• Open Menu Close Menu
  • Apple
  • Shopping Bag
  • Apple
  • Mac
  • iPad
  • iPhone
  • Watch
  • TV
  • Music
  • Support
  • Search apple.com
  • Shopping Bag

Lists

Open Menu Close Menu
  • Terms and Conditions
  • Lists hosted on this site
  • Email the Postmaster
  • Tips for posting to public mailing lists
Re: Shark question
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Shark question


  • Subject: Re: Shark question
  • From: kwiley <email@hidden>
  • Date: Mon, 16 Jun 2008 13:08:44 -0700


On Jun 16, 2008, at 12:33 PM, Wade Tregaskis wrote:

Self is the number of samples recorded where that function (or line of source, or instruction) was executing. When looking at the profile view (function breakdown), self should add up to 100%. An easy way to check is to select all the rows in the outline view. If you have a session where this doesn't happen, please send it our way.

I figured it out. Heavy view does not have this problem. Tree view has this problem when some functions are not expanded. If the stack trace is fully expanded (by option clicking the triangle for example), then the Self column sums to 100%. Otherwise, stack frames contained within closed triangles don't register in the Self column and consequently Self doesn't sum to 100%.


I suppose this behavior makes sense, I certainly can't think of an alternative that makes better sense.

The behaviour is correct; no problems here. Self will add to 100% across the entire calltree because it's straightforwardly the percentage of time by function, as noted. Total will sum to 100% for the roots of the trees (whether in heavy or tree view), but because it's an aggregate statistic that includes all the callers or callees (heavy and tree views, respectively), it "repeats", if you like to think of it that way.


This shouldn't be confusing; that it is to you makes me suspect you're simply wanting to see the stats calculated in a different way, for whatever purpose you currently have... if you can figure out what that is, or at least explain what you're trying to do, we might be able to either point you in a better direction, or take it onboard as an enhancement request.

Well, I already conceded on the current design, I'm not disagreeing. I think the thing that threw me was that under most circumstances the Self column in Tree view doesn't sum to 100%. Just took me a while to wrap my head around what's going on there. It's almost as if what I want to see is a Self column that reflects Self in the current way if the subtree is expanded, but which includes the underlying subtree if the subtree is closed and therefore removed from view. Perhaps it could be explicitly offset with a different color. This is, of course, the same thing as the Total column just to its right, except that it has the very pleasing property that the Self column alway sums to 100%. In the current design, neither column sums to 100%, you don't really feel like you are seeing the full distribution of time- spent...unless you fully expand the tree and then it is probably three to five screens tall and can't be seen all at once.


If no one agrees with me, then I'm happy to fold, I don't mean to be difficult about it...I'm just suggesting the *idea* of a column that always sums to 100% in Tree view. This is similar to the Total column in Heavy view, but of course the organization is completely inside- out. It is fundamentally different aside from the fact that the column sums to 100%.

What do other people think?  Any other responses to this?

________________________________________________________________________
Keith Wiley    email@hidden   http://www.cs.unm.edu/~kwiley

"Luminous beings are we, not this crude matter."
                                           --  Yoda
________________________________________________________________________



_______________________________________________
Do not post admin requests to the list. They will be ignored.
Xcode-users mailing list      (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden


  • Follow-Ups:
    • Re: Shark question
      • From: Rick Altherr <email@hidden>
References: 
 >Shark question (From: kwiley <email@hidden>)
 >Re: Shark question (From: Rick Altherr <email@hidden>)
 >Re: Shark question (From: kwiley <email@hidden>)
 >Re: Shark question (From: Wade Tregaskis <email@hidden>)

  • Prev by Date: Re: aggregate targets and build configurations
  • Next by Date: This is a test message
  • Previous by thread: Re: Shark question
  • Next by thread: Re: Shark question
  • Index(es):
    • Date
    • Thread