The implementation of offload_getaddrinfo in this driver failed
when the node it was called with was an IP address. This condition
was never detected, and as a consequence a DNS query was done on
the IP address instead of returning it directly.
Also, the port was set first after running the DNS query.
As a consequence, if the IP address would have been returned directly,
this would have been done without a port been set.
Both errors are fixed in this patch.
Signed-off-by: Hans Wilmers <hans@wilmers.no>
Prevent usage of system work queue before it is ready by delaying
initialization of temperature device. Work queue is not used until
application stage.
Signed-off-by: Krzysztof Chruscinski <krzysztof.chruscinski@nordicsemi.no>
There does not seem to be a simple way to express the fact that UART1
can be initialized later at the POST_KERNEL level. The existing code
is wrong in the sense that it always initializes DT instance 1 later,
instead of doing it for UART1 (which may not be instance 1).
In addition, UART instances on other platforms are
also initialized at the PRE_KERNEL_1 level regardless of the instance
index, so let's do the same on this platform for consistency.
Signed-off-by: Vincent Wan <vincent.wan@linaro.org>
The power configuration is dependent on which UART is physically used.
In order to allow DT_INST_FOREACH_STATUS_OKAY() to iterate through
instances without the assumption that index 0 corresponds to UART0
(which would be incorrect in the case when only UART1 is enabled),
we need to check the base address to identify which UART is being dealt
with.
Signed-off-by: Vincent Wan <vincent.wan@linaro.org>
This reimplements z_timer_cycle_get_32() so it works
when IRQs are locked and solves the hung
k_busy_wait() problem.
Fixes#23622.
Signed-off-by: Jakub Cebulski <jcebulski@internships.antmicro.com>
Signed-off-by: Tomasz Gorochowik <tgorochowik@antmicro.com>
Signed-off-by: Mateusz Holenko <mholenko@antmicro.com>
Co-authored-by: Tomasz Gorochowik <tgorochowik@antmicro.com>
Updating the eswifi driver to provide port information (along with
IP / net context) whenever an async message is received. This is
needed to fully populate the "remote" info in "net_context"
required whenever net_context_send is called.
Also adding code to populate "remote" info in "net_context"..
Tested with STM32 disco IOT kit. Fixes#25621
Signed-off-by: Bilal Wasim <bilalwasim676@gmail.com>
When the loop that tries to obtain the lock in poll_out() finishes
because of hitting the max number of trials (what normally should
never happen), force the lock to be taken instead of just giving up
with sending the data. The latter approach that was in use so far
could not deal with a situation when some thread was aborted while
keeping the lock. Other threads would then have no chance to send
anything with poll_out() until it was called from an ISR.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Correct dependency of the TEMP_NRF5 Kconfig option so that it can be
enabled only on nRF SoCs that feature the TEMP peripheral.
Signed-off-by: Andrzej Głąbek <andrzej.glabek@nordicsemi.no>
Some Kconfig defined devices may be defined using dt_chosen_label
function. Since there is no way to ensure a device enabled in dts
is also defined in Kconfig, it may happen that instance is not
actually defined.
In this case device_get_binding might return 0, leading to undefined
behavior in the function that calls it.
When not already done, systematically check return of function
device_get_binding on devices defined through dt_chosen_label macro.
Trigger ASSERT when required and return error when possible.
Fixes#20068
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>
Nordic driver must allow read from control OUT endpoint by
itself. For data stage transactions with length > MPS (64Bytes)
this must be performed for each 64B + Residue data packet.
Residue - data packet with len < 64B.
The exact length of data transfer is known from wLength field
form setup packet in setup stage. Until now driver was incorrectly
initializing the length of the data stage and at some point will
not allow for next data stage.
This commit addresses the issue #23980.
Signed-off-by: Emil Obalski <emil.obalski@nordicsemi.no>
The weak implementation returns 0 for all operations without doing
anything, which incorrectly suggests that an operation like
device_get_power_state() returned an accurate description of the
system clock power state. Return -ENOTSUP instead.
Signed-off-by: Peter Bigot <peter.bigot@nordicsemi.no>
If timeout is being overwrite exactly when previous one is expiring
then hardware event was cleared correctly but interrupt was already
triggered. Interrupt routine was assuming that compare event is set
and proceed with that assumption. However, in that corner case when
compare event was overwritten and event was cleared, that was not the
case.
As the outcome, timeout could be triggered prematurely. Fixed by
clearing pending interrupt after handling previous compare value.
Signed-off-by: Krzysztof Chruscinski <krzysztof.chruscinski@nordicsemi.no>
mcuboot and possibly other tools read single byte values to determine
the state of objects. Rather than fail to do the read of values too
short for this peripheral detect the situation and read into a stack
buffer that meets the length criteria, and on success copy the data
into the provided buffer.
Signed-off-by: Peter Bigot <peter.bigot@nordicsemi.no>
* still need to clear IP bit in timer irq handler
* last_time should be aligned to ticks, old code will miss some
cycles which are about (curret_time - last_time) % CYC_PER_TICK
* in timeout set, shorten the delay needed when tick is 0, this
will improve the response of timer irq
Signed-off-by: Wayne Ren <wei.ren@synopsys.com>
the pulse triggered timer irq doesn't work for all targets. In
iotdk, we found the clear of IP bit will clear int request
when elapsed called in thread context. So come back to level
triggered way which is supported in all targets, and use the sw
triggered irq to remember the irq request which may be cleared
in non timer int handler.
Signed-off-by: Wayne Ren <wei.ren@synopsys.com>
* rename overflow_cyc to overflow_cycles for better understanding
* use MIN macro to replace if .. else ..
* typo fix in comments
Signed-off-by: Wayne Ren <wei.ren@synopsys.com>
Allow configuring the clock prescaler divider for the NXP Kinetis
FlexTimer. Setting the prescaler to a lower value allows for much
higher resolution/accuracy for the generated PWM waveforms.
Signed-off-by: Henrik Brix Andersen <hebad@vestas.com>
User can now add extra Ethernet TAP parameters when starting QEMU.
This is useful if we want to set for example the MAC address
of the network interface.
Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
When PCIe is enabled for UART, the port address is probed during
initialization and is written back into the device config struct.
However, the device config struct is supposed to be const and
read only. This results in page faults when MMU is enabled as
the struct cannot be written into. So fix this by storing port
address in device data struct if a particular UART instance is
of PCIE.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
The init function returns successful even if the first
configuration function call fails. This may leave
a non-usable UART to be discoverable with
device_get_binding() which will definitely result
in lots of head scratching. So change the init function
to return properly.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
This adds the calls to read_timer_{start,end}_of_tick_handler()
to mark the start and end of ISR which will be used to display
the time spent in ISR with benchmarking tests.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
IRQ_CONNECT() can only be called at one location to connect the irq for
CPE0. This commit modifies the driver to call into the HwiP layer in TI
HAL so that TI's RF driver can do the same when connecting the irq.
Fixes#25216
Signed-off-by: Vincent Wan <vincent.wan@linaro.org>
New update of hal_ti requires DeviceFamily_CC13X2/DeviceFamily_CC26X2
to be defined in order to include the rfc.h header.
Signed-off-by: Vincent Wan <vincent.wan@linaro.org>
Some kernel tests use `CONFIG_TICKLESS_KERNEL=n` with
`CONFIG_SYS_CLOCK_TICKS_PER_SEC=1` to detect when a test runs longer
than 1 second. These tests break if a tick is announced every time a
timeout occurs. Only announce if the measured duration since the last
tick is at least the duration of a tick.
Signed-off-by: Peter Bigot <peter.bigot@nordicsemi.no>
Add +1 tick to alarm in order to compensate the partially started tick.
Alarm will expire between requested ticks and ticks+1.
In case only 1 tick is requested, it will avoid that +1 Tick event
occurs before alarm setting is finished.
Signed-off-by: Alexandre Bourdiol <alexandre.bourdiol@st.com>
The check for assertion on the "config_func" was added to
validate that the function pointer is valid. However, in
the code we are invoking the "config_func" and comparing
its output with NULL. This causes build failures with
CONFIG_ASSERT=1. Caused by PR-25393.
Tested on Nucleo F767Zi board.
Fixes#25427
Signed-off-by: Bilal Wasim <bilalwasim676@gmail.com>
The existing uart driver ns16550 did not have ISR locking that
effected IO APIC working in fixed delivery mode in SMP system
x86_64. This commit adds ISR locking mechanism using spinlock
for the interrupt related services.
The CONFIG_IPM_CONSOLE_STACK_SIZE is increased to lift
limitation of stack size experienced in IPM driver test with
this spinlock impelentation.
Fixes#23026
Signed-off-by: Jennifer Williams <jennifer.m.williams@intel.com>
Current dma struture code didn't allowed only rx channel removal,
disabling tx channel (in spi client node) was leading compilation
issue.
Fix this by moving conditional code inside SPI_DMA_CHANNEL macro and
get the part of code which is present or removed (SPI_DMA_CHANNEL_INIT)
outside of {}.
Additionally, fix indentation on '\' in whole instance init macros
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>
All initialization of the Ethernet interface is done in the
eth_initialize function which is invoked by the boot code.
This function sets up DMA, programs the Ethernet module and
enables IRQs. However, this function does not setup "netif"
interface info which is done when the ethernet device is
enumerated by the NET stack via the "iface_api.init" func.
However, after the eth_initialize func is called, it is
possible that the system receives RX interrupts, and the
"rx_thread" accesses the "netif" pointer to get iface info.
However, because the "netif" info is not necessarily
populated at this time, we get a crash (as OS does NULL
access).
Fixed by enabling Ethernet IRQ after the interface is
properly setup.
Tested on Nucleo F767Zi board.
Fixes#25408
Signed-off-by: Bilal Wasim <bilalwasim676@gmail.com>
Use device tree provided configurations for arm architecture timer
PPIs.
This fixes issue of timer ppi not working on most hardware where
edge-triggered PPI are not supported.
Signed-off-by: Sandeep Tripathy <sandeep.tripathy@broadcom.com>
Hardware does not seem to support triggering interrupts to
itself by setting line as both input/output and setting
output to desired level. So just say interrupt triggering
is not supported when line is set to output.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Change adds abort of ongoing write operation in ep_ctx_reset. This is
required to keep the state of Zephyr driver consistent with state of
nrfx driver. This fixes a bug where nrfx_usbd was stuck in busy state.
Signed-off-by: Marek Pieta <Marek.Pieta@nordicsemi.no>
If interrupt is previously enabled and triggering condition
changes, it might fire callback based on previous condition.
To avoid this issue, disable interrupt before trying to
change settings, and enable afterwards.
Signed-off-by: Daniel Leung <daniel.leung@intel.com>
Several reviewers agreed that DT_HAS_NODE_STATUS_OKAY(...) was an
undesirable API for the following reasons:
- it's inconsistent with the rest of the DT_NODE_HAS_FOO names
- DT_NODE_HAS_FOO_BAR_BAZ(node) was agreed upon as a shorthand
for macros which are equivalent to
DT_NODE_HAS_FOO(node) && DT_NODE_HAS_BAR(node) &&
- DT_NODE_HAS_BAZ(node), and DT_HAS_NODE_STATUS_OKAY is an odd duck
- DT_NODE_HAS_STATUS(..., okay) was viewed as more readable anyway
- it is seen as a somewhat aesthetically challenged name
Replace all users with DT_NODE_HAS_STATUS(..., okay), which is
semantically equivalent.
This is mostly done with sed, but a few remaining cases were done by
hand, along with whitespace, docs, and comment changes. These special
cases include the Nordic SOC static assert files.
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
The driver-specific config_info structure referenced from the device
structure is marked const. Some drivers fail to preserve that
qualifier when casting the pointer to the driver-specific structure,
violating MISRA 11.8.
Changes produced by scripts/coccinelle/const_config_info.cocci.
Some changes proposed by the script are not included because they
reveal mutation of state through the const pointer, though the
code works as long as the driver-specific object is defined without
the const qualifier.
Signed-off-by: Peter Bigot <peter.bigot@nordicsemi.no>
With this change, the spi transceive with dma function
waits for the spi busy flag reset and for the dma transfer end.
Then it reloads the channels until all buffers are consumed.
Signed-off-by: Francois Ramu <francois.ramu@st.com>
This change avoids the reload of the dma channel
in the callback function, just sets the corresponding Tx, Rx flag.
Signed-off-by: Francois Ramu <francois.ramu@st.com>
