This is the mail archive of the
systemtap@sources.redhat.com
mailing list for the systemtap project.
Re: FW: recent kprobe work
- From: Ananth N Mavinakayanahalli <ananth at in dot ibm dot com>
- To: "Lynch, Rusty" <rusty dot lynch at intel dot com>
- Cc: systemtap at sources dot redhat dot com
- Date: Wed, 29 Jun 2005 14:25:56 -0400
- Subject: Re: FW: recent kprobe work
- References: <032EB457B9DBC540BFB1B7B519C78B0E077BCF6C@orsmsx404.amr.corp.intel.com>
- Reply-to: ananth at in dot ibm dot com
On Wed, Jun 29, 2005 at 10:56:31AM -0700, Lynch, Rusty wrote:
> >I believe Dave was referring to our use of arch_init() rather than
> >arch_init_kprobe() or somesuch.
> >>
> >> Anyone have a problem with me kicking out a patch to move all the
> arch_*
> >> function names to kprobes_arch_* ?
> >
> >Hmm, I'd prefer to have the _kprobe() suffix rather than the prefix, as
> >we do now (arch_prepare/arm/disarm_kprobe()). How about making the call
> >in question arch_init_kprobe/kretprobe() instead?
>
> How about arch_init_kprobes() <kprobes instead of kprobe>, since this
> function is meant to initialize the architecture specific sections of
> the kprobes infrastructure (as opposed to a specific kprobe or kretprobe
> or jprobe or what ever else is next)?
>
> ... and then don't touch any of the other arch_ functions. (I was
> thinking we had some other arch_* functions that would be namespace
> problems, but now I realize they all have a kprobe specific component to
> them (like arch_prepare_kprobe)).
Sounds good!
Ananth