This is the mail archive of the systemtap@sourceware.org mailing list for the systemtap project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

[Bug translator/5667] Exploit kprobe.symbol_name for function boundary tracing


------- Additional Comments From srinivasa at in dot ibm dot com  2008-01-28 05:27 -------
(In reply to comment #2)

> 
> If this symbol list is acquired via elfutils during translation,
> then either the existing relocation mechanism (section+offset)
> versus the kprobes-specific (symbol-name) may be selected.  As
> you see from the code, in the attachment, there is little
> difference.

Frank
 Out intention is,  If user doesn't have the debuginfo package and he wants do 
function boundry tracing then kprobe-specific symbol-name approach helps.

In that direction, We need to read just function name from the script and 
assign it to kp.symbol_name. Then kernel API's will resolve the symbol_name to 
address and probes the instruction.

Please let me know your views on this.

Thanks
Srinivasa DS


-- 


http://sourceware.org/bugzilla/show_bug.cgi?id=5667

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]