#5282: Bizarre results from -P profiler on OS X
---------------------------------+------------------------------------------
Reporter: bos | Owner: simonmar
Type: bug | Status: new
Priority: high | Milestone: 7.4.1
Component: Runtime System | Version: 7.0.3
Keywords: | Testcase:
Blockedby: | Difficulty:
Os: Unknown/Multiple | Blocking:
Architecture: x86_64 (amd64) | Failure: Incorrect result at runtime
---------------------------------+------------------------------------------
Comment(by simonmar):
Here's a profile from today's HEAD with a patch I'm about to commit, on a
Mac:
{{{
Fri Nov 25 15:23 2011 Time and Allocation Profiling Report
(Final)
5282 +RTS -P -RTS idle 10000000
total time = 1.00 secs (1001 ticks @ 1000 us)
total alloc = 800,082,392 bytes (excludes profiling overheads)
COST CENTRE MODULE %time %alloc ticks bytes
counting.loop Main 98.7 100.0 988 800000000
idle.\ Main 1.3 0.0 13 0
individual
inherited
COST CENTRE MODULE no. entries %time %alloc
%time %alloc ticks bytes
MAIN MAIN 108 0 0.0 0.0
100.0 100.0 0 656
main Main 217 0 0.0 0.0
100.0 100.0 0 28848
main.count Main 223 1 0.0 0.0
0.0 0.0 0 10880
idle Main 220 1 0.0 0.0
100.0 100.0 0 0
counting Main 221 1 0.0 0.0
100.0 100.0 0 72
counting.loop Main 222 10000001 98.7 100.0
100.0 100.0 988 800000000
idle.\ Main 224 10000000 1.3 0.0
1.3 0.0 13 0
main.bm Main 218 1 0.0 0.0
0.0 0.0 0 0
CAF Main 215 0 0.0 0.0
0.0 0.0 0 48
main Main 216 1 0.0 0.0
0.0 0.0 0 0
main.bm Main 219 0 0.0 0.0
0.0 0.0 0 304
CAF GHC.Conc.Signal 206 0 0.0 0.0
0.0 0.0 0 672
CAF GHC.IO.Encoding 190 0 0.0 0.0
0.0 0.0 0 3360
CAF GHC.IO.Encoding.Iconv 188 0 0.0 0.0
0.0 0.0 0 248
CAF GHC.IO.Handle.FD 179 0 0.0 0.0
0.0 0.0 0 34944
CAF Text.Read.Lex 133 0 0.0 0.0
0.0 0.0 0 2360
}}}
I've switched to using wall-clock time for profile samples on both the
non-threaded and the threaded RTS (CPU time was far too unreliable), and
I've bumped the default sample interval to 1ms. The profile on Linux
looks the same, and on Windows I've switched to using a better timer so
the results there are better too.
--
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/5282#comment:19>
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
