dev: Add PERF_ATTACH_TASK_DATA unconditionally for all has_branch_stack()

Short running processes i.e those getting small cpu run time each time when
they get scheduled on might not accumulate much branch records before a PMU
IRQ really happens. This increases the possibility for the process to loose
much its branch records while it gets scheduled in and out of various cpus.

There is a need to save all occured branch records during the cpu run time
while the process gets scheduled out. It requires a event context specific
buffer for such storage.

This adds PERF_ATTACH_TASK_DATA flag unconditionally, for all branch stack
sampling events which would allocate task_ctx_data during an event init.
This also creates a platform specific kmem cache to serve such allcation
requests.

Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>