Re: [PATCH v11 00/14] unwind_user: x86: Deferred unwinding infrastructure

[Namhyung Kim <namhyung@kernel.org>]

Hello,

On Mon, Jun 30, 2025 at 02:50:52PM +0200, Florian Weimer wrote:
> * Steven Rostedt:
> 
> > SFrames is now supported in gcc binutils and soon will also be supported
> > by LLVM.
> 
> Is the LLVM support discussed here?
> 
>   [RFC] Adding SFrame support to llvm
>   <https://discourse.llvm.org/t/rfc-adding-sframe-support-to-llvm/86900>
> 
> Or is there a secone effort?
> 
> > I have more patches on top of this series that add perf support, ftrace
> > support, sframe support and the x86 fix ups (for VDSO). But each of those
> > patch series can be worked on independently, but they all depend on this
> > series (although the x86 specific patches at the end isn't necessarily
> > needed, at least for other architectures).
> 
> Related to perf support: I'm writing up the SFrame change proposal for
> Fedora, and I want to include testing instructions.  Any idea yet what a
> typical “perf top” or “perf report” command line would look like?

I think you can run "perf report -s dso,sym -g none" then it will show
"Children" and "Self" overheads.  If callchain in userspace works ok,
you will get non-kernel entries (symbols start with "[.]") having more
children overhead than the self.

  $ perf record -g -- perf bench sched messaging
  
  $ perf report -s dso,sym -g none | grep -F -e Children -e '[.]' | head
  # Children      Self  Shared Object           Symbol
      63.09%     0.01%  perf                    [.] run_bench
      63.09%     0.00%  libc.so.6               [.] __libc_start_call_main
      63.09%     0.00%  perf                    [.] cmd_bench
      63.09%     0.00%  perf                    [.] handle_internal_command
      63.09%     0.00%  perf                    [.] main
      63.09%     0.00%  perf                    [.] run_argv
      63.09%     0.00%  perf                    [.] run_builtin
      63.02%     0.00%  perf                    [.] bench_sched_messaging
      62.79%     0.00%  perf                    [.] group

Thanks,
Namhyung