ambd_sdk_with_chip_non_NDA/integration_notes.txt at master · hank820/ambd_sdk_with_chip_non_NDA · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
2021-02-26
I add three files in to src/Platform which reference from ESP32 and make compile pass.
Now the InitServer looks work, i can get "Event Received", “No Event” and “Event Timeout” from GenericPlatformManagerImpl_FreeRTOS.cpp by printf.
I can’t get output if use “ChipLogError”.

Because i don't know how does chip example work.
Please let me know if there is anything I need to do.
or I will take finish src/Platform/* functions as my next goal.

Thanks.


2021-02-25
I moved the application files back to the project_hp/Makefile.  When building the library it was trying to build everything.
I kept the CPP options flags in place in case they are needed for building the library.  But this keeps building the library independant of the application.

in src/lib/mdns/minimal/Query.h  setting the private member mClass = to QClass::IN fails.
very strange because setting it to any of the other enumerated values works
maybe a namespace thing but would seem odd since it seems to be properly namespaced.
Can you have a look at this further.

I got side tracked by the other side of the project for most of the afternoon:
I solved the start if InitServer()  I got rid of 1 unresolved symbol, Yeah, but now have 5 new ones.
I had planned to keep going on that but got side tracked.

I think that we should put in our backlog, and maybe sooner rather than later to have all the chip #defines in a seperate part of the make file.
I think we may be running into issues where the app has one configuration and the library another.
I do think that we need to keep the build of the library and the application independant.

It's still broken but some good progress today.


2021-02-25
1.
Update toolchain from 6.4.1 to 9.3.0.

2.
I found the file which cause _EXFUN macro problem. It come from “component/soc/realtek/amebad/swlib/string/strproc.h”
I commented codes which cause the problem. Will confirm with my colleague why define these.


2021-02-24

Implement AMEBADUtils.cpp basic function.
1. Implement basic funtion, without AP mode.
2. Need to revise behavior to match ConnectivityManagerImpl.cpp
TODO(Hank) : Add ConnectivityManagerImpl.cpp

Set $CHIPDIR in Makefile.include.gen, change repo path from here if needed.


2021-02-23
@Hank the patches Pass TestSystemTimer and TestSystemObject, and TestCHIPTLV : replace time by xTaskGetTickCount
will not work in general for chip.  We need to be only making changes withing the src/platform/AmdbaD directory.
I have put up a question on one of the chip slack channels regarding making such changes.  I may have more information later today


CHIP_ERROR PlatformManagerImpl::_InitChipStack(void)
Wifi initialization stuff that is in main() of the example app needs to be migrated into here


esp_event_loop_create_default
I am thinking that there is something analogous to the default event loop for AmebaD.

esp_event_handler_register is listed as to be deprecated in the esp documentation.
We should discuss how the system events for wifi are handled by AmebaD SDK

Updated src/platform/AmebaD to support the logging function - just uses printf()

Current state is that this will build and run
Chip Thread Started is shown.
I added some debug code in the chip thread function, I could see that it was being run, but I am not getting any events. So seems to be as expected

uncommenting the include server.h and appcallbacks and the 2 calls at the bottom of the ChipTest function cause the build to break.

I started by adding the server directory which then triggered a bunch of other build failures.

I am thinking the best thing to do is to go to your (Hank) gn build method, while it is a big clunky but we could be working with the more recent build of the code.
I have a branch in my fork which includes the Thermostat cluster which is my end game.

So the next goal is to be able to link against something that will build with all the commented out code in chipinterface.cpp
I would suggest doing this from my fork in the tstat-ambd branch

I am committing everything as is now to provide a good starting point if I don't get to work on this more tonight.