#5505: Program runs faster with profiling than without
---------------------------------+------------------------------------------
    Reporter:  simonpj           |        Owner:  simonmar    
        Type:  bug               |       Status:  new         
    Priority:  high              |    Milestone:  7.4.1       
   Component:  Compiler          |      Version:  7.0.3       
    Keywords:                    |     Testcase:              
   Blockedby:                    |   Difficulty:              
          Os:  Unknown/Multiple  |     Blocking:              
Architecture:  Unknown/Multiple  |      Failure:  None/Unknown
---------------------------------+------------------------------------------

Comment(by bgamari):

 Here are -sstderr logs for both the profiling and non-profiling cases.
 Note that in my case relatively little time is spent in GC,

 '''With profiling'''
 {{{
 ../../GGen/dist/build/GGen/GGen z_endstop_clamp_and_mount.stl +RTS
 -sstderr
   10,355,015,584 bytes allocated in the heap
      196,136,736 bytes copied during GC
        7,760,328 bytes maximum residency (27 sample(s))
          479,656 bytes maximum slop
               20 MB total memory in use (0 MB lost due to fragmentation)

   Generation 0: 19768 collections,     0 parallel,  0.44s,  0.47s elapsed
   Generation 1:    27 collections,     0 parallel,  0.15s,  0.16s elapsed

   INIT  time    0.01s  (  0.00s elapsed)
   MUT   time   13.56s  ( 15.97s elapsed)
   GC    time    0.59s  (  0.64s elapsed)
   RP    time    0.00s  (  0.00s elapsed)
   PROF  time    0.00s  (  0.00s elapsed)
   EXIT  time    0.00s  (  0.00s elapsed)
   Total time   14.15s  ( 16.60s elapsed)

   %GC time       4.2%  (3.8% elapsed)

   Alloc rate    763,501,989 bytes per MUT second

   Productivity  95.8% of total user, 81.6% of total elapsed
 }}}

 '''Without profiling'''
 {{{
 ../../GGen/dist/build/GGen/GGen z_endstop_clamp_and_mount.stl +RTS
 -sstderr
   18,234,515,824 bytes allocated in the heap
      128,446,200 bytes copied during GC
        5,211,504 bytes maximum residency (23 sample(s))
          325,792 bytes maximum slop
               13 MB total memory in use (0 MB lost due to fragmentation)

   Generation 0: 34960 collections,     0 parallel,  0.62s,  0.71s elapsed
   Generation 1:    23 collections,     0 parallel,  0.10s,  0.10s elapsed

   INIT  time    0.01s  (  0.00s elapsed)
   MUT   time   27.14s  ( 31.47s elapsed)
   GC    time    0.72s  (  0.81s elapsed)
   EXIT  time    0.00s  (  0.00s elapsed)
   Total time   27.86s  ( 32.27s elapsed)

   %GC time       2.6%  (2.5% elapsed)

   Alloc rate    671,650,697 bytes per MUT second

   Productivity  97.4% of total user, 84.1% of total elapsed
 }}}

-- 
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/5505#comment:6>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler

_______________________________________________
Glasgow-haskell-bugs mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs

Reply via email to