PB, request for code review and merge back(2)

Hello,

PhracturedBlue wrote: As I said. I've used it. It really doesn't feel that weird to me.

I agree with PB ....

Thanks for your unbelievable work!

PB, the keyboard is just a minor issue and I can change it as you want.
Currently I have another critical issue wanting your help.
I spent a whole day to port the telemetry config and telemetry test(telemetry monitor） pages. The telemetry config page is all done and test wiel with real devo10. However, after I flashed dfu with telemtest page into devo10, its LCD got screwed up and the firmware just doesn't boot.

Now I figure out the something is wrong inside the PAGE_TelemtestEvent(). If I have codes in this function commented out, everything is just fine, otherwise, the devo10 won't boot.
I have checkin codes into my repo. Please check them out (of coz don't merge to your repo now) and see if there are hardware conflicts between current telemetry settings and devo10. Thanks

PhracturedBlue wrote: Sorry, I didn't get to this tonight. looking at the code though, it seems that you are running GUI_Redraw() on objects that aren't necessarily on screen, and that is likely to cause issues.

I check all GUI_Redraw usages, the only place that might have problem is GUI_Redraw(listbox->scrollbar) in the GUI_DrawListbox method, however, this line won't be hit until entering the Load model or reorderlist pages. But the current issue is the devo10 cannot even boot.

I reproduced this issue in another way, if you checkout rev. 898 from my repo and build a dfu, the devo can't boot. Then you can try rev.898, commented out few updates related to the Telemetry stuffs--ported from devo8's main_page, it just works

PhracturedBlue wrote: both 898 and 899 work for me. We are starting to get low on memory though (Deviation uses a lot of stack space). It might be a good idea to try doing a:
make clean TARGET=devo10
make prd TARGET=devo10

That will do an optimized build, and if it works it would be useful info.
Also, please provide the dfu and list file for a build that is not working so I can compare (difference in compile version could be what is causing the discrepency)

You are right (but your previous command make prd is incorrect 。 The optimized build works, so we are in short of ROM

So 2 more questions :
1) Initially, I also suspected that the build might reach the limitation of ROM, however, the MCU's sheet mentions it has 256KB ROM and 48K RAM , but the debug build's size is only about 240KB and 22KB RAM. Where are the missing ROM?



2) Eventually the release build should be optimized , hence the release build 's size is only about 140KB, do we need to remove some redundant codes(all my traditional mixer files have been cut 2 days ago)?

Anyway, thanks for your help, the telemetry pages should be done today and request for your code review again

PB, please take time to do code review
1) Synced with your latest update
2) Telemetry config and monitor pages are ported, but I don't have an RX to test against. I also want you to provide an option to let user fully shut-off or turn-on the telemetry monitoring in the TX
3) Improve the appearance of the scroll bar
4) Change the orientation of the horizontal bar: 100 is to the left and -100 is to the right, aligning the orientation of the AIL/RUDD sticks
5） Fix bugs

Not only the AIL/RUDD sticks, but also their trims' bar are showing in reverse direction against sticks' movement. I am still looking for a better solution.

synced with your merged codes and verified that current consumption returns to normal even if putting the safety and binding checking back into the original 100ms loop. So the debug build slows down the program and does cause performance issue.

I just fixed some new found bugs, please have them merged to your trunk
1) Telemtest_page.c: should initialize scroll_bar, or the TX will hang when keeping pressing UP/DOWN in the 1st telemetry monitor page
2) Label.c: add 1 more line spacing
3) In the main.c , currently you set the initial brightness to 9, which leads to a visible flash light in booting the TX as most users might set the brightness to lower than 5. Just want to know what could happen in the devo8 if this initial value is set to 1 or 2

So far, all deviation 8 features , except for the scanner, are in the devo10, I am working on auto-dimm feature. The remaining known issue is that the more current consumption after manual binding, please let me know once you have further solution for it.