Re: trace-cmd and host-guest tracing

Re: trace-cmd and host-guest tracing

[Steven Rostedt]

On Thu, 4 Mar 2021 17:52:58 +0200
Tzvetomir Stoyanov <tz.stoyanov@gmail.com> wrote:

> Hi Dario,
> We decided to simplify opening trace files with host and guest trace data
> and changed a little bit the logic and the APIs. Now the timestamps in the
> guest file are recalculated by default, when the file is opened. Even if
> the file is opened standalone. This makes the logic, described in Bugzilla
> issue obsolete:
>     https://bugzilla.kernel.org/show_bug.cgi?id=211657

Tzvetomir,

Can you add this information into that bugzilla too?

Thanks!

-- Steve

> and the issue is not relevant any more. Moreover, the APIs described there
> are removed, as not needed any more.
> I hope that no one started working on that yet.
> 

Re: trace-cmd and host-guest tracing

[Dario Faggioli]

[-- Attachment #1: Type: text/plain, Size: 1789 bytes --]

On Thu, 2021-03-04 at 17:52 +0200, Tzvetomir Stoyanov wrote:
> Hi Dario,
>
Hi!

> We decided to simplify opening trace files with host and guest trace
> data and changed a little bit the logic and the APIs. Now the
> timestamps in the guest file are recalculated by default, when the
> file is opened. 
>
Yep, I saw the patch series earlier today... Seems nice. :-)

> Even if the file is opened standalone. This makes the logic,
> described in Bugzilla issue obsolete:
>     https://bugzilla.kernel.org/show_bug.cgi?id=211657
> and the issue is not relevant any more. Moreover, the APIs described
> there are removed, as not needed any more.
>
Right. As said, I noticed this already, when seeing the patches fly by.
But thanks a lot for the heads-up!

> I hope that no one started working on that yet. 
> 
Well, as a matter of fact Stefano (Cc-ed) was just almost ready
releasing his first RFC, based right on those APIs, which is kind of
funny! :-D :-P

Of course, this is perfectly fine... We knew that the API may change
and things like this one happens all the time in Open Source
development, so better get use to deal with them.

If you don't mind, though, we'd still like to submit the patches he was
working on, as RFC. Surely, the part of the code using the removed API
would have to be ignored... But there's a couple of other things in
those patches too, which we'd like, if possible, to have your feedback
about. Let us know if this could be a problem...

Regards
-- 
Dario Faggioli, Ph.D
http://about.me/dario.faggioli
Virtualization Software Engineer
SUSE Labs, SUSE https://www.suse.com/
-------------------------------------------------------------------
<<This happens because _I_ choose it to happen!>> (Raistlin Majere)

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: trace-cmd and host-guest tracing

[Steven Rostedt]

On Thu, 04 Mar 2021 23:06:26 +0100
Dario Faggioli <dfaggioli@suse.com> wrote:

> If you don't mind, though, we'd still like to submit the patches he was
> working on, as RFC. Surely, the part of the code using the removed API
> would have to be ignored... But there's a couple of other things in
> those patches too, which we'd like, if possible, to have your feedback
> about. Let us know if this could be a problem...

As you said, "happens all the time in Open Source development", which is
true here as well.

No problem sending RFCs based on the old APIs. It may give us a better idea
of what you expect of this code as well.

Thanks,

-- Steve