Hardfault during re-initialization of Cordio stack/link-layer

[donatieng]

Issue originally raised by @ryannowarm: https://github.com/MarceloSalazar/mbed-os-example-ble-GAP/issues/2

== The system has been rebooted due to a fatal error. ==
DEVICE MAC ADDRESS: d8:f3:64:8a:07:d2

Scanning started (interval: 2ms, window: 2ms, timeout: 0ms).
We have scanned for 5999ms with an interval of 4 timeslots and a window of 4 timeslots
We have been listening on the radio for at least 5997ms

Starting next GAP demo mode

++ MbedOS Fault Handler ++

FaultType: HardFault

Context:
R0   : 00000000
R1   : 00000000
R2   : 200070CC
R3   : 00014095
R4   : 20007EEC
R5   : 20002340
R6   : 20001940
R7   : 00000000
R8   : 0001F89D
R9   : 0001EE05
R10  : 20008E68
R11  : 2000C728
R12  : 20002318
SP   : 20008DF8
LR   : 00019581
PC   : 00016DA6
xPSR : 010F0000
PSP  : 20008D90
MSP  : 2003FFC0
CPUID: 410FC241
HFSR : 40000000
MMFSR: 00000082
BFSR : 00000000
UFSR : 00000000
DFSR : 00000008
AFSR : 00000000
MMFAR: 00014095
Mode : Thread
Priv : Privileged
Stack: PSP

-- MbedOS Fault Handler --



++ MbedOS Error Info ++
Error Status: 0x80FF013D Code: 317 Module: 255
Error Message: Fault exception
Location: 0x3D7
Error Value: 0x16DA6
Current Thread: main  Id: 0x20008F3C Entry: 0x1A275 StackSize: 0x1000 StackMem: 0x20007F08 SP: 0x2003FF58 
For more info, visit: https://armmbed.github.io/mbedos-error/?error=0x80FF013D
-- MbedOS Error Info --ˇ
== The system has been rebooted due to a fatal error. ==

[ciarmcom]

Internal Jira reference: https://jira.arm.com/browse/MBOCUSTRIA-272

[0xc0170]

This should be fixed by #8998