[ ghc-Bugs-710864 ] "entries" field is often 0

[SourceForge.net]

Bugs item #710864, was opened at 2003-03-27 17:48
Message generated for change (Comment added) made by simonmar
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=108032&aid=710864&group_id=8032

Category: Profiling
Group: 5.04.2
>Status: Closed
>Resolution: Fixed
Priority: 5
Submitted By: Nobody/Anonymous (nobody)
Assigned to: Nobody/Anonymous (nobody)
Summary: "entries" field is often 0

Initial Comment:
On a time profiling report the "entries" field is often
zero.
If I want to know why "getAppropriatePreds" takes up so
much time (14.7%) in my code, does the zero mean an
overflowed count or that the cost centre has never been
entered? I'm using 5.04.2 on Linux, and my program is
compiled with -O -prof -auto-all.
 
For example:

   compose         InfrequentCheck   1137    12280580 
 2.1    2.5   2.1    2.5
  align            InfrequentCheck   1127           0 
 3.7     6.2     8.3      6.7
   align2          InfrequentCheck   1139     8577943 
 0.4    0.0  0.4    0.0
   lookupModeArgs  InfrequentCheck   1136    86747940 
 4.2    0.4  4.2    0.4
   getAppropriatePreds InfrequentCheck  1126        0 
14.7    6.6  14.7    6.6

Amanda Clare (afc at aber dot ac dot uk)


----------------------------------------------------------------------

>Comment By: Simon Marlow (simonmar)
Date: 2004-08-09 13:44

Message:
Logged In: YES 
user_id=48280

I've fixed this one in the new code generator (well,
mostly).  6.4 will behave better here.


----------------------------------------------------------------------

Comment By: Simon Peyton Jones (simonpj)
Date: 2003-12-16 15:29

Message:
Logged In: YES 
user_id=50165

The code I noticed the problem with would take me days to 
distill to a 
small enough example. But I do have a small example program 
(about 150 
lines) that I think also wrongly shows zeros in the "entries" 
column, so 
I've tarred this up and put it at 
http://users.aber.ac.uk/afc/zerosexample.tar.bz2

There's a readme file that shows how the program should be 
run. The 
program reads in a tree of biological data and prunes the tree 
to just 
keep the branches that are represented in the input file.

The zero entry I don't understand in the profile in this case is 
for 
"maketree", which is the function argument to a fold and so 
should be 
called many times (its internals are certainly called many 
times). I'm 
also not sure why "trim" and "parseline" have zeros.

Amanda


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=108032&aid=710864&group_id=8032