On Thu, May 28, 2009 at 12:27 PM, Philip Mucci <mu...@eecs.utk.edu> wrote:
> Hi Stefane,
>
> Everything looks great. The only issue I have is that the new
> event::qual:qual:qual syntax will break a lot of tools out there that do
> strict checking of assumed libpfm event formats....so I'm not convinced the
> double colon is necessary. It can adequately be parsed by just assuming that
> the first item is always the event specifier. My 2 cents.

The issue is how you treat umask vs. attributes. You need a way to
separate those out.
This is the role of '::'. An event may not have umask, but attributes.
The other way to
do this is to detect attributes by identifying the = sign. Umask have
no = sign, attributes
always have one:

     INST_RETIRED:any_p:u=1:k=1 (u=user priv level, k=kernel priv level)

With this syntax, you could interleave umask and attributes:

     INST_RETIRED:u=1:any_p:k=1

Of course, that would be more work for the parsing code.

Attributes vs. umask:
   - attributes have user-controllable values, umasks don't

------------------------------------------------------------------------------
Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT 
is a gathering of tech-side developers & brand creativity professionals. Meet
the minds behind Google Creative Lab, Visual Complexity, Processing, & 
iPhoneDevCamp as they present alongside digital heavyweights like Barbarian 
Group, R/GA, & Big Spaceship. http://p.sf.net/sfu/creativitycat-com 
_______________________________________________
perfmon2-devel mailing list
perfmon2-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/perfmon2-devel

Reply via email to