[Patch 01/02] mieq threading prep: Only increment tail (push) when the event data is actually in the queue

[Simon Thum]

Jeremy Huddleston wrote:
> 
> On Nov 18, 2008, at 11:54, Simon Thum wrote:
> 
>> Jeremy Huddleston wrote:
>>> -    unsigned int           oldtail = miEventQueue.tail, newtail;
>>> +    unsigned int           oldtail = miEventQueue.tail;
>> All fine, but is there a specific reason to remove newtail? Not that I'd
>> expect this to save more than one or two cycles on any platform...
> 
> I just thought it was adding to namespace polution, TBH... no "real"
> reason... and with any level of optimization, you'll get back those two
> cycles.  I can revert that "part" if you really care.
I merely was curious about subtle effects I might have missed, nothing else.

Cheers,

Simon