* Added a figure explaining how layering of the posix arch
compares to the embedded builds
* One minor grammar change in last paragraph
Signed-off-by: Alberto Escolar Piedras <alpi@oticon.com>
In STM32 family, RNG IP is available only on F4 and L4 series.
Besides STM32F401 does not support is neither.
Get entropy driver available on STM32 devices supporting it and
generate a compilation issue on STM32 devices that do not own
the RNG IP.
Solves #5448
Signed-Off-By: Erwan Gouriou <erwan.gouriou@linaro.org>
On STM32L4 SoCs RNG is clocked by 48MHz domain.
Hence, besides clock activation, it requires 48M
domain to be enabled.
Tested on:
*nucleo_l476rg
*stm32l476g_disco
*disco_l475_iot1
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>
SPI_MODE_GET() returns a bitfield. It is thus wrong to test if a bit is
set using the equality operator. The bit-wise AND operator must be used
instead.
This can be tested by setting the SPI in mode 3 (CPOL + CPHA). Currently
both tests will fail and the result is a SPI configured in mode 0. This
was confirmed using an oscilloscope. Applying the patch fixes the
polarity.
Signed-off-by: Florian Vaussard <florian.vaussard@gmail.com>
We can have a spurious error while performing a transfer using IRQ. This
happens when the last message of the transfer is a read with a STOP
condition. We must disable the RX interrupt while waiting for the STOP
interrupt, otherwise we will get a spurious RX interrupt which will lead
to an error.
Signed-off-by: Florian Vaussard <florian.vaussard@gmail.com>
STM32 UART driver uses a macro for clock initialization, that
is difficult to read and incompatible with needed changes to
fix STM32F0 series UART problems.
This change switches to using the full clock bus names in UART
init functions removing the macro-magic and increasing
readability.
Signed-off-by: Daniel Wagenknecht <wagenknecht@clage.de>
The combination of
CONFIG_CLOCK_STM32_SYSCLK_SRC_PLL and
CONFIG_CLOCK_STM32_PLL_SRC_HSI
on SOCs with PREDIV1 support made use of the LL define
LL_RCC_PLLSOURCE_HSI_DIV_2, which is not defined for SOCs with
PREDIV1 support.
This exchanges LL_RCC_PLLSOURCE_HSI_DIV_2 with LL_RCC_PLLSOURCE_HSI
which is the appropiate source according to stm32f0xx_ll_rcc.h
line 473 and stm32f3xx_ll_rcc.h line 795.
Tested by compiling hello world for nucleo_f091rc board with HSI as
PLLSOURCE.
Signed-off-by: Daniel Wagenknecht <wagenknecht@clage.de>
Current implementation of LL_SPI_TransmitData16 on F0 family
generates following warning:
"warning: dereferencing type-punned pointer will break strict-aliasing
rules [-Wstrict-aliasing]"
Besides being forbidden by rule, this cast is not needed, as register is
16 bits wide. Modification has been tested on F0 SoC.
stm32yyxx_ll_spi.h being included in soc.h file, warning is generated
at each compiled object, this commit allows a clean build.
This issue is referenced in ST and tracked under
reference 13359. Code will be updated on upcoming stm32cube updates.
Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
The native_posix board will accept command line arguments
which the application / test may pick by calling
native_get_cmd_line_args()
Signed-off-by: Alberto Escolar Piedras <alpi@oticon.com>
When asserts are enabled we run into an issue with newlib and types of
printf style formatters not matching. The easy fix to this is to cast
the uint32_t to u32_t to make things consistent with or without newlib
enabled.
This fixes#5645
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
Re-order the execution of the arch/ and subsys/ CMakeLists.txt code to
work around a manifestation of issue #6505. When OpenThread created an
External project in subsys, it did not have access to important
toolchain flags added in arch/.
Intuitively, subsystems might depends on how the ARCH is configured,
but the ARCH shouldn't depend on any subsystem.
Signed-off-by: Sebastian Bøe <sebastian.boe@nordicsemi.no>
Instead of hard-coding the path use find_program(), since that allows
CMake to sort out platform-specific details such as .exe extensions on
Windows.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
The function returns an enum, not a u8_t, so we should cast
appropriately to avoid an implicit conversion.
This construct was triggering this compilation error when compiling
with CXX:
/home/sebo/zephyr/include/bluetooth/buf.h:85:9: error: invalid
conversion from ‘u8_t {aka unsigned char}’ to ‘bt_buf_type’
[-fpermissive] return *(u8_t *)net_buf_user_data(buf);
Signed-off-by: Sebastian Bøe <sebastian.boe@nordicsemi.no>
Compiling this declaration with a CXX compiler triggers the compiler
error:
/home/sebo/zephyr/include/bluetooth/gatt.h:898:10: error: ‘struct
bt_gatt_read_params::<anonymous union>::__single’ invalid; an
anonymous union can only have non-static data members [-fpermissive]
Reading up on the standard, I was unable to find any mention of this
being valid C or CXX code (But reading the standard is not
straightforward). And I was unable to find any mechanism to make the
CXX compiler accept it (e.g. Changing the -std, or adding this as a
language extension e.g. -fms-extensions).
So we rewrite it to not declare the struct with the tag
"__single". There does not seem to be any reason for it to be declared
like this.
Signed-off-by: Sebastian Bøe <sebastian.boe@nordicsemi.no>
Instead of accessing the environment variable ZEPHYR_BASE every time we
require accessing the source code root, use an intermediate variable
that has OS path separators correctly set to '/' to avoid issues on
Windows.
Note: This removes the ZEPHYR_SOURCE_DIR CMake variable. External
applications using that will need to change to use the new ZEPHYR_BASE
variable.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
The new images has package updates including:
- ninja-build
- lcov
- new versions of doc generations packages
- gcovr
- gcc-6-multilib
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
In order to support both "/" and "\" as OS path separators,
use the correct Python os.sep helper.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
I've been successfully using the latest sphinx/breathe/docutils
and doxygen versions for local doc build testing. The CI system
already uses the latest doxygen, so this patch updates the
pip-installed sphinx, breathe, and docutils tools too.
Signed-off-by: David B. Kinder <david.b.kinder@intel.com>
It has been agreed in the project TSC to reject commit messages without
any content. Every commit message needs some explaination beyond what
was put in the title, even the most trivial ones.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
When testing ping6 with net shell, it was noticed that after some
sucessive calls the applications stopped to handle rx packets.
Analyzing other icmpv6 register callbacks it was verified that is
necessary to unref packets before returning NET_OK.
Signed-off-by: Pedro Martucci <pedropaulomartucci@gmail.com>
commit 2a7546fb5a ("net: lwm2m: add support for coap2coap proxy")
erroneously changed the COAP_OPTION_* used to specify the coap2coap
or coap2http proxy resource used from COAP_OPTION_URI_PATH to
COAP_OPTION_PROXY_SCHEME.
Changing it back to COAP_OPTION_URI_PATH requires us to re-order how
the coap options are appended to the packet as the coap options must
be added in the order specified by the numbers in:
https://tools.ietf.org/html/rfc7252#section-12.2
Signed-off-by: Michael Scott <michael@opensourcefoundries.com>
CoAP packet w/ confirmation flag set is required to be retransmitted
before it got the ACK message from the peer.
However, the packet is usally unreference once it's sent to the network.
Although we set the timeout as no wait when calling function
net_app_send_pkt(), it's still possible that the packet is unreferenced
before we got a chance to increase the packet reference by calling
coap_pending_cycle().
Usually, the IP stack will generate an ARP packet first and then send
out the packet. However, this is not the case when the remote is a
loopback address.
As issue #5101 described, when asking client to perform a firmware pull
on URL "coap://127.0.0.1:7783/large". The packet will be unreferenced
immediately after calling net_app_send_pkt(). Which then result in
client hang.
The solution to the issue is to increase the reference count on the
sending packet and decrease it after the process is finished.
Signed-off-by: Robert Chou <robert.ch.chou@acer.com>