pkg/semtech_loramac: fix memcpy to uninitialized pointer [backport 2019.07]

[fjmolinas]

Backport of #11847

Contribution description

#11552 introduced an illegal memcpy to an uni-initialized pointers. This broke loramac for some boards e.g: lobaro-lorabox and broke ABP for most of them. This pr fixes the illegal access.

The PR should be backported to Release 2019.07.

Testing procedure

• On im880b (using boards: add support for im880b lora board #11315) or lobaro-lorabox on master the following causes a hardfault:

BOARD=lobaro-lorabox LORA_DRIVER=sx1272 make -C tests/pkg_semtech-loramac/ flash term

loramac save
2019-07-16 11:05:16,557 - INFO #  loramac save
2019-07-16 11:05:16,562 - INFO # [semtech-loramac] saving configuration on EEPROM
2019-07-16 11:05:17,010 - INFO # [semtech-loramac] saving uplink counter: 0 
2019-07-16 11:05:17,042 - INFO # [semtech-loramac] saving rx2 freq: 869525000
2019-07-16 11:05:17,070 - INFO # [semtech-loramac] saving rx2 dr: 3
reboot

with this PR it works as expected.

• On b-l072z-lrwan1 and other boards in master trying to set session keys fails or hangs:

loramac set appskey 1EBB10A1AE58065BB4540A4181DC7BB8
2019-07-15 17:16:52,784 - INFO #  loramac set appskey 1EBB10A1AE58065BB4540A4181DC7BB8
loramac get appskey
2019-07-15 17:16:55,244 - INFO #  loramac get appskey
2019-07-15 17:16:55,248 - INFO # APPSKEY: E9030008000000000000000000000000

With this PR it succeeds.

• Release tasks 11-lorawan succeed with b-l072z-lrwan1:
• Task Add inet_pton and inet_ntop (POSIX.1-2001) #1 - LoRaWAN example
• Task Start a RIOT #2 - OTAA join procedure
• Task Interrupt-Handling on msp430 is buggy #3 - ABP join procedure
• Task Remove swtimer #4 - LoRaWAN device parameters persistence

Issues/PRs references

fixes #11626
alternative to #11783

[jia200x]

I confirm it's the same commit as aff7ee3

[jia200x]

it's pointing to the right branch and everything looks OK. ACK

[fjmolinas]

@jia200x rebased since 868b55c was merged into 2019.07-branch

[fjmolinas]

@jia200x rebased since 868b55c was merged into 2019.07-branch

@jia200x your review somehow got dismissed while doing so... could you take a look again (should just mean ACK'ing again, nothing has changed).

[jia200x]

Sure! ACK