* [Bug translator/1304] support command line args for scripts
[not found] <20050906161228.1304.fche@redhat.com>
@ 2006-02-17 4:29 ` fche at redhat dot com
2006-02-23 22:07 ` fche at redhat dot com
1 sibling, 0 replies; 2+ messages in thread
From: fche at redhat dot com @ 2006-02-17 4:29 UTC (permalink / raw)
To: systemtap
------- Additional Comments From fche at redhat dot com 2006-02-17 04:29 -------
It occurs to me that we will need *two* kinds of parametrization after all.
The first would be a parse/translate-time one, to enable a generic script file
to apply to e.g. user-specified probe points. Since probe point expansion
occurs during elaboration, this parameter substitution would have to take place
by then. Perhaps something like dtrace's "$1" etc. notation could be used,
right within the parser (actually, the scanner), and would allow:
probe kernel.function("$1") { ... }
probe timer.ms($2) { ... }
One could get creative
A separate parametrization would be useful for compiled .ko reuse. Since such
parameters can't reasonably affect probe points, they can't take the place of
literals. One simplish possibility is to allow initialization-time assignment
to global systemtap variables (scalars) from standard linux module parameters.
This could be made to happen using nothing other than clever MODULE_PARAM type
macro calls being emitted near the globals' declarations, and removal of their
current unconditional initialization.
--
What |Removed |Added
----------------------------------------------------------------------------
AssignedTo|fche at redhat dot com |systemtap at sources dot
| |redhat dot com
Status|NEW |ASSIGNED
http://sourceware.org/bugzilla/show_bug.cgi?id=1304
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
^ permalink raw reply [flat|nested] 2+ messages in thread