RE: [PATCH 2/2] btrace: set/show record btrace cpu

["Metzger, Markus T" <markus dot t dot metzger at intel dot com>]

Hello Eli,

> > > I think we need to say something about just what those "errata
> workarounds"
> > > are, and what are they used for.
> >
> > I rephrased this to "... for enabling workarounds for processor errata
> > when decoding the trace".
> 
> It's better, but still not clear enough.  What kind of "errata" are we talking about?
> The kind described in https://community.amd.com/thread/186609, for example?
> And what do the workarounds do?
> 
> If you can explain that to me or give an example, I will try to propose some text to
> describe that in the manual.

Processor errata are bugs that, in our case, may cause the trace to not match the spec.
This typically causes unaware decoders to fail with some error.

An erratum workaround will try to detect an erroneous trace packet sequence and
correct it.

In our case, each workaround needs to be enabled separately.  The decoder determines
the workarounds to be enabled based on the processor on which the trace was recorded.


> > > > +  add_prefix_cmd ("cpu", class_support, cmd_set_record_btrace_cpu,
> > > > +		  _("\
> > > > +Set the cpu to be used for trace decode.\n\n\ The format is
> > > > +\"<vendor>: <identifier>\" or \"none\" or \"auto\" (default).
> > >                            ^^
> > > So should there be a blank after the colon, or shouldn't there be?
> > > The example in the manual says no blank.
> >
> > White space is ignored.  Do we write this explicitly?
> 
> Not necessarily.  But I'd prefer us to consistently use one of the forms.

I removed the optional space here and in the commit-message.

Regards,
Markus.
Intel Deutschland GmbH
Registered Address: Am Campeon 10-12, 85579 Neubiberg, Germany
Tel: +49 89 99 8853-0, www.intel.de
Managing Directors: Christin Eisenschmid, Christian Lamprechter
Chairperson of the Supervisory Board: Nicole Lau
Registered Office: Munich
Commercial Register: Amtsgericht Muenchen HRB 186928