[Fwd: Re: [PATCH] Return probe]

Hien Nguyen hien@us.ibm.com
Wed Apr 27 00:49:00 GMT 2005


Vara Prasad wrote:

>
>
>> Just a thought if a probe inside the tampoline can be avoided, since 
>> we change
>> the return address on stack to tampoline and now within the tampoline 
>> we can call the handlers. Then the return from tampoline can be 
>> changed to actual return address.  We may need to get the pt_regs 
>> structure and avoid preemption during the handler execution .
>>
>> Do you feel that there are some performance benefits by avoiding int3 
>> in a tampoline?
>>  
>>
> I will let Jim/Hien answer this one.
>
Actually Jim has thought about it, we have to code the trampoline for 
all different architectures (the code probably similar to int3). We 
avoid that in the expense of couple traps.



More information about the Systemtap mailing list