search

cuitao2046 / gperftools

Automatically exported from code.google.com/p/gperftools
BSD 3-Clause "New" or "Revised" License
0 stars 0 forks source link

<cycle 1> is abusive #253

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago 
What steps will reproduce the problem?
1. Wrote some for statement in your source code.
2. Compiler and link with profiler.
3. CPU profiling and then watch its output.

What is the expected output? What do you see instead?
<cycle 1>, <cycle 2>, ... <cycle n> are expected.

What version of the product are you using? On what operating system?
Linux x86_64

Please provide any additional information below.
NONE.

Original issue reported on code.google.com by wang.wen...@gmail.com on 18 Jun 2010 at 9:15

GoogleCodeExporter commented 9 years ago 
I'm sorry, I don't understand what problem you're seeing.  Perhaps you can 
attach a sample program that illustrates the problem, along with the profile 
that's generated.  What does the profile say when you run pprof --text?  What 
do you expect it to say?

Original comment by csilv...@gmail.com on 18 Jun 2010 at 6:04

GoogleCodeExporter commented 9 years ago 
Any more word on this?

Original comment by csilv...@gmail.com on 2 Aug 2010 at 4:47

GoogleCodeExporter commented 9 years ago 
Perhaps the problem is that in Kcachegrind when there are cycles it's hard to 
tell what's going on.  With pprof currently removing template and argument 
overload information there are some false cycles making it worse.  First step 
is to use kcachegrind's view -> skip cycle detection option.  If you have the 
patch in my issue 272, to keep overload information, you'll also get fewer 
cycles.

(I think this is more a Kcachegrind issue than a pprof issue.)

Original comment by CHKings...@gmail.com on 5 Oct 2010 at 5:55

GoogleCodeExporter commented 9 years ago 
Thanks for this analysis, CHKingsley.  Given what you said, do you think it's 
reasonable to close this bug WillNotFix?  It doesn't sound like there's much 
fixing I can do in any case (besides applying the patch in issue 272).

Original comment by csilv...@gmail.com on 8 Oct 2010 at 1:32

GoogleCodeExporter commented 9 years ago 
I'm a little sad, since this is the best issue-report subject line of any 
perftools issue I've ever gotten :-), but am closing this as per my idea in 
comment 4.

Original comment by csilv...@gmail.com on 31 Aug 2011 at 11:46