We have two rt1050 evaluation boards modified to use qspi flash instead of the default hyper flash. One board was modified over a month ago and was fine until recently. One day, performance started to degrade in that the J-Link download dialog would repeatedly program the same area of flash before starting a debug session. Eventually it would fail altogether with something like
It is also impossible to debug an application that is resident in the external DRAM, but it is possible to debug an application that completely sits in the internal RAM of the rt1050.
The second evaluation board was modified only days ago but showed this behavior from the beginning.