Dominik Ermel <[email protected]> Revert "tests: drivers: build_all: add fake serial device for modem tests" This reverts commit 9e58a1e475473fcea1c3b0d05ac9c738141c821a. This change is in conflict with commit 94f7ed356f0c ("drivers: serial: add a dummy driver for vnd,serial"). As a result two equal serial devices are defines, resulting in link error. Signed-off-by: Marcin Niestroj <[email protected]> dts: Introduce DT_STRING_TOKEN and DT_STRING_UPPER_TOKEN To be able to get a tokenize DT string without the quotes. Deprecate also DT_ENUM_TOKEN and DT_ENUM_UPPER_TOKEN. Signed-off-by: Carlo Caione <[email protected]> reserved-memory: Introduce support for reserved-memory node Introduce a set of header files to be able to define and declare sections and regions in the linker script. Introduce also DT helpers to retrieve data back. Signed-off-by: Carlo Caione <[email protected]> samples: reserved_memory: Introduce sample application Introduce sample application to test reserved-memory helpers. Signed-off-by: Carlo Caione <[email protected]> drivers: gpio: add new driver for STMPE1600 gpio expander The STMPE1600 is an I2C based GPIO expander. This initial patch only supports reading from/writing to the pins on the STMPE1600, and there is currently no support for interrupts. Signed-off-by: Titouan Christophe <[email protected]> drivers: i2c: fix timing and stop bit Current impĺementation fails due to missing stop bit when reading data. This fixes it and refactor the implementation by adding k_busy_wait when waiting I2C bus completion. Also, this implementation is based on esp-idf v3.0. It will be refactored based on latest esp-idf v4.3 using proper low level calls. Signed-off-by: Sylvio Alves <[email protected]> Fixes #34015 west.yml: hal_stm32: stm32cube/common_ll: Update vs last Cube update Update hal_stm32 module to integrate latest version of the stm32cube/common_ll headers. Signed-off-by: Erwan Gouriou <[email protected]> dts: bindings: sx12xx: add ST stm32wl-subghz-radio node Add a dts node for st,stm32wl-subghz-radio. The device is integrated in STM32WL series SoCs, and based on the sx1262, but it does not use any gpio as all the necessary signals are internally connected to various SoC units. To account for that, make the redundant gpio optional in the template files, but mark them as required in the sx1261 and sx1262 definitions, to match with what's used in the driver. Signed-off-by: Fabio Baltieri <[email protected]> drivers: spi_ll_stm32: add SUBGHZSPI support This adds support for controlling the SUBGHZSPI NSS line in STM32WL devices. This is a special dedicated SPI port only connected to the radio device internally, chip select happens through a bit in the PWR module. Adding a special dt-property to identify the port, it all gets built out on non-WL devices. Deduplicate the existing dts bindings in the process, and add the new one for the special spi with the new property. Signed-off-by: Fabio Baltieri <[email protected]> drivers: sx126x: refactor few functions for stm32wl support Move few functions around in preparation of adding the STM32WL variant of the driver. Signed-off-by: Fabio Baltieri <[email protected]> drivers: sx126x_standalone: convert to gpio_dt_spec APIs Convert the standalone part of the sx126x driver to the new gpio_dt_spec APIs. This allows moving the specific GPIOs in that part of the driver and out of struct sx126x_data. Signed-off-by: Fabio Baltieri <[email protected]> drivers: sx126x: add support for the stm32wl chip Add support for the STM32WL integrated radio, based on the sx1262. The STM32WL implementation does not use any GPIO, and the signals are routed to other units of the SoC and accessed with the ST HAL. Signed-off-by: Fabio Baltieri <[email protected]> dts: stm32wl: define a subghz-spi radio node Add a device node for the subghz radio in stm32wl.dtsi. The radio is present in all chips of the series, so having the node there with the common properties simplifies the board dts files. Signed-off-by: Fabio Baltieri <[email protected]> boards: nucleo_wl55jc: add lora support Enable the subghz radio for the lora samples in nucleo_wl55jc. Signed-off-by: Fabio Baltieri <[email protected]> boards: lora_e5_dev_board: add lora support Define the subghz radio node in the module dtsi file, with the internal RF switch connection and TCXO setting, and point at it from the corresponding dev board file. Signed-off-by: Fabio Baltieri <[email protected]> samples: lora, class_a: allow building on nucleo_wl55jc Add nucleo_wl55jc to platform_allow for lora send/receive and class_a samples. The board has a variation of the sx1262 driver, this should allow to build test that code path. Fix the existing config as well to specify the platforms on a single line, as currently only the last entry is considered. Signed-off-by: Fabio Baltieri <[email protected]> Bluetooth: Controller: nRF5: Use conservative Tx-Rx turnaround timing Use a conservative turnaround time of 150us covering slower nRF51 and faster nRF52 CPU that influence the turnaround timing. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: controller: Extended/Periodic Advertising time reservations Add Extended and Periodic Advertising radio event time reservations. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Refactor out advertising time calculations Refactor out advertising time calculations so that it can be reused when PDU length is updated as part of advertising data and scan response data updates. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Legacy advertising time reservation Implementation to update legacy advertising time reservations when advertising and scan response data are updated. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Refactor out Extended Advertising time calc Refactor out Extended Advertising time calculations so that it can be reused when PDU length is updated as part of advertising data and scan response data updates. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Update DID when scan response data removed Update DID value when scan response data is remove, length set to 0. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Export primary adv channel event time calc Export the function calculating the primary channel advertising event time reservation, so that Extended Advertising implementation can reuse it. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Extended advertising time reservation Implementation to update Extended Advertising time reservations when advertising and scan response data are updated. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Refactor out Periodic Advertising time calc Refactor out Periodic Advertising time calculations so that it can be reused when PDU length is updated as part of advertising data updates. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Disallow Periodic Adv Enable if already enabled Add check for already enabled and started Periodic Advertising when HCI LE Set Periodic Advertising Parameters command is called. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Periodic advertising time reservation Implementation to update Periodic Advertising time reservations when advertising data and/or CTE length is updated. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Fix Extended and Periodic Coded PHY time Fix Extended and Periodic Coded PHY time reservation. The define earlier used does not have calculations for Coded PHY. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Use BT_HCI_ERR_SUCCESS return and other changes Apply suggestions from code review, and change to using BT_HCI_ERR_SUCCESS instead of returning 0. Co-authored-by: Piotr Pryga <[email protected]> Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> Bluetooth: Controller: Refactor Periodic Advertising time update Refactor Periodic Advertising time update function to use caller supplied Periodic Advertising PDU to calculate the time reservations. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]> MAINTAINER: remove jocelyn-li from collaborat…
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR supports #33104 in updating the use of k_work delayable timers in the Bluetooth mesh implementation. The API has been updated, and appears to work on one sample with limited testing.
It is in draft format, and all commits tagged DNM, because this needs significant review and rework from domain experts to address the following concerns. More details are in comments in the code, and in the individual commit messages. At least one unrelated probable bug has been highlighted (in cfg_srv). Overall:
On the whole this code does not appear to be correct in the presence of preemptive threads or SMP systems. The use of atomic flags without an enclosing lock for operations that are not read-modify-write is not robust in those contexts.
There were many uses of
k_delayed_work_cancel(), none of which took into account the possibility that cancellation can fail, and several of which would likely cause faults if the work item continued to be after the post-cancellation state updates were made.It is generally not clear whether schedule or reschedule operations are appropriate.
My preferred path forward is for a mesh maintainer to take these commits and rework them as necessary.
(NB: This is currently based on and incorporates #33614 to support the sample.)