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/10849] make MAXSKIPPED overflow trigger an error message


------- Additional Comments From fche at redhat dot com  2009-10-28 23:17 -------
> Previous patch does not work if there is no probe end {} in the script. 

Why is that?


> Just adding the error messages to the appropriate places results in a println
> for each skipped probe, so I used an 'atomic_read(& session_state)' and a
> comparison to STAP_SESSION_ERROR to determine whether or not to print the
> error.

A better way might be to use 

atomic_compxchg(& skipped_count, STAP_SESSION_RUNNING, STAP_SESSION_ERROR)


> The problem, though, is that if another error sets session state to
> STAP_SESSION_ERROR, then the skipped message will not be printed. The only
> error that does this is when error_count > MAXERROR -- here if we encounter
> error_count > MAXERROR before skipped > MAXSKIPPED then the latter will not
> print any error message.

That's not so bad.  What was worse is there appearing only a warning to
indicate exceeding MAXSKIPPED.


-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|systemtap at sources dot    |chwang at redhat dot com
                   |redhat dot com              |
             Status|NEW                         |ASSIGNED


http://sourceware.org/bugzilla/show_bug.cgi?id=10849

------- 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]