This is the mail archive of the
systemtap@sources.redhat.com
mailing list for the systemtap project.
Re: architecture paper draft
- From: William Cohen <wcohen at redhat dot com>
- To: "Frank Ch. Eigler" <fche at redhat dot com>
- Cc: systemtap at sources dot redhat dot com
- Date: Fri, 28 Jan 2005 10:56:37 -0500
- Subject: Re: architecture paper draft
- References: <20050127212504.GH22921@redhat.com>
Frank Ch. Eigler wrote:
Hi -
I committed a first draft version of one proposed architecture in the
systemtap CVS repository. Check out the "archpaper" module using
anoncvs [1]. Or use a real ssh account so you can eventually contribute
directly (have you signed up yet?). Just for gags, temporarily
(since it's such an early draft) I put formatted version of the
LaTeX paper up at home:
<http://web.elastic.org/~fche/systemtap.pdf>
- FChE
[1]
cvs -d :pserver:anoncvs@sources.redhat.com:/cvs/systemtap login
{enter "anoncvs" as the password}
cvs -d :pserver:anoncvs@sources.redhat.com:/cvs/systemtap co archpaper
General comments on the systemtap paper
Clearer to label section "3.1 Input" as "SystemTap Probe Language"
If indirection is not allow then, then how does SystemTap probe data
structures in the kernel that use pointers or arguments that are
passed in to functions as pointers (assuming using jprobes).
The language seems verbose in placing the probe. For example:
break kernel.function("sys_write")
The equivalent for DTrace would be
fbt::sys_write:entry
The first is the instrumentation provider, fbt which appears to be for
the kernel only. For userspace there is the pid provider so the format
would be ($1 is passed in from the command line):
pid$1::sys_write:entry
Might want to use some a keyword besides "break" to indicate placement
of probe. C already uses that keyword. Are switch-case statements not
going to be language?
-Will