Re: [RFC] systemtap: begin the process of using proper kernel APIs (part1: use kprobe symbol_name/offset instead of address)

["Frank Ch. Eigler" <fche@redhat.com>]

Hi -

On Fri, Jul 18, 2008 at 03:10:27PM +0200, Peter Zijlstra wrote:
> [...]
> > For example, we're about to do unwinding/stack-traces of userspace
> > programs.  To what extent do you think the kernel unwinder (should one
> > reappear in git) would welcome patches that provide zero benefit to
> > the kernel, but only enable a peculiar (nonintrusive) sort of
> > unwinding we would need for complex userspace stacks?
> 
> I think sysprof (kernel/trace/trace_sysprof.c) already does user-space
> stack unwinding. So pushing that capability further up the chain when a
> second user (stap) comes along makes perfect sense.

trace_sysprof relies on dump_stack, which is x86-only and does not do
elf/dwarf unwinding proper.  (Likewise oprofile, etc.)  They can't
even start, because they don't have unwind data available - something
we plan to make available to the systemtap runtime.


- FChE