Re: [Qemu-devel] [RFC] tracing: consistent usage of "disable" in "trace-events"

[Lluís]

Daniel P Berrange writes:

> On Mon, Nov 08, 2010 at 04:55:10PM +0100, Lluís wrote:
>> In any case, there might appear other events that could have performance
>> implications, although I understand the ease of usage of having all
>> trace events available by default.

> Ok, I agree that if we have tracepoints in the kind of places in TCG
> you describe, then this could have measurable performance impact.

>> That's why I would rather declare all trace-events without the "disable"
>> keyword, and leave it only on those events that are known to have a high
>> frequency, as no backend should have so poor performance as to force
>> events to "disappear".

> This sounds like a reasonable plan to me. 


Then I'll cook a simple patch for that later.

Lluis

-- 
 "And it's much the same thing with knowledge, for whenever you learn
 something new, the whole world becomes that much richer."
 -- The Princess of Pure Reason, as told by Norton Juster in The Phantom
 Tollbooth