FreeTDS Development Group

["Craig A. Berry" <craigberry AT mac.com>]

At 01:29 PM 1/28/2003 -0500, Lowden, James K wrote:

>Indeed, the TDS protocol does put binary floating point data on the wire.
>For TDS 5.0, the format is established at login time (see Sybase's "TDS 5.0
>Functional Specification", page 175).  The 8-byte format is the IEEE 754
>float type, unless it's VAX 'D' or 'ND5000'.  The 4-byte form is labelled
>just "IEEE floating point numbers", unless it's "VAX 'F' or 'ND5000'" form.

Looks like we'll have to punt on VAX for now.  One would have to implement 
conversions from IEEE and/or make FreeTDS request VAX floating point (thus 
depending on the server to do the conversions, which may or may not be 
available in recent versions of Sybase even, much less Microsoft).

>As a matter of architecture, this is really a link-time issue:  

[discussion of link-time approaches snipped]

If you assume that you never relink without reconfiguring and rebuilding, 
then it is purely link-time.  If you don't assume that then it also involves 
other stages of the build process.  Even aside from that, though, I haven't 
come up with a way of doing it at link time.

>The only solution I have atm is not much different from yours: rather than
>rename the functions, wrap their entirety in #if HAVE_AUTOCONF ||
>!HAVE_STRTOK or whatever.  IOW, "produce this code if autoconf was used or
>there's no strtok()".  Unixy systems will always produce the code (insofar
>as the preprocessor is concerned) but the makefile will never compile it.
>VMS systems will not define HAVE_AUTOCONF, leaving code production to the
>HAVE_STRTOK test.  
>
>Your compiler would produce a module with no function;  the linker would
>scoop it up but no name would be exported.  That approach would be slightly
>better, I assert, because 
>
>1.  the names are preserved
>2.  the preprocessor isn't interceding in linkage
>3.  the resulting binary is smaller
>
>The unixy systems would be unaffected by this change, except for the
>complexity.  


This approach was actually my first thought before I went the redefining 
route.  If it's cleaner to you, it works great for me and I include a patch 
to do that below.  I've omitted the HAVE_AUTOCONF bit since if I understand 
correctly it will be redundant:  systems that use autoconf to decide they 
don't need the functions will never bother to compile them and thus won't 
see that anyway; is that correct?


--- src/replacements/asprintf.c;-0      Sun Nov 17 05:28:12 2002
+++ src/replacements/asprintf.c Tue Jan 28 23:00:56 2003
@@ -15,6 +15,8 @@
 static void *no_unused_var_warn[] = { software_version, no_unused_var_warn };
 
 
+#if !HAVE_ASPRINTF
+
 int
 asprintf(char **ret, const char *fmt, ...)
 {
@@ -26,3 +28,5 @@
        va_end(ap);
        return len;
 }
+
+#endif /* !HAVE_ASPRINTF */
--- src/replacements/atoll.c;-0 Sun Nov 17 05:28:12 2002
+++ src/replacements/atoll.c    Tue Jan 28 23:05:10 2003
@@ -29,8 +29,12 @@
 static char software_version[] = "$Id: atoll.c,v 1.7 2002/11/17 11:28:12 
freddy77 Exp $";
 static void *no_unused_var_warn[] = { software_version, no_unused_var_warn };
 
+#if !HAVE_ATOLL
+
 tds_sysdep_int64_type
 atoll(const char *nptr)
 {
        return atol(nptr);
 }
+
+#endif /* !HAVE_ATOLL */
--- src/replacements/strtok_r.c;-0      Sun Nov 17 05:28:12 2002
+++ src/replacements/strtok_r.c Tue Jan 28 23:05:11 2003
@@ -19,6 +19,8 @@
 
 #include "replacements.h"
 
+#if !HAVE_STRTOK_R
+
 char *
 strtok_r(char *str, const char *sep, char **lasts)
 {
@@ -39,3 +41,5 @@
        }
        return str;
 }
+
+#endif /* !HAVE_STRTOK_R */
--- src/replacements/vasprintf.c;-0     Mon Dec  9 10:40:59 2002
+++ src/replacements/vasprintf.c        Tue Jan 28 23:03:18 2003
@@ -39,6 +39,9 @@
 #endif
 
 #define CHUNKSIZE 512
+
+#if !HAVE_VASPRINTF
+
 int
 vasprintf(char **ret, const char *fmt, va_list ap)
 {
@@ -103,3 +106,6 @@
        return len;
 #endif /* HAVE_VSNPRINTF */
 }
+
+#endif /* !HAVE_VASPRINTF */
+
[end of patch]