[Bug 16954] New: #defines of LONG_MAX, LONG_MIN and INT_MIN are wrong

http://bugs.winehq.org/show_bug.cgi?id=16954

Summary: #defines of LONG_MAX, LONG_MIN and INT_MIN are wrong
           Product: Wine
           Version: unspecified
          Platform: Other
        OS/Version: other
            Status: UNCONFIRMED
          Severity: enhancement
          Priority: P2
         Component: -unknown
        AssignedTo: [e-mail]
        ReportedBy: [e-mail]

a build of python2.5.2 under msys under wine goes completely tits-up
due to very strange long / int interactions.

the addition of a (long) typecast (around 0xNNNNL!) fixes the problem.

the definitions in the standard wine / mingw32 header files are
completely at odds with standard linux #defines for LONG_MAX and LONG_MIN.

the definition below for INT_MIN matches the standard linux #define,
and has the advantage that it actually works.

+#ifdef __WINE__  /* weird: you have to typecast 0x7fffffffL to long */
+#undef LONG_MAX
+#undef LONG_MIN
+#define LONG_MAX ((long)0x7FFFFFFFL)
+#define LONG_MIN ((long)(-LONG_MAX-1))
+#else

+#ifdef __WINE__
+/* wine's #define of INT_MIN slightly faulty...
+ */
+#undef INT_MIN
+#define INT_MIN  (-INT_MAX - 1)
+#endif
+
+

[Bug 16954] #defines of LONG_MAX, LONG_MIN and INT_MIN are wrong

http://bugs.winehq.org/show_bug.cgi?id=16954

Dan Kegel  changed:

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

--- Comment #1 from Dan Kegel   2009-01-19 11:08:01 ---
I can't quite see what's wrong with Wine's definitions.
(Why would (long) be needed around an NNNL constant?)

Can you provide a small C test program that fails to
compile (or compiles wrong)?  Thanks...

[Bug 16954] #defines of LONG_MAX, LONG_MIN and INT_MIN are wrong

http://bugs.winehq.org/show_bug.cgi?id=16954

--- Comment #2 from Dmitry Timoshkov   2009-01-20 11:15:15 ---
What exactly the problem is? If you are compiling with Wine's msvcrt
headers then you shouldn't do that, use libc provided by your platform.

[Bug 16954] #defines of LONG_MAX, LONG_MIN and INT_MIN are wrong

http://bugs.winehq.org/show_bug.cgi?id=16954

--- Comment #3 from Luke Kenneth Casson Leighton   2009-01-20 11:29:14 ---
hi tim, sorry there's quite a bit of context missing, from
other bugreports over the past few days.

i'm compiling python 2.5.2 using msys running under wine.

so as far as msys is concerned, there _is_ no libc: it's as if i was
compiling under native win32.

so ... er... actually, this bug should probably be under
mingw32 not wine! ha ha :)

[Bug 16954] #defines of LONG_MAX, LONG_MIN and INT_MIN are wrong

http://bugs.winehq.org/show_bug.cgi?id=16954

Dmitry Timoshkov  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |INVALID

--- Comment #4 from Dmitry Timoshkov   2009-01-20 11:33:35 ---
Exactly, Wine headers has nothing to do here.

[Bug 16954] #defines of LONG_MAX, LONG_MIN and INT_MIN are wrong

http://bugs.winehq.org/show_bug.cgi?id=16954

Dmitry Timoshkov  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |CLOSED

--- Comment #5 from Dmitry Timoshkov   2009-01-20 11:33:45 ---
Closing.

[Bug 16954] #defines of LONG_MAX, LONG_MIN and INT_MIN are wrong

http://bugs.winehq.org/show_bug.cgi?id=16954

--- Comment #6 from Dan Kegel   2009-01-20 12:13:33 ---
But they might, if somebody was building with winegcc.

A minimal test case showing the problem would be enough to reopen this bug.