Accessing compilation errors #284

royemmerich · 2014-02-13T23:25:28Z

So I am thrashing away trying to get MODBUS communication working...and it is not...yet :( I have one BIG script with all of the modbus_tk python library in it to compensate for the inability to import other python modules.

I also added some LED blinking to see if things are actually working. Initially they didn't blink so I commented out all the MODBUS related code (numerous classes, function definitions, etc.), re-uploaded the script and they started to blink again. Eventually through trial and error I figured out it was this piece of code which was causing things to break (specifically line 3):

1 _hooks = {}
2 def install_hook(name, fct):
3     _hooks[name].append(fct)

This got me thinking:

How is one supposed to know when there is a compilation error after uploading a script and what the error is?

P.S. Until importing is working and the python standard libraries are better represented it is looking more feasible to implement the MODBUS stuff in C and call it from python. I'll give that a bash over the next few days.

JonHylands · 2014-02-13T23:51:07Z

So, if you are using an actual pyboard (or a discovery board, or like me rolled your own) you will usually get a compile error show up on the USB console.

If you use my brother's usb-ser-mon, it automagically re-connects as soon as the console is available (after flashing), so you won't miss anything.

Jon

----- Original Message -----

How is one supposed to know when there is a compilation error after
uploading a script and what the error is?

pfalcon · 2014-02-14T07:04:51Z

for the inability to import other python modules.

There's no such known inability (with #282 applied which cover last "obscure" bit). If you have issues with imports, please submit a ticket with a testcase which shows the issue.

How is one supposed to know when there is a compilation error after uploading a script and what the error is?

Everyone would have different opinion for this one, so here's mine: you should do all porting, debugging, testing, etc. using "unix" port. For "unix" port, such questions don't hold - you get complete (with tracebacks) error reporting in split seconds. For hardware access, you should have appropriate mock/semihosting modules. Once you have testsuite which does ~100% code coverage, you move to device, where it will just work.

dpgeorge · 2014-02-16T00:30:28Z

If there is an error that throws an exception, then it should print out on the USB console. You don't have to be connected to the console when the error happens, it should buffer the output and you will see it when you connect (eg picocom /dev/ttyACM0).

royemmerich · 2014-02-17T09:56:39Z

I have tried both picocom and minicom and have yet to see any compilation error printed to the console. I prefer minicom as it doesn't drop the connection during a script upload and automatically reconnects after I hit the reset button on my STM32F407 discovery board.

Based on my little import test script in #304, I modified the print line to something I know creates an issue in python 3:

This works 👍

print(counter)

And this doesn't work 👎

print counter

However I don't see any compilation error in the second case.

JonHylands · 2014-02-17T15:42:46Z

Using my brother's usb-ser-mon, I get the following:

USB Serial device with vendor 'STMicroelectronics' serial '00000000050C' connected @/dev/ttyACM0
Use Control-C to exit.
File "0:/src/test.py", line 2 column 7
SyntaxError: invalid syntax
Micro Python build on 25/1/2014; PYBv4 with STM32F405RG
Type "help()" for more information.

My test.py program has the following line (after a blank line):

print "Hello"

This is running on my custom STM32F405 based MicroPython board.

What are you running on?

You can't use normal terminal emulators - they don't connect fast enough when the board is powered or reset.

Clone this repo: https://github.com/dhylands/usb-ser-mon

Read and follow the readme, and then see what happens.

Jon

----- Original Message -----

I have tried both picocom and minicom and have yet to see any compilation
error printed to the console.

Based on my little import test script in #304, I modified the print line
to something I know creates an issue in python 3:

This works 👍
print(counter)
And this doesn't work 👎
print counter
However I don't see any compilation error in the second case.

Reply to this email directly or view it on GitHub:
#284 (comment)

JonHylands · 2014-02-17T17:37:21Z

Yes, I'm aware of that. I was trying to force a syntax error, to show that the compile errors show up on the USB console, which is what the point of this issue is...

----- Original Message -----

This is Python 3, and print is a function - try:
print("hello")
instead.

Reply to this email directly or view it on GitHub:
#284 (comment)

jugglerchris · 2014-02-17T17:39:50Z

@JonHylands Yeah, sorry - I replied having seen the bug e-mail, without spending the time to read the rest of the bug. My bad - I deleted the comment immediately afterwards, but presumably notifications still went through.

dhylands · 2014-02-17T19:00:14Z

I think that exceptions which occur as part of boot.py or main.py might not
be getting reported

Dave Hylands
Shuswap, BC, Canada
http://www.davehylands.com
On Feb 17, 2014 9:37 AM, "Jon Hylands" notifications@github.com wrote:

Yes, I'm aware of that. I was trying to force a syntax error, to show that
the compile errors show up on the USB console, which is what the point of
this issue is...

----- Original Message -----

This is Python 3, and print is a function - try:
print("hello")
instead.

Reply to this email directly or view it on GitHub:

#284 (comment)

Reply to this email directly or view it on GitHubhttps://github.com//issues/284#issuecomment-35305989
.

JonHylands · 2014-02-17T19:22:36Z

On my board, I've got the following in my boot.py:

print("Executing uCee's boot.py")
pyb.source_dir("src")
pyb.main("test.py")

So test.py is the equivalent of main.py in this case...

Jon

----- Original Message -----

I think that exceptions which occur as part of boot.py or main.py might not
be getting reported

pfalcon · 2014-02-22T16:37:13Z

@royemmerich: Did suggestions provided help? Can we close this ticket?

royemmerich · 2014-02-23T21:57:42Z

Did suggestions provided help?

Making use of @dhylands usb-ser-mon python serial app I can now see the compilation errors after copying my file to the /SRC directory and hitting the reset button on my STM32F4 discovery board.

However after @dpgeorge's comment:

You don't have to be connected to the console when the error happens, it should buffer the output and you will see it when you connect (eg picocom /dev/ttyACM0)

which doesn't seem to work, then I would consider usb-ser-mon a (rather good) workaround and this still an unfixed bug.

dhylands · 2014-02-24T02:11:21Z

The USB serial console doesn't seem to buffer characters.

Using usb-ser-mon connected to a UART (and main.c modified appropriately),
I always see all of the startup messages on the HW UART. But only sometimes
do I see all of the messages from a hard RESET on the USB serial console.

I haven't yet established if the fault is in usb-ser-mon or in the
firmware. I should be able to tell by doing a wireshark capture.

Dave Hylands

On Sun, Feb 23, 2014 at 1:57 PM, Roy Emmerich notifications@github.comwrote:

Did suggestions provided help?

Making use of @dhylands https://github.com/dhylands usb-ser-monhttps://github.com/dhylands/usb-ser-monpython serial app I can now see the compilation errors after copying my
file to the /SRC directory and hitting the reset button on my STM32F4
discovery board.

However after @dpgeorge https://github.com/dpgeorge's comment:

You don't have to be connected to the console when the error happens, it
should buffer the output and you will see it when you connect (eg picocom
/dev/ttyACM0)

which doesn't seem to work, then I would consider usb-ser-mon a (rather
good) workaround and this still an unfixed bug.

Reply to this email directly or view it on GitHubhttps://github.com//issues/284#issuecomment-35845085
.

pfalcon · 2014-04-26T14:21:21Z

I assume this was resolved/clarified, and there were further changes to console code to make other concerns obsolete. If issues still hold, please reopen/resubmit.

pfalcon added the question label Feb 15, 2014

royemmerich mentioned this issue Feb 17, 2014

Module import not functioning as expected on STM #304

Closed

pfalcon closed this as completed Apr 26, 2014

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

Accessing compilation errors #284

Accessing compilation errors #284

royemmerich commented Feb 13, 2014

JonHylands commented Feb 13, 2014

Uh oh!

pfalcon commented Feb 14, 2014

Uh oh!

dpgeorge commented Feb 16, 2014

Uh oh!

royemmerich commented Feb 17, 2014

Uh oh!

JonHylands commented Feb 17, 2014

Uh oh!

JonHylands commented Feb 17, 2014

Uh oh!

jugglerchris commented Feb 17, 2014

Uh oh!

dhylands commented Feb 17, 2014

Uh oh!

JonHylands commented Feb 17, 2014

Uh oh!

pfalcon commented Feb 22, 2014

Uh oh!

royemmerich commented Feb 23, 2014

Uh oh!

dhylands commented Feb 24, 2014

Uh oh!

pfalcon commented Apr 26, 2014

Uh oh!

Uh oh!

Accessing compilation errors #284

Accessing compilation errors #284

Comments

royemmerich commented Feb 13, 2014

JonHylands commented Feb 13, 2014

Uh oh!

pfalcon commented Feb 14, 2014

Uh oh!

dpgeorge commented Feb 16, 2014

Uh oh!

royemmerich commented Feb 17, 2014

Uh oh!

JonHylands commented Feb 17, 2014

Uh oh!

JonHylands commented Feb 17, 2014

Uh oh!

jugglerchris commented Feb 17, 2014

Uh oh!

dhylands commented Feb 17, 2014

Uh oh!

JonHylands commented Feb 17, 2014

Uh oh!

pfalcon commented Feb 22, 2014

Uh oh!

royemmerich commented Feb 23, 2014

Uh oh!

dhylands commented Feb 24, 2014

Uh oh!

pfalcon commented Apr 26, 2014

Uh oh!