[Bug tapsets/10712] New: standardize syscall trace rendering

["fche at redhat dot com" <sourceware-bugzilla at sourceware dot org>]

We currently have three ways of intercepting syscalls
(kernel.function, kprobe.function(), process.syscall),
and may well get a few more (kernel-side tracepoints,
glibc uprobes).  It would be nice if the tapset logic
that collected/formatted the various syscall arguments
were more standardized.

Specifically, it would be nice if the "argstr" and "retstr"
values were calculated by some shared code (some auxiliary
function), so that the basic scenario of syscall tracing
produced apprx. identical results, whether done by
unprivileged users via utrace or by sysadmins via kprobes.

Exposing the same parameter variables is also desirable,
but a little more tricky (since at the utrace syscalls
level, all syscalls currently come through one probe,
though see bug #6613).

-- 
           Summary: standardize syscall trace rendering
           Product: systemtap
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: tapsets
        AssignedTo: systemtap at sources dot redhat dot com
        ReportedBy: fche at redhat dot com


http://sourceware.org/bugzilla/show_bug.cgi?id=10712

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