The death of a network context was sort of a mess. There was one
function, net_context_put(), which was used both by the user as a way
to "close" the connection and by the internals to delete it and to
"clean up" a TCP connection at the end of its life.
This has led to repeated gotchas where contexts die before you are
ready for them (one example: when a user callback decides the
transation is complete and calls net_context_put() underneath the
receive callback for the EOF, which then returns and tries to inspect
the now-freed memory inside the TCP internals). I've now stepped into
this mess four times now, and it's time to fix the architecture:
Swap the solitary put() call for a more conventional reference
counting implementation. The put() call now is a pure user API (and
maybe should be renamed "close" or "shutdown"). For compatibility,
it still calls unref() where appropriate (i.e. when the context can be
synchronously deleted) and the FIN processing will still do an unref()
when the FIN packets have been both transmitted and acked. The
context will start with a refcount of 1, and all TCP callbacks made on
it will increment the refcount around the callback to prevent
premature deletion.
Note that this gives the user a "destroy" mechanism for an in-progress
connection that doesn't require a network round trip. That might be
useful in some circumstances.
Change-Id: I44cb355e42941605913b2f84eb14d4eb3c134570
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
net_addr_ntop() will convert IPv4|6 address to string form.
Renamed existing net_sprint_ip_addr_buf() to net_addr_ntop()
and adjusted parameters as per API.
Jira: ZEP-1638
Change-Id: Ia497be6bf876ca63b120529acbadcfd9162a96e3
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@linux.intel.com>
Currently, the function accepts a struct sockaddr * but the code
immediately type casts this to either in_addr or in6_addr. This is
incorrect behavior as the first field in a sockaddr is sa_family_t
and not address data.
So without special knowledge, a developer will use a sockaddr structure
as the parameter and then wonder why the address information isn't being
set correctly.
Let's change this parameter to void * which makes this function similar
to inet_pton().
Jira: ZEP-1616
Change-Id: I1fc9368da999d90feb07c03fac55dcc749d4eba6
Signed-off-by: Michael Scott <michael.scott@linaro.org>
It was in the static initializers, but was missing from the object
runtime init functions.
Change-Id: I10d519760eabdbe640a19cc5cfa9241c1356b070
Signed-off-by: Benjamin Walsh <walsh.benj@gmail.com>
This will allow users to install a way of finding out what the event and
the objects are used for without looking at the object itself, or to
tag a bunch of objects that belong together.
The runtime init function _does not_ take a tag so that there is no
runtime hit if not needed. The static initializer macro _does_ take the
tag, so that it does not have to be initialized at runtime if needed,
and thus avoids a runtime hit.
Change-Id: I89a36c6f969ff952f9d1673b1bb5136e407535c6
Signed-off-by: Benjamin Walsh <walsh.benj@gmail.com>
These interrupts are for ISRs that need the lowest possible latency.
They do not take parameters and are installed directly in the interrupt
vector table.
Issue: ZEP-1038
Change-Id: I7583e9191dd32d9253ad933181d2103a6e191dea
Signed-off-by: Andrew Boie <andrew.p.boie@intel.com>
k_poll() is similar to the POSIX poll() API in spirit in that it allows
a single thread to monitor multiple events without actively polling
them, but rather pending for one or more to become ready. Such events
can be a direct event, or kernel objects (currently only semaphores and
fifos).
When a kernel object being polled on is ready, it is not "given" to the
poller: the poller must then acquire it via the regular API for the
object (e.g. k_sem_take()). Only one thread can poll on a particular
object at one time. These restrictions mean that k_poll() is most
effective when a single thread monitors multiple events that are not
subject for contention. For example, being the sole reader on multiple
fifos, or the only thread being signalled by multiple semaphores, or a
combination of both.
Change-Id: I7035a9baf4aa016fb87afc5f5c0f5f8cb216480f
Signed-off-by: Benjamin Walsh <walsh.benj@gmail.com>
Allow peeking at the fifo to see if there is an element without
dequeuing it.
Change-Id: I99cbe4495c81f1d7b77ad6a37cef4ec8c24d48eb
Signed-off-by: Benjamin Walsh <walsh.benj@gmail.com>
As cpu_idle.S is the only bit of code that is using the SCB asm defines,
so to allow us to remove scb.h in the future lets move the defines that
are used just into cpu_idle.S
Jira: ZEP-1568
Change-Id: I3c3a6f145ec4c1a43f076d079d5fe1694c255b78
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Kill of nvic.h and use either CMSIS helper functions for NVIC or direct
NVIC register access via CMSIS for IRQ handling code.
Jira: ZEP-1568
Change-Id: If21910b9293121efe85c3c9076a1c2b475ef91ef
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Move the exception priority related defines into exc.h out of nvic.h
Change-Id: I7ded917a3f6f7cdbc506b2f70c22f37eddc5f5c7
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Replace _ScbExcPrioSet with calls to NVIC_SetPriority as it handles both
interrupt and exception priorities. We don't need to shift around the
priority values for NVIC_SetPriority.
Jira: ZEP-1568
Change-Id: Iccd68733c3f7faa82b7ccb17200eef328090b6da
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Renames the flash security section so it makes sense for other Kinetis
devices, not just k64. In Kinetis reference manuals, this section is
referred to as the 'flash configuration field'.
Change-Id: I2b7c7cc1ec2541419d77878d367d96c9ceb7a0cf
Signed-off-by: Maureen Helm <maureen.helm@nxp.com>
The functions that implement usb_request_handlers are already using an
int32_t for transfer_len, so lets make the typedef match. This address
a potential issue in the future when the typedef of int32_t changes (or
when building with newlib).
Change-Id: I6e478551c38f2040b0dcec47c2e4c565c27acdd0
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Added structure definition for stream end points
and the a2dp stream.
Change-Id: I6d0cc08611f5179397bea6200eb9244d7c1cc8d6
Signed-off-by: Piyush Itankar <piyush.t.itankar@intel.com>
This is mainly for backward compatibility with 1.0b devices and for
spec compliance. CFC is mandatory post 1.0b spec where in MSC FC
shall not be used.
FC bit in MSC is used to manage the flow control. If FC is 1 then
the device is unable to accept frames.
Implementation is done by reusing "tx_credit" as a binary semaphore
wherein it will be blocked if MSC is recieved with FC bit 1 and
unblocked if FC bit is 0. Once tx thread is scheduled then semaphore
should be always available until all the buf in queue is sent.
Change-Id: I91181668ec0f46ff0b02905dd97e4503fc1fa7a7
Signed-off-by: Jaganath Kanakkassery <jaganathx.kanakkassery@intel.com>
Add SENSOR_CHAN_*_XYZ enum values to be used instead of
SENSOR_CHAN_*_ANY, because the new naming takes into account that
they are used to fetch all the 3 axes for a channel (X, Y and Z)
and not just any number of them.
Also deprecate old SENSOR_CHAN_*_ANY enum values.
Change-Id: I59e9901c1f8879d084bdb7c95583c2b28aa1e025
Signed-off-by: Bogdan Davidoaia <bogdan.davidoaia@linaro.org>
The cause for this change is TCP. Until now, the radio strategy driver
(ALOHA or CSMA) was providing the actual nbuf, and not the buffer
fragment, counting on the fact that the loop was using
net_buf_frag_del() which made so, iteration after iteration, buffer
framgent to be always buf->frags. The problem with this logic is loosing
the fragments that might be still referenced by TCP, in case the whole
buffer did not make it so TCP can retry later and so on.
Instead, TX now takes the nbuf and the actual frag to send. It could
have been working with just a pointer on the data, and the whole length
of the frame. But it has been avoided due to possible future devices,
that will be smarter and run CSMA directly in the hw, thus it will
require to access the whole buffer list through the nbuf.
Change-Id: I8d77b1e13b648c0ec3645cb2d55d1910d00381ea
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
Telnet "Interpret As Command" (IAC) code is ignored.
Change-Id: I882397389d77b8adfcbce62fbd9654c0b0412ae3
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
Let's use the generic console input type now. This will be useful for
other console input drivers such as telnet.
Change-Id: I787a1e9d86481d5f8c4803453726d9042a89dea4
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
Input line length can be modified via Kconfig.
Change-Id: I3423fce9814e04b11d11e5d391f85fe1efbe8d17
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
As for IN6ADDR_ANY_INIT, let's have an INADDR_ANY_INIT.
Change-Id: I07c9ec6d2bb20d3a228edaac2e3380942feac5fd
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
Check if input parameters are valid or not. It might lead to crash
NULL address input.
Change-Id: Ib446ab0467268bca01f478cca3ece868c7c9e49b
Signed-off-by: Ravi kumar Veeramally <ravikumar.veeramally@linux.intel.com>
The net_context_connect() callback was being invoked synchronously
with the transmission of the SYN packet. That's not very useful, as
it doesn't tell the user anything they can't already figure out from
the return code. Move it to the receipt of the SYNACK instead, so the
app can know that it's time to start transmitting. This matches the
Unix semantics more closely, where connect(2) is a blocking call that
wakes up only when the connection is live.
Change-Id: I11e3cca8572d51bee215274e82667e0917587a0f
Signed-off-by: Andy Ross <andrew.j.ross@intel.com>
Fold the OPEN_DRAIN GPIO interface into the DRIVE_STRENGTH interface.
The latter is more flexibile, suporting a larger range of hardware
capability.
Change-Id: I7ee01c8d375e3e4fe8ad10c075e365d246e83c75
Signed-off-by: Marcus Shawcroft <marcus.shawcroft@arm.com>
Provide a mechanism to configure different pin drive strengths on
hardware with that capability.
Configuration flags are provided to select each of DFLT, ALT and
DISCONNECT drive strengths independently for low and high outputs.
This provides sufficient flexbility to configure all of the drive
strength capability in at least nRF5 hardware.
The flags are chosen such that in the absence of a drive strength
flag a driver selects DFLT drive strength. This ensures that an
existing application that omits a flag continues to observe the
preexisting behaviour.
The behaviour of the drive strength flags is documented such that a
driver for hardware that does not support a particular drive strength
will simply defalt to the standard drive strength for that hardware.
Change-Id: I9894cc5e739a1899a4ecf795f2a5980b95b0c7a0
Signed-off-by: Marcus Shawcroft <marcus.shawcroft@arm.com>
The K_<thread option> flags/options avaialble to users were hidden in
the kernel private header files: move them to include/kernel.h to
publicize them.
Also, to avoid any future confusion, rename the k_thread.execution_flags
field to user_options.
Change-Id: I65a6fd5e9e78d4ccf783f3304b607a1e6956aeac
Signed-off-by: Benjamin Walsh <walsh.benj@gmail.com>
This will be needed for some thread user options that will move to
kernel.h since they are part of the user API.
Change-Id: I46e302b6cafcdddbad3458134b98feb5b8d45d9b
Signed-off-by: Benjamin Walsh <walsh.benj@gmail.com>
Replace _scs_relocate_vector_table with direct CMSIS register access and
use of __ISB/__DSB routinues. We also cleanup the code a little bit to
just have one implentation of relocate_vector_table() on ARMv7-M.
Jira: ZEP-1568
Change-Id: I088c30e680a7ba198c1527a5822114b70f10c510
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
As a first step towards removing the custom ARM Cortex-M Core code
present in Zephyr in benefit of using CMSIS, this change replaces
the use of the custom core code with CMSIS macros in
enable_floating_point().
Jira: ZEP-1568
Change-id: I544a712bf169358c826a3b2acd032c6b30b2801b
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Support using CMSIS defines and functions, we either pull the expect
defines/enum from the SoC HAL layers via <soc.h> for the SoC or we
provide a default set based on __NVIC_PRIO_BITS is defined.
We provide defaults in the case for:
IRQn_Type enum
*_REV define (set to 0)
__MPU_PRESENT define (set to 0 - no MPU)
__NVIC_PRIO_BITS define (set to CONFIG_NUM_IRQ_PRIO_BITS)
__Vendor_SysTickConfig (set to 0 - standard SysTick)
Jira: ZEP-1568
Change-Id: Ibc203de79f4697b14849b69c0e8c5c43677b5c6e
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
In preperation for removing the scb/scs layers and using CMSIS directly
lets remove all the _Scb* and _Scs* functions that are not currently
used.
Jira: ZEP-1568
Change-Id: If4641fb9a6de616b4b8793d4678aaaed48e794bc
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Now that we have a more generic mcux spi driver that can be used across
multiple Kinetis SoCs, remove the specific k64 spi driver.
Jira: ZEP-1374
Change-Id: Ifc324374f305837f5e3d2cfd7ad30d3608865b5b
Signed-off-by: Maureen Helm <maureen.helm@nxp.com>
On other targets, CONFIG_TEXT_SECTION_OFFSET allows the entire image to
be moved in memory to allow space for some type of header. The Mynewt
project bootloader prepends a small header, and this config needs to be
supported for this to work.
The specific alignment requirements of the vector table are chip
specific, and generally will be a power of two larger than the size of
the vector table.
Change-Id: I631a42ff64fb8ab86bd177659f2eac5208527653
Signed-off-by: David Brown <david.brown@linaro.org>
Implementation includes adding some defines in the pinmux,
adjusting gpio driver to specific defines for STM32F3X family,
adding specific functionality in the F3X SoC definition.
Change-Id: I465c66eb93e7afb43166c4585c852e284b0d6e67
Signed-off-by: Adam Podogrocki <adam.podogrocki@rndity.com>
_NvicSwInterruptTrigger is only utilized by a testcase for irq handling
on ARM-V7M. Just put the code into the testcase so we dont need to
support an additional interface.
Change-Id: I763c63c32a7a52918250458351d08b8fa54069dd
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
The k20_sim.h used by the PWM driver has some defines that also exist in
the MCUX HAL. Lets pickup the values from the HAL and drop the one's
that are in k20_sim.h. Also, they aren't used by anything at this
point.
Change-Id: Iaed4e8e5cec7d57a5ce9e89480cecfb7dc90a5e6
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
It is useful that the user API can know whether the connection
was established properly or not. So this commit adds status
parameter to connect callback in net_context API.
The call to connect callback needs to be set properly in TCP
code. This commit does not fix the connect callback call which
is not properly done right now in net_context.c.
Change-Id: I284a60ddd658ceef9e65022e96591f467a936a09
Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
If the parameter "timeout" is set in net_context_connect(), the
assumption by the user is that the function would wait for SYNACK
to be received before returning to the caller.
Currently this is not the case. The timeout parameter is handed
off to net_l2_offload_ip_connect() if CONFIG_NET_L2_OFFLOAD_IP is
defined but never handled in a normal call.
To implement the timeout, let's use a semaphore to wait for
tcp_synack_received() to get a SYNACK before returning from
net_context_connect().
Change-Id: I7565550ed5545e6410b2d99c429367c1fb539970
Signed-off-by: Michael Scott <michael.scott@linaro.org>
net_context is used for more than just TCP contexts. However,
the accept_cb field is only used for TCP. Let's move it from
the generic net_context structure to the TCP specific net_tcp
structure.
Change-Id: If923c7aba1355cf5f91c07a7e7e469d385c7c365
Signed-off-by: Michael Scott <michael.scott@linaro.org>
The recv_data_wait field in struct net_context is described as a "Mutex"
when in fact it's a semaphore signal.
Change-Id: I3bef8d1a07ceb3da5894ae4cdc8f1fe3c61c5dbe
Signed-off-by: Michael Scott <michael.scott@linaro.org>
