Bootloader merging fails with python3, [ERROR] 'l' format requires -2147483648 <= number <= 2147483647

[teetak01]

Description

Severity: Critical (breaks build with limited workaround), cannot work in CI-automation.

Mbed OS master #31da50e8fa68a2bbe8c92d70b258a5d19a41b5e3 (19.02.2019)
mbed-cli 1.9.1
Python 3.6.1
ARM Compiler 6.11 Professional or GCC 6.3.1
K64F

mbed-cloud-client-example (internal development version)

Bootloader merging fails with:

Merging Regions
  Filling region bootloader with /home/teetak01/devel/mbed-cloud-client-example-internal/tools/mbed-bootloader-k64f-block_device.bin
[ERROR] 'l' format requires -2147483648 <= number <= 2147483647
[mbed] ERROR: "/usr/bin/python3" returned error.

This fails two times, and third retry in compilation passes.

Merging Regions
  Filling region bootloader with /home/teetak01/devel/mbed-cloud-client-example-internal/tools/mbed-bootloader-k64f-block_device.bin
  Filling region header with ./BUILD/K64F/ARMC6/mbed-cloud-client-example-internal_header.hex
  Filling region application with ./BUILD/K64F/ARMC6/mbed-cloud-client-example-internal_application.bin
Space used after regions merged: 0x6df34
Merging Regions
  Filling region application with ./BUILD/K64F/ARMC6/mbed-cloud-client-example-internal_application.bin
Space used after regions merged: 0x63b34

Issue request type

[ ] Question
[ ] Enhancement
[x] Bug

[0xc0170]

There was this fix but caused some other failures #8788, reverted via #8915

This is not a regression caused in the recent hours? I've reviewed the logs, not much landed to tools in the last days

cc @ARMmbed/mbed-os-tools

[teetak01]

I tested this with Mbed OS from last friday, and this morning.

[bridadan]

Hi @teetak01 I wasn't able to reproduce this, here's my environment:

Windows 10
Mbed CLI 1.9.1, installed on python 3.6.3
Hash for mbed-cloud-client-example-internal: 991357060e45569393ded95d6dc37b4f438df3f2
Hash for mbed-os: 31da50e
Command: mbed compile -m K64F -t ARMC6

Do you have a setup that reproduces the issue consistently?

[cmonr]

@bridadan Sanity check. Could you retry the previous command with -v and -vv?

Iirc, I came across this problem last week (was retesting this: #8146 (comment)), and this odd behavior led me to opening #9675

[theotherjimmy]

@cmonr That was related to the time stamp, so running with -v was eliminated as the cause.

[bridadan]

(for completeness I ran it with -v and -vv, no change)

[ciarmcom]

Internal Jira reference: https://jira.arm.com/browse/MBOCUSTRIA-892

[teetak01]

@bridadan Ubuntu 18.04.1, Python 3.6.7, fresh installation. Happens quite reliably on my setup.

[teetak01]

Also happens with GCC 6.3.1

[bridadan]

I had a spare moment to setup an Ubuntu VM. I am able to reproduce the error!