Remove HAL_Init and related code from SystemInit and move it to
mbed_sdk_init. The function SystemInit is called early in the boot
sequence before RAM is initialized or the VTOR is setup, so it should
not be used to perform the HAL initialization.
This fixes crashes due the vector table being used before it has been
relocated.
The function _eth_arch_low_level_input() is meant to pass data into
LWIP and to prepare the ethernet buffers to receive more data.
If the LWIP heap is empty and the call to pbuf_alloc() in
_eth_arch_low_level_input returns null, the ethernet receive buffers
are not updated to receive data. Because of this the ethernet RX
interrupt will not fire. Since the RX interrupt is the only thing that
triggers a call to _eth_arch_low_level_input(), the receive buffers
will never get cleared, and the device stops receiving data.
To prevent this from happening, this patch ensures that the function
_eth_arch_low_level_input() clears the receive buffers even if a new
pbuf for the data couldn't be allocated.
This issue can be reproduce by running the test
"features-feature_lwip-tests-mbedmicro-net-udp_echo_parallel"
and on the same machine running the below python script to flood the
device with UDP broadcast packets:
MY_IP = #ADD your local IP here
from socket import *
s = socket(AF_INET, SOCK_DGRAM)
s.bind((MY_IP, 1234))
s.setsockopt(SOL_SOCKET, SO_REUSEADDR, 1)
s.setsockopt(SOL_SOCKET, SO_BROADCAST, 1)
for _ in range(1000):
s.sendto("test data", ('255.255.255.255', 1234))
print("Message sent")
This test was removed because of its memory footprint.
Most of the memory consummed by the test itself was related to the instantion of test cases in RAM.
With the introduction of case_t which allows tests cases to live in FLASH, the memory footprint has been drastically reduced (860 bytes of RAM saved) and the ticker test can be reintroduced in the code base.
This creates a macro for the UUID length used by Greentea. This cuts
down on the use of "magic numbers" in test cases that use
the GREENTEA_SETUP_UUID function.
This is a workaround for IAR's lack of flexibility with memory regions.
Otherwise these tests would use very little heap and be mostly global
allocations.
This matches the timeout used in linux:
https://linux.die.net/man/5/dhclient.conf
This resolves several issues noticed during testing when we
have a very large number of devices that try to get an IP address
around the same time.
Before, the UDP test was very strict on the number of packets it would
try to match before failing. Now it will keep trying for the whole test
to get enough passing packets. It also includes the test's UUID so you
can validate which packets are being received.
The dtls test already has the ability to retry upon a UDP failure.
However the sockets are currently configured to be blocking and to wait
forever. I added a timeout of 1.5 seconds in order the test to correctly
timeout.
Initialize all values of timeinfo in make_time_info. This prevents
the field 'tm_isdst' from getting inadvertently set to 1 causing
time to be off by 1 hour.
Latest cmsis files provide virtual nvic implementation, therefore all nvic
set/get vectors were removed. As the result, we did not reallocate vectors
for mbed SDK. This should fix it for most of the platforms (cortex m0 and
cortex a9 need to provide own if they need it).
* Make memory sections configurable in linker files
* Dynamically determine vector location in flash for NVIC relocation
* Advertise bootloader support in targets.json
Martin Kojtal
Russ Butler
Anna Bridge
Yuguo Zou
Jimmy Brisson
Vincent Coubard
Mahadevan Mahesh
Brian Daniels
Christopher Haster
Sam Grove
Aksel Skauge Mellbye
Mika Leppänen
Kevin Gilbert
Andrzej Puzdrowski
Laurent MEUNIER