FreeTDS Development Group

[Chris Eleveld <ihermit2 AT yahoo.com>]

If we change it to UINT or INT we haev to verify there
are no swaps on it at on a Big Endian system like mine
we'll swap the wrong 2 bytes.  I believe we've already
taken care of this one though.  

I must admit not knowing the deep rustligns of tds I
don't know if there is some unspoken rule that all
negative message number should be suppresed or not.
on the wire it certianly appears to be 4 bytes.  The
values apssed through DB-Lib appear to be 4 byte
signed integers as well.  So I'm inclined to vote for
a signed 4 byte integer myself and maybe a warning to
the log in addition to suppresing the message since
it's likely a sign of a possible endian bug.

other oppinions?

--- Steve Langasek <vorlon AT netexpress.net> wrote:
> On Sat, Jan 26, 2002 at 10:10:42AM -0000, Dave
> Brotherstone wrote:
> > As discussed, a diff (I hope it's the right format
> - shout if it's not) of
> > the couple of changes to make the msg_number a
> TDS_UINT under dblib.
> 
> > Tested with SQL2000 and PHP/Apache.
> 
> My concern is that this is being done specifically
> *because* of the line 
> in dblib/dbutil.c:
> 
>         if( tds->msg_info->msg_number > 0 )
> 
> If you change msg_number to be a uint, then this
> comparison will always 
> be false.  But either this comparison was put here
> for a good reason, in 
> which case making the variable a uint is wrong; or
> it wasn't, in which 
> case the comparison should simply be eliminated.
> 
> If it's just a question of the storage size being
> too small, then it 
> should be changed to a TDS_INT, not a TDS_UINT.
> 
> Steve Langasek
> postmodern programmer
> 
> > Index: include/tds.h.in
> >
>
===================================================================
> > RCS file:
> /cvsroot/freetds/freetds/include/tds.h.in,v
> > retrieving revision 1.9
> > diff -u -r1.9 tds.h.in
> > --- include/tds.h.in        2001/12/03 00:06:14     1.9
> > +++ include/tds.h.in        2002/01/26 09:58:01
> > @@ -476,7 +476,7 @@
> >  typedef struct tds_msg_info {
> >        TDS_SMALLINT priv_msg_type;
> >        TDS_SMALLINT line_number;
> > -      TDS_SMALLINT msg_number;
> > +      TDS_UINT msg_number;
> >        TDS_SMALLINT msg_state;
> >        TDS_SMALLINT msg_level;
> >        TDS_CHAR *server;
> > Index: src/dblib/dbutil.c
> >
>
===================================================================
> > RCS file:
> /cvsroot/freetds/freetds/src/dblib/dbutil.c,v
> > retrieving revision 1.5
> > diff -u -r1.5 dbutil.c
> > --- src/dblib/dbutil.c      2001/11/14 04:52:33     1.5
> > +++ src/dblib/dbutil.c      2002/01/26 09:58:01
> > @@ -29,7 +29,7 @@
> >  
> >  
> >  extern int
>
(*g_dblib_msg_handler)(DBPROCESS*,TDS_SMALLINT,TDS_SMALLINT,TDS_SMALLINT,TDS_CHAR*,TDS_CHAR*,TDS_CHAR*);
> > -extern int
>
(*g_dblib_err_handler)(DBPROCESS*,TDS_SMALLINT,TDS_SMALLINT,TDS_SMALLINT,TDS_CHAR*,TDS_CHAR*);
> > +extern int
>
(*g_dblib_err_handler)(DBPROCESS*,TDS_SMALLINT,TDS_UINT,TDS_SMALLINT,TDS_CHAR*,TDS_CHAR*);
> >  
> >  /* The next 2 functions will be the reciever for
> the info and error messages
> >   * that come from the TDS layer.  The address of
> this function is passed to
> > 
> > 
> 
> Content-Description: footer
> > ---
> > You are currently subscribed to freetds as:
> [vorlon AT netexpress.net]
> > To unsubscribe, forward this message to
> $subst('Email.Unsub')
> 
> 

> ATTACHMENT part 2 application/pgp-signature 
> ---
> You are currently subscribed to freetds as:
> [ihermit2 AT yahoo.com]
> To unsubscribe, forward this message to
> $subst('Email.Unsub')
> 


__________________________________________________
Do You Yahoo!?
Great stuff seeking new owners in Yahoo! Auctions! 
http://auctions.yahoo.com