[miniclient] align external repository for miniclient with feature branch #5726
Conversation
functions in all projects to work properly.
mbed HAL API and NRF52 prototype implementation. Stdout can be send to SWO using platform.swo-enable and serial can be disabled using platform.stdio-enable.
- move swo init to SystemInit (in system_nrf52.c) - add swo init code that resumes tracing after device reset - change swo enable to use already existing macro name
The serial serial_free() tried to compare address of a member
of a struct, instead of the struct itself. This gives a warning
on GCC&ARMCC, but a error on IAR.
The fix here is not beautiful, a proper one would likely
avoid the memcpy() on serial_init() and have a proper way of
identifying the stdio-uart.
---8<---8<---8<---
Compile [100.0%]: serial_api.c
[Error] serial_api.c@382,0: [Pe042]: operand types are
incompatible ("serial_t *" and "struct serial_s *")
[ERROR]
if(&stdio_uart == &(obj->serial))
The nanostack hal's critical section uses a mutex for mutual exclusion, which is nice for many use cases. But when one needs to use the critical section from interrupts, the RTX will have a assertion failure and panic. Add a configurable for mbed_lib, which can be used to enable a alternative version of critical section, which uses the underlying OS primitives, which disables the interrupts. Note: the default behavior is not changed, one needs to override the "nanostack-hal.critical-section-usable-from-interrupt" to have "true". Reason for this change is that there is a need for sending events using nanostack event queue from interrupt context, eg. from a socket callback.
marcuschangarm
changed the title
|
@dannybenor FYI |
|
This PR passes at least my tests on K64F & NRF52_DK boards. |
|
The SWO-for-stdout implementation in here for mbed_retarget.cpp is a variant of a more-generic "alternate stdio retarget" proposal already put forward by #5356, to which I responded with an maximally-generic proposal in #5571. (The submitter of #5356 then rewrote his example using #5571, which is a bit confusing for the history, but means it is now a good example implementation of it). #5571 isn't currently making progress, but it should be borne in mind. No need to do anything on your feature branch now, but if #5571 does go in to master, the conflict will need to be cleared up - you could instead do the SWO output without touching mbed_retarget.cpp by providing an Or if the SWO interface is thought to be generically-applicable enough, there could be a JSON option that to use the SWO FileHandle built-in to mbed_retarget.cpp, as for the existing serial. |
|
@kjbracey-arm thank you for the heads up! My plan is to make a proper SWO HAL API using your redirect once it's been merged. |
Build : SUCCESSBuild number : 889 Triggering tests/morph test |
Exporter Build : SUCCESSBuild number : 562 |
Test : FAILUREBuild number : 734 |
|
The latest test results -flash tests fail (the addition here) |
|
@marcuschangarm Any update on this one ? We need to get this in otherwise we cannot move our development to mbedOS repo. |
Fix legacy size from NRF51.
|
/morph build |
Build : SUCCESSBuild number : 960 Triggering tests/morph test |
|
@yogpan01 sorry for the delay! I was a bit stumped by why enabling flash writes would suddenly deplete the heap! 😨 Let's see if this makes the CI happy! |
Exporter Build : SUCCESSBuild number : 645 |
Test : FAILUREBuild number : 787 |
|
Great, now the NUCLEO_F746ZG fails! |
|
#5419 also is seeing spurious(?) fails on lp_ticker tests on Nucleo F401. Do we have a general problem on master? |
There was a patch few days ago, since then havent seen these. I would rebase to get the latest master in |
|
/morph test |
Test : FAILUREBuild number : 800 |
|
Please take a look at test around the NUCLEO_F746ZG with this PR. More specifically, low power timer test are failing. |
|
I'm seeing low power tests failing on Nucleo boards on #5419. They seem spurious to me, or at least unrelated to either actual PR. Are they showing up more widely? Looks like either a precision problem or interrupt latency spikes. (Ah, I did already comment above, and apparently there's a master patch. Let's both try rebasing). |
0xc0170
force-pushed
the
feature-miniclient
branch
from
945d75f
to
ca04eff
Compare
|
Subset of changes merged into new feature branch: https://github.com/ARMmbed/mbed-os/tree/feature-cloud-client-lite |
This PR takes the external fork and aligns it with the miniclient feature branch.
These changes are not meant to be compatible with mbed-os/master but will find their way there through separate pull-requests once we are happy with the results.
Current changes include: