Re: [PATCH] Win32: Ruby & APR; build problems for Ruby Subversion SWIG bindings

Hi,

At Sat, 14 May 2005 18:06:32 +0900,
Erik Huelsmann wrote in [ruby-core:04977]:
> Index: configure.in
> ===================================================================
> RCS file: /src/ruby/configure.in,v
> retrieving revision 1.275
> diff -u -r1.275 configure.in
> --- configure.in	30 Apr 2005 02:59:41 -0000	1.275
> +++ configure.in	14 May 2005 07:45:02 -0000
> @@ -226,6 +226,8 @@
>  AC_CHECK_SIZEOF(double, 8)
>  AC_CHECK_SIZEOF(time_t, 0)
>  
> +AC_CHECK_TYPES([pid_t, gid_t, uid_t])

I think AC_TYPE_UID_T and AC_TYPE_PID_T also would be needed to
remove.

> Index: process.c
> ===================================================================
> RCS file: /src/ruby/process.c,v
> retrieving revision 1.129
> diff -u -r1.129 process.c
> --- process.c	4 Mar 2005 06:47:41 -0000	1.129
> +++ process.c	14 May 2005 07:45:03 -0000
> @@ -1847,14 +1847,14 @@
>      if (pid < 0) rb_sys_fail(0);
>      return INT2FIX(pid);
>  #elif defined(HAVE_SETPGRP) && defined(TIOCNOTTY)
> -  pid_t pid;
> +  rb_pid_t pid;
>    int ret;
>  
>    rb_secure(2);
>    pid = getpid();
>  #if defined(SETPGRP_VOID)
>    ret = setpgrp();
> -  /* If `pid_t setpgrp(void)' is equivalent to setsid(),
> +  /* If `rb_pid_t setpgrp(void)' is equivalent to setsid(),

This is overdone, isn't it?

> Index: ruby.h
> ===================================================================
> RCS file: /src/ruby/ruby.h,v
> retrieving revision 1.111
> diff -u -r1.111 ruby.h
> --- ruby.h	30 Apr 2005 02:59:41 -0000	1.111
> +++ ruby.h	14 May 2005 07:45:03 -0000
> @@ -82,6 +82,25 @@
>  typedef unsigned long VALUE;
>  typedef unsigned long ID;
>  
> +#if ! HAVE_PID_T
> +  typedef int rb_pid_t;
> +#else
> +  typedef pid_t rb_pid_t;
> +#endif

Why not #ifdef?

-- 
Nobu Nakada