New release New firmware release 1.2.2.b1

[mohpor]

@daniel
Hi again,
Are we still up for today's update? (despite of the early 1.2.2.b1 update)

[livius]

@bmarkus
this is only not updated web page label
updater tool update to version expected 1.2.2.b1

[bmarkus]

On the download page current firmware version is 1.2.0.b1 see

https://www.pycom.io/support/supportdownloads/

[daniel]

@pdp8user thanks. Yes, please let us know the details about the issues that you are experiencing so that we can analyze the problem.

[pdp8user]

@daniel Thanks for that and the prompt reply.

Keep up the good work! The memory issues have been frustrating, because as @mohpor says, something would work some times and not others.

I'm still having some issues in that respect, which is why the question about the finalizer. I've still to clarify the details, then I'll report it to you.

Cheers,
Brian

[daniel]

@pdp8user it's automatic.

[daniel]

@mohpor cool :-) thanks for the feedback!

Cheers,
Daniel

[mohpor]

@daniel

Wonderful update I gotta admit. Fixed a lot of memory problems! One problem that has been fixed and I haven't see anybody reporting was that you could run your code 2 times and then it would crash on the 3rd try. It is now fixed and I can upload my code as many time as I want to! Fantastic.

Cheers,
M.

[pdp8user]

Do we have to call the finalizer explicitly, or is it automatic?

[daniel]

@jmarcelino thanks for the report, we are investigating why those crashes are occurring. We'll let you know once we discover the root cause.

Cheers,
Daniel

[jmarcelino]

OK first let me say the new release is a massive improvement, before this same code would crash after only 140 or so Alarm events. However it's still crashing without fail in the 2500-3000 range.

Note I may well be doing something wrong as I just started with MicroPython this Xmas!

First the crash:

[...some output here]
Elapsed 2690 s 
Hourly average 22.84 mAh
Consumed 17.07
Elapsed 2691 s 
Hourly average 22.83 mAh
Consumed 17.07
Elapsed 2692 s 
Hourly average 22.82 mAh
Guru Meditation Error of type InstrFetchProhibited occurred on core  0. Exception was unhandled.
Register dump:
PC      : 0x00692cc0  PS      : 0x00060430  A0      : 0x800ecdd4  A1      : 0x3ffc4bb0  
A2      : 0x3ffc5490  A3      : 0x00000001  A4      : 0x00000000  A5      : 0x3ffc4bd0  
A6      : 0x400dc2cc  A7      : 0x3ffc1340  A8      : 0x800ecd90  A9      : 0x3ffc4b90  
A10     : 0x3ffc5490  A11     : 0x00000001  A12     : 0x00000000  A13     : 0x3ffc4bd0  
A14     : 0xb33fffff  A15     : 0x3ffc4b70  SAR     : 0x0000001c  EXCCAUSE: 0x00000014  
EXCVADDR: 0x00692cc0  LBEG    : 0x4000c2e0  LEND    : 0x4000c2f6  LCOUNT  : 0xffffffff  

Backtrace: 0x00692cc0:0x3ffc4bb0 0x400ecdd4:0x3ffc4bd0 0x400dc2dc:0x3ffc4c00 0x400e0e3e:0x3ffc4c20 0x400d81d8:0x3ffc4ca0

Rebooting...

and my code (shortened version I'm using for testing)

from machine import Pin
from machine import Timer
import micropython 
micropython.alloc_emergency_exception_buf(100)

class LTC4150(object):
    ah_quanta = 0.17067759
    
    def __init__(self):
        self.p_in = Pin('P15', mode=Pin.IN, pull=None)
        self.p_in.callback(Pin.IRQ_FALLING, self.__pin_handler) 
        self.chrono = Timer.Chrono()
        self.chrono.start()
        self.count = 0
        self.time = 0 
        self.lastTime = 0 
        
    def __pin_handler(self, arg):
        self.count = self.count + 1



def periodic_handler(alarm):
    mA = cc.count* cc.ah_quanta
    seconds = cc.chrono.read()
    print("Consumed %.2f" % mA ) 
    print ("Elapsed %d s " % seconds)
    print("Hourly average %.2f mAh" % ((3600*mA)/seconds))


cc = LTC4150()

alarm = Timer.Alarm(periodic_handler, 1, periodic=True)

If I increase the period say to 10s it still crashes just takes 10x as long, so I think it's more to do with the number of events

Thanks for the help

[robert-hh]

As a hint for everyone who is building the image her- or himself:
Please update the pycom-esp-idf too. Otherwise the build will not run.

[bmarkus]

@jmarcelino said in New release New firmware release 1.2.2.b1:

@bmarkus

>>> import gc
>>> gc.mem_free()
42464

Thanks, getting better.

[jmarcelino]

@bmarkus

>>> import gc
>>> gc.mem_free()
42464

[bmarkus]

How much free RAM is available?

[jmarcelino]

Oh this is a very welcome early bug fix release.

I was trying some periodic Alarm code which was crashing with a Guru Meditation error at about every 142 events on the previous beta. Happy to report that's now running well into the thousands and still ticking away nicely.

This even before I even had a proper chance to report it! Amazing :-)