andy git 06/15 suspend/resume observations

[Werner Almesberger]

Mike (mwester) wrote:
> The patch adds some paranoid tests to ensure that none of the remaining
> GSM suspend/resume handling code dereferences pointers without testing
> for NULL.

Hmm, is this just paranoia or do you have evidence that a NULL pointer
is indeed valid in those cases ? If it's just paranoia, it would be
better to trip over the NULL pointer as early as possible.

Just imagine how many more bugs would be caught every day if free(NULL)
wasn't valid :)

- Werner