[Bug bpf/24656] New: Target variables for tracepoints differ between traditional sytemtap backend and bpf backend

wcohen at redhat dot com sourceware-bugzilla@sourceware.org
Mon Jun 10 14:35:00 GMT 2019


https://sourceware.org/bugzilla/show_bug.cgi?id=24656

            Bug ID: 24656
           Summary: Target variables for tracepoints differ between
                    traditional sytemtap backend and bpf backend
           Product: systemtap
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: bpf
          Assignee: systemtap at sourceware dot org
          Reporter: wcohen at redhat dot com
  Target Milestone: ---

After getting the target() function working I expected that the cycle_thief.stp
example would work with the bpf backend.  However, found that the bpf backend
has a different set of target variables available than the original systemtap
kernel module backend:

[wcohen@localhost systemtap]$ ../install/bin/stap -L
'kernel.trace("sched_switch")'
kernel.trace("sched:sched_switch") $preempt:bool $prev:struct task_struct*
$next:struct task_struct*
[wcohen@localhost systemtap]$ ../install/bin/stap --bpf -L
'kernel.trace("sched_switch")'
kernel.trace("sched:sched_switch") $prev_comm:char[] $prev_pid:pid_t
$prev_prio:int $prev_state:long int $next_comm:char[] $next_pid:pid_t
$next_prio:int

Did not see this type of difference between probes for a kernel function. For
example the vfs_read probes are the same:

[wcohen@localhost systemtap]$ ../install/bin/stap -L
'kernel.function("vfs_read")' 
kernel.function("vfs_read@fs/read_write.c:446") $file:struct file* $buf:char*
$count:size_t $pos:loff_t*
[wcohen@localhost systemtap]$ ../install/bin/stap --bpf -L
'kernel.function("vfs_read")' 
kernel.function("vfs_read@fs/read_write.c:446") $file:struct file* $buf:char*
$count:size_t $pos:loff_t*

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


More information about the Systemtap mailing list