Re: updated bootprobe

["Bryn M. Reeves" <breeves at redhat dot com>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Jose R. Santos wrote:
> Hy Bryn,
> 
> This looks like a systemcall and fork/exec trace.  Wouldn't it make

Hi Jose,

Yes - that's basically what the script does now. I was thinking of
extending it a bit to add some IO statistics, but that should still be
possible with the changes you suggest.

> sense to use the tracing facility already available in SystemTap and
> porting the Perl post-processing script to use that instead of rolling
> your own trace format?  I would love to have post-processing scripts
> that do the same thing for the SystemTap trace facility.

Many thanks for the pointer - I've only spent a very brief time looking
at LKET so far, and I agree that this is likely a better approach to
this problem.

I still think Dan's script has some didactic value as-is, but I will
definitely have a go at extending your example below to make more use of
stap & lket's own reporting abilities.

Kind regards,

Bryn.

> Please take a look at "man lket"
> 
> Here is the an example of a script that should be able to gather the
> information needed for generating the data that your post-processing
> creates.
> 
> ====================
> #! stap
> process_snapshot() {}
> addevent.tskdispatch.cpuidle {}
> addevent.process {}
> addevent.syscall.entry { printf ("%4b", $flags) }
> addevent.syscall.exit {}
> addevent.tskdispatch.cpuidle {}
> 
> =====================
> 
> -JRS

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iD8DBQFE03D66YSQoMYUY94RApSBAKCfVvV9zfprm4GDuC1BYq9h81/15wCfW7g3
LXB8NdUeCJw+agznUOtN+GM=
=GHwI
-----END PGP SIGNATURE-----