Fix the leak of node rx buffer used to generate the
connection complete and CSA#2 event introduced in the
commit 4a5f263e ("Bluetooth: controller: split: Validate
chan map and hop value") and the
commit 94d5f085

 ("Bluetooth: controller: fixing error
re. all zero chmap in conn-ind").
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>

Replace a development assertion in the implementation of
Connection Parameter Request Procedure with an internal
comment and handle transaction violation be ignoring the
PDU.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>

Remove the redundant connection initiated check as the event
is closed on connection initiated and it is sufficient to
check in the prepare_cb function to abort any events in the
pipeline after the connection has been initiated.

Relates to commit 5ce5dc05 ("Bluetooth: controller:
Avoid race between ULL and LLL when initiating conn") and
commit 18f5fb99

 ("Bluetooth: controller: Remove use of
lll_stop").
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>

Remove use of lll_stop and lll_is_stop and rely on "initiated" flag in
lll_conn struct instead.
Signed-off-by: Wolfgang Puffitsch <wopu@demant.com>

Use an "initiated" flag in the lll_conn struct to guard the processing
of PDUs related to connection initiation (CONNECT_IND,
AUX_CONNECT_RSP). This avoids races between ULL and LLL when creating
a connection.
Signed-off-by: Wolfgang Puffitsch <wopu@demant.com>

Commit 0a7b65ef

 tweaked the CREATE_OBJ_LEVEL macro in such a way
that it would break the expected sorting order.

For example if you had 2, 19, 20, 30 as the level, we'd end up sort
these to be 19, 2, 20, 30.

Fix this by adding aditional "_" symbol after the init level counter.
That allows to keep correct sort order (for both GNU and MWDT
toolchains) and distinguish init level counter from section suffix
(for MWDT toolchain).

Fixes zephyrproject-rtos#33464
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com>

CAN_SYNC_SEG and ts1 are in common units. Both need to be scaled by 1000
to calculate the sample point.
Signed-off-by: Andriy Gelman <andriy.gelman@gmail.com>

We should verify that the buffer has sufficient data before attempting
to parse the SDU length field. If we get a too short packet just
disconnect the channel.

Fixes #32497
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>

Fix control Tx buffer leak into data Tx pool that happens
after a cross-over control procedure response was paused due
to currently active encryption setup procedure, and a new
control Tx PDU in addition to the paused one is enqueued
thereafter.

When the control tx PDUs is resumed but not yet enqueued
towards the radio, if there is a new control Tx PDU enqueued
then the paused control Tx PDU is not set as the head of the
control PDUs in the Tx queue. This caused the paused control
Tx PDU to be associated with data Tx pool, hence causing the
incorrect release into data Tx pool.

Relates to the commit bff76b4c ("Bluetooth: controller:
split: Fix control tx queue handling") and to the
commit 6991d099

 ("Bluetooth: controller: Fix control tx
queue handling").

Fixes #32898.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>

The order of Destination CIDs shall correspond the order of Source CIDs
including its amount so errors that don't result in all connection being
refused shall not break the order of CIDs.
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>

For errors that means all connections have been refused there is no need
to add dcids since none will be valid.
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>

When receiving L2CAP_CREDIT_BASED_CONNECTION_REQ the remote may request
more channels than allowed so this checks if amount of channel surpasses
the maximum channels (5) and return an error.
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>

Overwrite the existing bond when the IRK of the existing bond could not
resolve the RPA of the peer. This would happen if the peer has deleted
the bond and replaced the IRK that was used.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>

Refactor update_keys_check helper function to operate on input keys
input. This allows the function to be re-used on a keys structure that
is not the current connection keys.

This also avoids the helper function changing the connection state.
The conn->le.keys pointer should at this point always have been
assigned, as central when sending the pairing request, and as peripheral
when receiving the pairing request at the very latest.
Signed-off-by: Joakim Andersson <joakim.andersson@nordicsemi.no>

Generated device and driver names have changed with commits [1] and [2],
but net offload drivers missed the conversion. Update net offload DT
device macros now.

[1] commit 8c1bef53 ("device: support generating defines from
  devicetree nodes with no label")
[2] commit f91e9fba

 ("device: fix potential truncation of DT-derived
  device names")
Signed-off-by: Marcin Niestroj <m.niestroj@grinn-global.com>

A different configuration is applied when running this test
using west or using twister.
Both will use the disco_l475_iot1 overlay, but the extra config
for clock bus tunning is applied only when using twister and
samples is failed when run with west.
Move this extra config from sample.yaml to board .conf file so
it is applied in both cases.

Fixes #32376
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>

If user sets CONFIG_NUM_PREEMPT_PRIORITIES=0, then the priority
of the net_mgmt thread will be -1 which is the same as idle thread.
This will trigger assert in kernel as then the minimum coop priority
is -2 in this case. Remove the net_mgmt thread priority setting from
Kconfig file as it is low value and set the coop thread priority
the same way as other network threads are doing it.

Fixes #32375
Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>

CLOCK_STM32_PLL_XTPRE Kconfig symbol default value is n.
Then there is no need to explicitly set it to 'n' in stm32f1 boards
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>

This reverts commit "drivers/clock_control: Remove useless
CLOCK_STM32_PLL_XTPRE config" 9be1f7e22f3b3c42009eeba15061cad3c0988b22.

Fixes #32382
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>

zassert_within should also compare on equality instead
of only greater/lower.
example:
zassert_within(1,1,0); // should return true
zassert_within(1,2,1); // should return true
Signed-off-by: Alexander Wachter <alexander@wachter.cloud>

Correct a bunch of precision/analysis errors in this test:

* Test items weren't consistent about tick alignment and resetting of
  the timestamp, so put these steps into init_timer_data() and call
  that immediately before k_timer_start().

* Many items would calculate the initial timestamp AFTER
  k_timer_start(), leading to an extra (third!) point where the timer
  computation could alias by an extra tick.  Always do this
  consistently before the timer is started (via init_timer-data()).

* Tickless systems with high tick rates can easily advance the system
  uptime while the timer ISR is running, so the system can't expect
  perfect accuracy even there (this test was originally written for
  ticked systmes where the ISR was by definition happening "at the
  same time").

  (Unfortunately our most popular high tick rate tickless system,
  nRF5, also has a clock that doesn't divide milliseconds exactly, so
  it had a special path through all these precision comparisons and
  avoided the bugs.  We finally found it on a x86 HPET system with 10
  kHz ticks.)

* The interval validation was placing a minimum bound on the interval
  time but not a maximum (this mistake was what had hidden the failure
  to reset the timestamp mentioned above).

Longer term, the millisecond precision math in these tests is at this
point an out of control complexity explosion.  We should look at
reworking the core OS tests of k_timer to use tick precision (which is
by definition exact) pervasively and leave the millisecond stuff to a
separate layer testing the alternative/legacy APIs.

Fixes #31964 (probably -- that was reported against up_squared, on
which I had trouble reproducing, but it was a common failure on
ehl_crb).
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>

If userspace is enabled, then the TLS context needs to be
made a NET_SOCKET kernel object. Without this the userspace
cannot access TLS sockets.
Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>

A recent patch removed use of the cfg structure, but left a pointer to
it defined which causes build failures.
Signed-off-by: Peter Bigot <peter.bigot@nordicsemi.no>

On I2C V1 parts, LL_I2C_EnableIT_TX() translates to EVT and BUF
IRQ enabling.
In stm32_i2c_msg_write function, LL_I2C_EnableIT_TX is called right
after stm32_i2c_enable_transfer_interrupts which already enables BUF
IRQ, which starts the transfer.
As a consequence it could happen that transfer is already complete
at the time LL_I2C_EnableIT_TX is called. This case is not expected
by remaining part of the code which loops forever waiting for BUF IRQ
to be raised.
Remove the superfluous LL_I2C_EnableIT_TX call.

Fixes: #32265
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>

nRF52840, nRF52810, nRF52811 and nRF52805 are affected by anomaly 242.
This patch introduces workaround for this anomaly as follow:
Power-fail comparator is disabled before any attempt to erase or write.
Either erase or write is not proceed if EVENT_POFWARN is
already asserted.
Signed-off-by: Andrzej Puzdrowski <andrzej.puzdrowski@nordicsemi.no>

Introduce support for situation when synchronization back-end
aborts operation before it is done. synchronization API will
transfer operation return code to the driver shim back.

Additionally:
FLASH_OP_ONGOING value was switched to be positive in order to
not been mislead with a negative error code.
Signed-off-by: Andrzej Puzdrowski <andrzej.puzdrowski@nordicsemi.no>

Add example to handle OOB host responses asynchronously.
Signed-off-by: Jose Alberto Meza <jose.a.meza.arellano@intel.com>

Define macro for number of times the temperatures will
be retrieved.
Signed-off-by: Jose Alberto Meza <jose.a.meza.arellano@intel.com>

Keep OOB Rx channel and interrupt always enabled.

Send callback when packet is received in OOB Rx channel
if asynchronous host-initiated handling is enabled.

Note that driver doesn't perform any buffering from packets,
so access to OOB Rx channel is gated by client's driver
packet retrieval.
Signed-off-by: Jose Alberto Meza <jose.a.meza.arellano@intel.com>

Currently assumption is all OOB traffic over eSPI bus is always client
initiated.
Add option for systems where host can initiate OOB traffic.
Signed-off-by: Jose Alberto Meza <jose.a.meza.arellano@intel.com>

The calculated alarm time starts from 2000 but the gmtime_r needs as
input the time from epoch (1970). This causes the alarm time to be
miscalculated due to leap years, as 2000 is a leap year and 1970 is not.
To fix the issue, the 2000 timestamp can be added to the input time of
gmtime_r.

Fixes #32260
Signed-off-by: Antonis Sioutas <antonis.si510@gmail.com>

Enable the LL/CON/INI/BV-24-C [Network Privacy - Connection
Establishment using resolving list, initiator, Ignore
Identity Address, with address resolution disabled] EDTT
test.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>

When address resolution is disabled, an identity address has
been added into the resolving list with peer IRK, and device
privacy has not been selected for the peer device then
connection indication shall not be sent to the peer that is
advertising using its identity address.

Fixes #24731.
Signed-off-by: Vinayak Kariappa Chettimada <vich@nordicsemi.no>

Fixes: #23449

This commit adds additional OUTPUT and BYPRODUCTS to custom command
and targets in the Zephyr build system.

This ensures that files produced during the build will be removed again
when invoking ninja clean / make clean.

The generated syscalls headers include folder is added to the syscall
target using ADDITIONAL_CLEAN_FILES property.
However, this property is new in CMake 3.15, so will not work when using
older CMake with ninja.

For CMake versions <=3.15 the ADDITIONAL_MAKE_CLEAN_FILES property is
used. However, this only supports Makefile generator.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>

Fixes: "WARNING:CONSTANT_COMPARISON: Comparisons should place the
constant on the right side of the test"
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>

STM32 uart driver doesn't support 9bits transactions in any case,
so remove case were it was declared as supported.

Fixes #31799
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>

Clear Floating Point Status and Control Register (FPSCR),
to prevent from having the interrupt line set to pending again,
in case FPU IRQ is selected by the test as "Available IRQ line"

Fixes #31982
Signed-off-by: Alexandre Bourdiol <alexandre.bourdiol@st.com>

acl_read_cb does not handle incoming (ACL) data
if BT_CTLR_TX_BUFFER_SIZE is greater than
USB_MAX_FS_BULK_MPS - BT_HCI_ACL_HDR_SIZE.

Since the host adjusts the data according to
the BT_CTLR_TX_BUFFER_SIZE and does not use ZLP
we cannot start usb_transfer over the possible length of
the whole packet, with or without USB_TRANS_NO_ZLP flag.
But we can read the packet length from the header and
call net_buf_put() when the whole packet is received.

Fixes: #31922
Reported-by: Matias Karhumaa <matias.karhumaa@gmail.com>
Signed-off-by: Johann Fischer <johann.fischer@nordicsemi.no>

Work items can be legally resubmitted from within their own handler.
Currently the p4wq detects this case by checking their thread field to
see if it's been set to NULL.  But that's a race, because if the item
was NOT resubmitted then it no longer belongs to the queue and may
have been freed or reused or otherwise clobbered legally by user code.

Instead, steal a single bit in the thread struct for this purpose.
This patch adds a K_CALLBACK_STATE bit in user_options and documents
it in such a way (as being intended for "callback manager" utilities)
that it can't be used recursively or otherwise collide.

Fixes #32052
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>

By the time hostname configuration was implemented, driver was switching
only between STA and STA+AP modes. After dynamic selection between NONE,
STA, AP and STA+AP was implemented (commit referenced below), hostname
configuration no longer takes effect when ESP chip obtains address over
DHCP (and sends hostname in the DHCP request).

Set hostname each time after enabling STA mode, so that it takes effect
in DHCP requests.

Fixes: 03ce6100

 ("drivers: wifi: esp: control CWMODE depending on
  current needs")
Signed-off-by: Marcin Niestroj <m.niestroj@grinn-global.com>