mcuxpresso debugging issues ...

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

mcuxpresso debugging issues ...

1,498 Views
frank_m
Senior Contributor III

Sorry for ranting about some debugger issues, perhaps it might be useful to others...


First, I had some strange installation-related issues under Windows.
Installing the IDE (V11.1.1_3241) and drivers went well, and building an example project went smoothly.
However, I was unable to debug when connected the target (an OM13098, MCU is LPC54628) for the first time.
The IDE found no debug probe, and the device manager told me I need to reboot the PC to reinstall the drivers (CMSIS-DAP). The PC had been rebooted more then once after installing the IDE, only the target (debug probe) was connected the first time.
After a reboot, the debug worked.
Host was a HP Z4 workstation with Win10, but I could replicate it on an older HP Z420 under Win8.2.
The funny thing is - I tried at home on a freshly installed Linux Mint 19.3 Mate 64 Bit (Athlon II X2 core), and not only was the installation much faster and without annoying queries, but debugging worked immediately, out of the box.

The second issue:
I have an application that accesses the external SD RAM on the OM13098 board.
With the debugger stopped at main(), I tried to open a memory view to display the SD RAM contents (address 0xA000.0000).
That was my bad, EMC and SD RAM were not initialized at this point, so the debugger error message was reasonable.
However, I had to acknowledge 3 or 4 consequential error messages afterwards, and the debugger stopped working at all.
I had to restart the IDE to be able to debug again.

On a positive note, the IDE works reasonably well otherwise.

Especially considering the fact I don't like Eclipse ...

0 Kudos
6 Replies

1,363 Views
millerbarbara93
Contributor I

I got sufficient information from your soul of heart.Thank you so much

tellpizzahut

0 Kudos

1,364 Views
frank_m
Senior Contributor III

Another thing I noticed - I obviously can't do instruction single stepping in the debugger. Which is quite unfortunate.

Is this a restriction of the IDE, or the debug interface (CMSIS-DAP) ?

0 Kudos

1,363 Views
converse
Senior Contributor V

Yes you can. It is described in the User Guide. (Have you read it?) Also in this blog

https://mcuoneclipse.com/2012/10/27/assembly-instruction-stepping/

(old version of eclipse but is the same in MCUXpresso)

0 Kudos

1,363 Views
frank_m
Senior Contributor III

Perhaps based on the misunderstanding of this:

pastedImage_1.png

I don't reall like Eclipse, especially it's convoluted terminology ...

I successfully added it to the menu now, and will check if it works ASAP.

0 Kudos

1,364 Views
converse
Senior Contributor V

My observations:

Installing drivers on Windows is a real pain. When a driver is installed, it doesn't necessarily make it available to use. It just tells Windows "when you come across this device, here is a driver to use". Plugging in the device is when the driver is really activated and loaded. At that point, there may be some things required by the driver that aren't currently available in Windows, hence the reboot.

Linux USB drivers work in a different way - the USB support is already present in the kernel and (most often) no further driver is required, so it 'just works'.

The install is much quicker because there are lots of hoops to jump through when installing on Windows (including lots of checking done by the OS), that just doesn't happen with Linux.

Regarding the debug - did you try 'clean up debug'? There is a button on the toolbar (red box with a black cross overlayed) that normally restores debug functionality without needing to restart the IDE.

0 Kudos

1,364 Views
frank_m
Senior Contributor III

In regard to Windows, I came across other toolchain during the years, and cannot remember to be it a persistent problem.

Nonwithstanding the numerous reboots during any installation under Windows ... But I'm pretty sure other toolchains did not show that problem.

Anyway, I had pinned it to Windows already, and put up with it.

> Regarding the debug - did you try 'clean up debug'? There is a button on the toolbar (red box with a black cross overlayed) that normally restores debug functionality without needing to restart the IDE.

I only tried the "Reset Target" option, which caused another error message I now forgot. Or was it "clean up debug" ... ???

I reckoned this would reset the core via the JTAG interface/periphera,l regardless of current core state.

I can just tell I have had some encounters with other Eclipse-based toolchains, with much worse debugging experiences ...

0 Kudos