This is the mail archive of the
systemtap@sourceware.org
mailing list for the systemtap project.
Re: [TAPSETS] Linux Kernel Event Trace Tool
Mathieu Desnoyers <compudj@krystal.dyndns.org> writes:
> > [...] The runtime functions callable from
> > probes should assume atomic calls: no reentrancy, no interrupts, and
> > so on.
> So it leaves the locking to the probe provider. You know that it
> limits the set of instrumentable kernel functions to the ones that
> can be protected by disabling interrupts.
Not exactly. Those functions could still in theory be instrumented,
but an actual run-time reentry into the probing system would be
detected, the probe short-circuited, and a "missed probe" would be
recorded.
> Page faults and NMI handlers are two examples where this locking
> scheme fails. All traps, exceptions and fault handlers cannot then
> be safely instrumented,
I am probably missing something. Could you sketch out a scenario that
causes a safety problem?
> or at the cost of doing each one a special case (for each
> architecture). [...]
Depending on what's involved, writing per-architecture support is not
necessarily beyond our means.
- FChE