This patch adds function-return probes (AKA exit probes) to kprobes. 
 When establishing a probepoint at the entry to a function, you can also 
establish a handler to be run when the function returns.

The subsequent post give example of how function-return probes can be used.

Two new registration interfaces are added to kprobes:

int register_kretprobe(struct kprobe *kp, struct rprobe *rp);
Registers a probepoint at the entry to the function whose address is 
kp->addr.  Each time that function returns, rp->handler will be run.

int register_jretprobe(struct jprobe *jp, struct rprobe *rp);
Like register_kretprobe, except a jprobe is established for the probed 
function.

To unregister, you still use unregister_kprobe or unregister_jprobe. To 
probe only a function's returns, call register_kretprobe() and specify 
NULL handlers for the kprobe.

The following fields of struct retprobe are of interest to the user:

handler - This function is run after the ret instruction executes, but 
before control returns to the return address in the caller.

maxactive - The maximum number of instances of the probed function that 
can be active concurrently.  For example, if the function is 
non-recursive and is called with a spinlock or mutex held, maxactive = 1 
should be enough.  If the function is non-recursive and can never 
relinquish the CPU (e.g., via a semaphore or preemption), NR_CPUS should 
be enough. maxactive is used to determine how many rprobe_instance 
objects to allocate for this particular probed function.  If maxactive 
<= 0, it is set to a default value.  See register_kretprobe().

nmissed - Initialized to zero when the rprobe is registered, and 
incremented every time the probed function is entered but there is no 
rprobe_instance object available for establishing the function-return probe.

kprobe - When the rprobe is registered, this field is set to the kprobe 
at the entry to the function.


Signed-off-by: hien Nguyen <hien@us.ibm.com>