Re: [PATCH v2 00/12] simpletrace: refactor and general improvements

[Stefan Hajnoczi]

On Thu, May 04, 2023 at 01:53:38PM -0400, John Snow wrote:
> On Thu, May 4, 2023, 1:48 PM Stefan Hajnoczi <stefanha@redhat.com> wrote:
> 
> > On Tue, May 02, 2023 at 11:23:27AM +0200, Mads Ynddal wrote:
> > > From: Mads Ynddal <m.ynddal@samsung.com>
> > >
> > > I wanted to use simpletrace.py for an internal project, so I tried to
> > update
> > > and polish the code. Some of the commits resolve specific issues, while
> > some
> > > are more subjective.
> >
> > An internal project based on qemu.git or a completely separate codebase?
> >
> > Sometimes I've wondered whether tracetool should be extracted from
> > qemu.git and moved into its own QEMU-independent place. That way other
> > C/C++ applications and libraries could use it easily.
> >
> > Now that Alex Bennee removed the vcpu trace events that were specific to
> > QEMU, the tracing code is less tightly coupled to QEMU. There are
> > probably still a number of places that need to be cleaned up in order
> > for the tracing code to be independent of QEMU though.
> >
> > If there is interest in doing this, I support the effort, although I'm
> > not sure how much time I have to actually do the work myself.
> >
> 
> I meant internal, but if there's interest in fully extracting it, I have a
> playbook for that now based on my efforts to do the same for qemu.qmp and I
> can offer good stewardship for that process.

I was curious how Mads is using simpletrace for an internal (to
Samsung?) project.

Maybe only simpletrace.py is needed because Mads' own code emits trace
logs in the simpletrace binary format, but in the general case we could
extract most of QEMU's tracing infrastructure (including tracetool).

Stefan

signature.asc
Description: PGP signature