FreeTDS Development Group

[Brian Bruns <camber AT ais.org>]

Thanks Steve,

I've merged it into my local copy and made the other changes discussed,
however the cvs repository seems to be down...so I'll merge it in when I
can.  Also, I added your name in the AUTHORS file.

Cheers,

Brian

On Fri, 26 Jan 2001, Steve Langasek wrote:

> Brian,
> 
> On Fri, 26 Jan 2001, Brian Bruns wrote:
> 
> > The issue with alpha is mostly a set of typedefs that need to have
> > autoconf determine which types to use for 8, 16, and 32 bit ints and 4 and
> > 8 byte floats...It's on my list for the next release but if someone wants
> > to take it on, I wouldn't complain.  I'm basically just going to copy 
> > what glib/gtk does
> 
> Ask, and you shall receive.  Below is a patch which will check sizes and set
> TDS_SMALLINT, TDS_INT, TDS_REAL, and TDS_FLOAT appropriately.  I've assumed
> here that 'char' will always be our 8-byte int, because I can't think of a
> case where that isn't true, and so I don't know what those platforms use in
> place of 'char'.  The checks for 'real' and 'float' don't do a whole lot
> either, but the good news is they don't have to for Linux-Alpha.  (Actually,
> the only one that has to change for Alpha is TDS_INT, but I might as well 
> try
> to design for the future. :)
> 
> Regards,
> Steve Langasek
> postmodern programmer
> 
> Index: configure.in
> ===================================================================
> RCS file: /Repository/freetds/configure.in,v
> retrieving revision 1.22
> diff -u -w -r1.22 configure.in
> --- configure.in      2000/11/04 20:41:43     1.22
> +++ configure.in      2001/01/26 06:08:16
> @@ -31,6 +31,43 @@
>  AC_HEADER_STDC
>  
>  dnl ------------------------------------------------------------
> +dnl Checks for integers/floats of different sizes
> +dnl ------------------------------------------------------------
> +AC_CHECK_SIZEOF(char)
> +AC_CHECK_SIZEOF(short)
> +AC_CHECK_SIZEOF(long)
> +AC_CHECK_SIZEOF(int)
> +AC_CHECK_SIZEOF(float)
> +AC_CHECK_SIZEOF(double)
> +
> +case 2 in
> +$ac_cv_sizeof_short) smallint=short;;
> +$ac_cv_sizeof_int)   smallint=int;;
> +esac
> +case 4 in
> +$ac_cv_sizeof_short) int=short;;
> +$ac_cv_sizeof_int)   int=int;;
> +$ac_cv_sizeof_long)  int=long;;
> +esac
> +
> +case 4 in
> +$ac_cv_sizeof_float) real=float;;
> +$ac_cv_sizeof_double)        real=double;;
> +*)                   AC_ERROR(No 32-bit float found.)
> +esac
> +
> +case 8 in
> +$ac_cv_sizeof_float) float=float;;
> +$ac_cv_sizeof_double)        float=double;;
> +*)                   AC_ERROR(No 64-bit float found.)
> +esac
> +
> +AC_SUBST(smallint)
> +AC_SUBST(int)
> +AC_SUBST(real)
> 
> +AC_SUBST(float)
> +
> +dnl ------------------------------------------------------------
>  dnl Checks for typedefs and structures
>  dnl ------------------------------------------------------------
>  
> @@ -99,7 +136,7 @@
>  dnl ------------------------------------------------------------
>  dnl Final output
>  dnl ------------------------------------------------------------
> -AC_OUTPUT(include/tdsver.h \
> +AC_OUTPUT(include/tdsver.h include/tds.h \
>       Makefile \
>       include/Makefile \
>       src/Makefile \
> Index: include/tds.h
> ===================================================================
> RCS file: /Repository/freetds/include/tds.h,v
> retrieving revision 1.66
> diff -u -w -r1.66 include/tds.h
> --- include/tds.h     2001/01/21 20:47:11     1.66
> +++ include/tds.h     2001/01/26 06:08:17
> @@ -82,12 +82,12 @@
>  typedef char         TDS_CHAR;      /*  8 bit char     */
>  typedef unsigned char        TDS_UCHAR;     /*  8 bit uchar    */
>  typedef unsigned char        TDS_TINYINT;   /*  8 bit int      */
> -typedef short                TDS_SMALLINT;  /* 16 bit int      */
> -typedef unsigned short       TDS_USMALLINT; /* 16 bit unsigned */
> -typedef long         TDS_INT;       /* 32 bit int      */
> -typedef unsigned long        TDS_UINT;      /* 32 bit unsigned */
> -typedef float                TDS_REAL;      /* 32 bit float    */
> -typedef double               TDS_FLOAT;     /* 64 bit float    */
> +typedef @smallint@           TDS_SMALLINT;  /* 16 bit int      */
> +typedef unsigned TDS_SMALLINT        TDS_USMALLINT; /* 16 bit unsigned */
> +typedef @int@                        TDS_INT;       /* 32 bit int      */
> +typedef unsigned TDS_INT     TDS_UINT;      /* 32 bit unsigned */
> +typedef @real@                       TDS_REAL;      /* 32 bit float    */
> +typedef @float@              TDS_FLOAT;     /* 64 bit float    */
>  typedef struct tdsnumeric
>  {
>          unsigned char         precision;
> 
> 
> ---
> You are currently subscribed to freetds as: [camber AT ais.org]
> To unsubscribe, forward this message to $subst('Email.Unsub')
>