Hiya,
I've spent a number of frustrating hours trying to create an SDK that will work with the Linux version of MCUXpresso.
I'm using the online tool, selecting "Linux" as the target and "GCC ARM Embedded" as the Toolchain.
Here is a screen shot of the build information from the SDK dashboard:
I am downloading the SDK, un gunzipping it, un tarring it and trying to drag the folder or a .zip file into the "Install SDKs" view at which point I get asked if I'm sure I want to import it followed by a message saying it was unable to load the SDK. When I look at the MCUXpressoIDE's error log (regardless of whether I am trying to load the folder, .zip or even the .tar.gz), I get:
eclipse.buildId=unknown
java.version=1.8.0_201
java.vendor=Oracle Corporation
BootLoader constants: OS=linux, ARCH=x86_64, WS=gtk, NL=en_CA
Command-line arguments: -os linux -ws gtk -arch x86_64
MCUXpresso IDE SDK import
Error
Tue Jan 07 16:12:14 EST 2020
The following items are not supported or not valid SDKs
[/home/myke/LinuxSDK/SDK_2.3.1_MK22FN1M0Axxx12.tar.gz]
I have no idea why that line is at the bottom as, as noted, I have ungunzip'd and untar'd the file I downloaded so this is in one of the various files. I have attached the file that was downloaded from the SDK site. I am running Ubuntu 18.04.03 LTS
I should also point out that this happens if I use something like a Freedom board SDK.
I'm having no trouble with Windows (but want to transfer my software development from Windows 7 to Linux).
I'm sure I'm doing something simple wrong but I've got no idea what it is - any help is appreciated!
myke
Thank you for the reply - I'll give it a try.
No, one minor point of confusion all the documentation says that the latest version of MCUXpressoIDE is v11.1.0 - but when I look at "About" for MCUXpressoIDE for Linux 0 it says that it's version v11.0.1 as you can see from this screen shot:
What is the correct version number for the Linux version of MCUXpressoIDE? When I click on "Check for Updates" is says "No updates found".
Thanx,
myke
Mr. Verse,
Thank you for your reply. I have made some progress but I need some more help when it comes ot the PEmicro Multilink Universal programmer.
First off, I downloaded MCUXpressoIDE on January 7th of this year. I hadn't deleted the tab yet and here it is:
When I saw your email, I went to the NXP MCUXpressoIDE download page today and discovered that the version was now different:
Any ideas what's happening here? The Linux PC was booted on January 7th and maybe it picked up an cached page but I have no idea how that happened - I've been on a Windows PC primarily (which does have v11.1.0) and started moving to the Linux system. I had not loaded in any previous versions on to this system.
Loading SDKs Under Linux:
Regardless of the version weirdness, after installing v11.1.0, I was able to load in the same SDK I was using on my Windows machine (this was for an MK20DN512VLL10, which we're using for our product) and then build one of the projects I created. What does the "Host OS" mean in this case? Regardless, you might want to make it a bit more intuitive for the user so that if they're going to use an SDK with MCUXpressoIDE, then they shouldn't touch anything.
Could I suggest that you disable the "Host OS" dropdown unless the user selects anything other than "MCUXpressoIDE" as the "IDE/Toolchain"? I tried a number of approaches to making different SDKs, all with the "Host OS" being "Linux" and I can tell you categorically none of them work.
PEmicro Multilink Universal programming/debugging problem:
I connected my PEmicro Multilink Universal which seemed to be recognized by the system without any issues (I did not change anything). Clicking on "Debug As" ==> "PEmicro probes" the programming operation seemed okay and the buttons for debugging became enabled. I started the program running but it stopped immediately with the log messages being:
Final - program crash after clicking on "Run":
eclipse.buildId=unknown
java.version=1.8.0_201
java.vendor=Oracle Corporation
BootLoader constants: OS=linux, ARCH=x86_64, WS=gtk, NL=en_CA
Command-line arguments: -os linux -ws gtk -arch x86_64
org.eclipse.cdt.dsf.gdb
Error
Sat Jan 11 14:57:39 EST 2020
Failed to execute MI command:
-data-list-register-values x 25
Error message from debugger back end:
Thread id: 1 has terminated
java.lang.Exception: Thread id: 1 has terminated
at org.eclipse.cdt.dsf.mi.service.command.AbstractMIControl$RxThread.processMIOutput(AbstractMIControl.java:993)
at org.eclipse.cdt.dsf.mi.service.command.AbstractMIControl$RxThread.run(AbstractMIControl.java:822)
Second, Part of the Load:
eclipse.buildId=unknown
java.version=1.8.0_201
java.vendor=Oracle Corporation
BootLoader constants: OS=linux, ARCH=x86_64, WS=gtk, NL=en_CA
Command-line arguments: -os linux -ws gtk -arch x86_64
com.nxp.mcuxpresso.trace.core
Info
Sat Jan 11 14:57:26 EST 2020
PEMicro continuous memory service session connected on port 10224->35808 @ 127.0.0.1
First message:
eclipse.buildId=unknown
java.version=1.8.0_201
java.vendor=Oracle Corporation
BootLoader constants: OS=linux, ARCH=x86_64, WS=gtk, NL=en_CA
Command-line arguments: -os linux -ws gtk -arch x86_64
com.nxp.mcuxpresso.trace.core
Info
Sat Jan 11 14:57:26 EST 2020
SWO session connected on port 10224->35806 @ 127.0.0.1
If I try single stepping through the application, I get the following message when I try to step over a second "GPIO_TogglePinsOutput" macro:
eclipse.buildId=unknown
java.version=1.8.0_201
java.vendor=Oracle Corporation
BootLoader constants: OS=linux, ARCH=x86_64, WS=gtk, NL=en_CA
Command-line arguments: -os linux -ws gtk -arch x86_64
org.eclipse.cdt.dsf
Error
Sat Jan 11 15:05:27 EST 2020
Request for monitor: 'RequestMonitor (org.eclipse.cdt.dsf.mi.service.MIRunControl$9@43c555aa): Status ERROR: org.eclipse.cdt.dsf.gdb code=10001 Context cannot be suspended. null' resulted in an error.
followed by:
eclipse.buildId=unknown
java.version=1.8.0_201
java.vendor=Oracle Corporation
BootLoader constants: OS=linux, ARCH=x86_64, WS=gtk, NL=en_CA
Command-line arguments: -os linux -ws gtk -arch x86_64
org.eclipse.cdt.dsf.gdb
Error
Sat Jan 11 15:05:27 EST 2020
Context cannot be suspended.
Note that changing the USB cable, like the SEGGER.
I should also point out that the operation of the PEmicro Multilink Universal is pretty marginal on the Windows version of MCUXpressoIDE v11.1.0. It is rock solid reliable under CodeWarrior.
SEGGER J-Link programming/debugging problem Okay:
I connected my Segger J-Link which seemed to be recognized by the system without any issues (I did not change anything). Clicking on "Debug As" ==> "SEGGER J-Link probes" the programming operation seemed okay but I got a number of errors saying that there were communications issues/"Failed to execute MI command".
On a hunch, I decided to try a shorter USB cable (I was using a 10 foot cable) and replaced it with a 2 metre cable and things started working fine and repeatedly.
So, if you get someone irate like me saying that things don't work for their SEGGER J-Link, ask about the USB cable - this "fix" doesn't seem to be anywhere on the web, I will put it on the SEGGER forumas a suggestion for other people.
I should also point out that the J-Link's operation under Windows is a bit spotty with, interestingly enough, the program doesn't seem to save in Windows (ie removing the J-Link, resetting the board or powering up and powering down, it won't come back up) - this is not an issue under CodeWarriior.
I don't know what you want to do with this, but I thought it would be helpful.
myke
