[Pdbg] [PATCH 7/7] pdbg/htm: Enforce powersave=off

[Alistair Popple]

On Friday, 16 March 2018 2:10:14 PM AEDT Cyril Bur wrote:
> On Fri, 2018-03-16 at 13:42 +1100, Alistair Popple wrote:
> > > +		i = 0;
> > > +		pdbg_for_each_class_target("core", target)
> > > +			if (!target_is_disabled(target))
> > > +				i++;
> > > +		if (i == 0)
> > > +			fprintf(stderr, "You haven't selected any HTM Cores\n");
> > > +		if (i > 1) {
> > > +			fprintf(stderr, "It doesn't make sense to core trace on"
> > > +				" multiple cores at once. %d\n", i);
> > 
> > For my own background why not? Does the HW not support it? Or do we just not
> > support it here because it's "hard"?
> 
> Depends on your definition of hard. I don't believe there is a hardware
> limitation (which makes me think that the printf is misleading, will
> fix) - although I've never actually tried.
> 
> The reason I don't want to allow it for now is that doing so would
> require partitioning up the memtrace buffer provided by the kernel and
> remembering how it has been split up across multiple calls. I've not
> seen anyone require such functionality and I don't think any of our
> current tooling can do it either.

Sounds somewhat annoying, I'm fine with not doing that although it would be good
to reword the printf to just say we don't currently support it (and to email
Cyril if you would like it to be supported :-P)

- Alistair

> > 
> > > +			fprintf(stderr, "What you probably want is -p 0 -c x\n");
> > > +		}
> > > +		if (i != 1)
> > > +			return 0;
> > > +	}
> > > +
> > >  	optind++;
> > >  	for (i = 0; i < ARRAY_SIZE(actions); i++) {
> > >  		if (strcmp(argv[optind], actions[i].name) == 0) {
> > > 
> > 
> > 
>