[Bug tapsets/3833] New: nfs_proc.stp and LKET/nfs_proc.stp tapset files broken

["mmlnx at us dot ibm dot com" <sourceware-bugzilla at sourceware dot org>]

During testing on SLES 10 we found numerous failures in the nfs_proc.stp and
LKET/nfs_proc.stp tapset files.  They appear to be the result of conditionals
like this from nfs_proc.stp:

    733 probe nfs.proc2.read_done = kernel.function("nfs_read_done")?,
    734                             module("nfs").function("nfs_read_done")?
    735 {
    736 %( kernel_v >= "2.6.10" %?
    737         server_ip = __nfs_read_data_info($data,0)
    738         prot = __nfs_read_data_info($data,1)
    739         count = __nfs_read_data_info($data, 2)
    740 %:
    741         server_ip = __nfs_read_data_info($task->tk_calldata,0)
    742         prot = __nfs_read_data_info($task->tk_calldata,1)
    743         count = __nfs_read_data_info($task->tk_calldata, 2)
    744 %)

The code assumes the "data" parameter appears in nfs_read_done() as of 2.6.10,
but a scan of previous major releases shows that "data" isn't a parameter until
2.6.17.

Other errors are probably the result of similar assumptions.  Someone needs to
go through these files and fix the conditionals to work on all versions.

-- 
           Summary: nfs_proc.stp and LKET/nfs_proc.stp tapset files broken
           Product: systemtap
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: tapsets
        AssignedTo: systemtap at sources dot redhat dot com
        ReportedBy: mmlnx at us dot ibm dot com


http://sourceware.org/bugzilla/show_bug.cgi?id=3833

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