rubinius

Comments and changes to this ticket

• 

  zimbatm February 8th, 2008 @ 04:40 AM

  • → State changed from “new” to “open”

  asinh works well under OS X :

  #include <stdio.h>
  #include <math.h>
  
  int main() {
    printf("%f\n", asinh(1));
  }
  

  # gcc test.c
  # ./a.out
  0.881374
  

• 

  zimbatm February 8th, 2008 @ 04:22 AM

  I also checked the ffi signature in kernel/platform/math.rb. I don't see any problem here.

• 

  zimbatm February 8th, 2008 @ 09:31 AM

  For the record :

  <dbussink> zimbatm: i see you've been looking at Math#asinh
  <dbussink> zimbatm: i've wasted some time on that too, it's a really weird problem
  <dbussink> zimbatm: there are some problems though with ffi which evan is investigating, it could be cause of this problem too
  

  indeed, all other math operations work and are defined the same way. I wish I knew how to delve into the ffi code...

• 

  Federico Builes February 10th, 2008 @ 06:03 PM

  I couldn't reproduce the behavior on OS X 10.5.1 using HEAD:

  irb(main):005:0> Math.asinh(1)

  => 0.881373587019543

• 

  zimbatm February 11th, 2008 @ 07:46 AM

  Removed excludes in 55a52181... Evan fixed the FFI over the weekend

• 

  zimbatm February 11th, 2008 @ 07:46 AM

  • → State changed from “open” to “resolved”