This is the mail archive of the
systemtap@sourceware.org
mailing list for the systemtap project.
Re: SystemTap Language Reference: Userspace probing patch
- From: Jim Keniston <jkenisto at us dot ibm dot com>
- To: Robb Romans <robb at linux dot vnet dot ibm dot com>
- Cc: systemtap at sourceware dot org, Prerna Saxena <prerna at linux dot vnet dot ibm dot com>
- Date: Wed, 15 Jul 2009 16:11:10 -0700
- Subject: Re: SystemTap Language Reference: Userspace probing patch
- References: <200907141809.45230.robb@linux.vnet.ibm.com>
On Tue, 2009-07-14 at 18:09 -0500, Robb Romans wrote:
> Hi,
>
> Here is a patch that adds information about userspace probing to the
> Language Reference Guide. Please review. I would like clarification on
> the final paragraph about PATH. Is it correct? Thanks.
>
> 02-LangRef-userspace_probing.diff (attached)
> Signed-off-by: Robb Romans <robb@linux.vnet.ibm.com>
>
>
> Regards,
> Robb
>
+Prototype support for \emph{userspace probing} is available in the
+form of the following non-symbolic probe point:
+
+\begin{vindent}
+\begin{verbatim}
+process(PID).statement(ADDRESS).absolute
+\end{verbatim}
+\end{vindent}
+
+This probe point is analogous to
+\texttt{kernel.statement(ADDRESS).absolute} in that both use raw
+(unverified) virtual addresses and provide no
+\texttt{\$variables}. The target PID parameter must identify a running
+process and ADDRESS must identify a valid instruction address. All
+threads of that process will be probed.
+
This (along with the stapprobes man page) seems about a year out of
date. Support for
probe process(PID).statement(ADDRESS).absolute
was dropped last summer, when Frank & co. added support for the
following:
probe process("PATH").function(PATTERN)
probe process("PATH").function(PATTERN).return
probe process("PATH").statement(PATTERN)
And pretty much all the context-variable constructs documented in the
(kernel) DWARF section above are also supported for user-space probes.
I'm no longer actively participating on the SystemTap team, but I doubt
that all this work has been backed out in the past couple of months.
You should probably get fche or dsmith to take a closer look at this.
Jim