[Bug 38266] New: Precision of math functions needs to be improved

https://bugs.winehq.org/show_bug.cgi?id=38266

Bug ID: 38266
           Summary: Precision of math functions needs to be improved
           Product: Wine
           Version: 1.7.39
          Hardware: x86
                OS: Linux
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: -unknown
          Assignee: [e-mail]
          Reporter: [e-mail]
      Distribution: ---

Created attachment 51094
  --> https://bugs.winehq.org/attachment.cgi?id=51094
Testcase showing the precision of exp() function

The math functions don't provide enough precision, which causes part of the
bugs 37149 and 37150.
Attached is a testcase, shows the exp() function holds about 13 precision
digits while there are about 16 in Windows/Linux.

[Bug 38266] Precision of math functions needs to be improved

https://bugs.winehq.org/show_bug.cgi?id=38266

Kevin Chan  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

[Bug 38266] Precision of math functions needs to be improved

https://bugs.winehq.org/show_bug.cgi?id=38266

Kevin Chan  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |testcase

[Bug 38266] Precision of math functions needs to be improved

https://bugs.winehq.org/show_bug.cgi?id=38266

Austin English  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |download, source

[Bug 38266] Precision of math functions needs to be improved

https://bugs.winehq.org/show_bug.cgi?id=38266

Austin English  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

[Bug 38266] Precision of math functions needs to be improved

https://bugs.winehq.org/show_bug.cgi?id=38266

[e-mail] changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail],
                   |                            |[e-mail],
                   |                            |[e-mail]

--- Comment #1 from [e-mail] ---
bug 37150 has been marked fixed.
bug 37149 has a staged patch...

Is there any reason this bug is still open?

[Bug 38266] Precision of math functions needs to be improved

https://bugs.winehq.org/show_bug.cgi?id=38266

--- Comment #2 from Bruno Jesus  ---
Not sure how to test this. I compiled the testcase in mingw and the results in
vanilla or patched wine (from bug 37149) is the same:
6741961161019537394602307493150652354869067776.0000000000

If I compile the testcase on Visual Studio 6 the result is (which is the same
as XP with this binary):
6741961161019537400000000000000000000000000000.0000000000

Native binary compiled with gcc (same result as mingw):
6741961161019537394602307493150652354869067776.0000000000

So I guess the the compiler is resolving the function on its own and not
linking.

[Bug 38266] Precision of math functions needs to be improved

https://bugs.winehq.org/show_bug.cgi?id=38266

Akarsha  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]