Author Topic: x86 Machine Code Statistics  (Read 3729 times)

Vortex

  • Member
  • *****
  • Posts: 1725
x86 Machine Code Statistics
« on: May 26, 2012, 04:15:11 AM »
An interesting article by Peter Kankowski :

x86 Machine Code Statistics

jj2007

  • Member
  • *****
  • Posts: 7740
  • Assembler is fun ;-)
    • MasmBasic
Re: x86 Machine Code Statistics
« Reply #1 on: May 26, 2012, 07:07:41 AM »
6% call, 1% ret??  ::)

KeepingRealBusy

  • Member
  • ***
  • Posts: 426
Re: x86 Machine Code Statistics
« Reply #2 on: May 26, 2012, 08:21:57 AM »
JJ,

Yes, I believe that. You have many calls to a procedure, but that procedure only has 1 ret (maybe more than 1 with conditional early exits). Note the stats are for the code itself, not a profile of the execution of the code.

Dave.

jj2007

  • Member
  • *****
  • Posts: 7740
  • Assembler is fun ;-)
    • MasmBasic
Re: x86 Machine Code Statistics
« Reply #3 on: May 26, 2012, 04:23:17 PM »
Dave,
I overlooked the obvious. Thanks for putting me back on track :biggrin:

dedndave

  • Member
  • *****
  • Posts: 8749
  • Still using Abacus 2.0
    • DednDave
Re: x86 Machine Code Statistics
« Reply #4 on: May 26, 2012, 10:39:20 PM »
API functions are probably used ~5:1 over local PROC's

K_F

  • Member
  • *****
  • Posts: 1292
  • Anybody out there?
Re: x86 Machine Code Statistics
« Reply #5 on: May 27, 2012, 06:29:15 AM »
The way software is done now days.. there are many proc calls with just as many crashes, without rets...
Not surprising  :P
'Sire, Sire!... the peasants are Revolting !!!'
'Yes, they are.. aren't they....'