This is the mail archive of the
systemtap@sourceware.org
mailing list for the systemtap project.
Re: Accessing target variables in return probes (was Re: src ChangeLog tapsets.cxx)
- From: "Frank Ch. Eigler" <fche at redhat dot com>
- To: "Stone, Joshua I" <joshua dot i dot stone at intel dot com>
- Cc: systemtap at sources dot redhat dot com
- Date: Sun, 17 Dec 2006 14:08:12 -0500
- Subject: Re: Accessing target variables in return probes (was Re: src ChangeLog tapsets.cxx)
- References: <C56DB814FAA30B418C75310AC4BB279D01117ABA@scsmsx413.amr.corp.intel.com>
Hi, Josh -
> > [stack-pointer-sensitive kretprobe skip detection]
> [...] Have you prototyped this at all?
Nope. I figure one might simulate it with an explicit .return probe
that periodically skips processing.
> We also discussed the possibility of extending kretprobes with a
> pre_handler -- a handler that is called during entry, but only if
> the return-probe instance was successfully allocated. [...]
If that were the only condition under which data for return & entry
probes might become unsynchronized, that could be sufficient. It
would be nice if the early prologue debugging information were good
enough to extract parameters at that point.
Perhaps a way around this is to have the translator test the debugging
info at the entry point address. If it fails for some reason, then
fall back to prologue analysis. Depending on the reliability of this
heuristic, this could apply to all .function() probes.
- FChE