FreeTDS Development Group

["ZIGLIO Frediano" <Frediano.Ziglio AT vodafone.com>]

> 
> On Fri, 17 Oct 2003, James K. Lowden wrote:
> > On Fri, 17 Oct 2003, Peter Giorgilli 
> > <pgiorgilli AT theage.fairfax.com.au>
> > wrote:
> 
> > >Basically, what was happening was that the "_SQLExecute" 
> function was 
> > >not trapping the "TDS_FAIL" return code (generated as a result  of 
> > >the connection dying) and returning "SQL_SUCCESS" by default. The  
> > >"SQLMoreResults" was in turn being called, when strictly 
> it shouldn't  
> > >have been, the end result being that the process would end up in a 
> > >tight  loop.
> > ...
> > > After all of that, the patch below seems to have done the trick.
> 
> > Thanks for the work and the patch, Peter.  :-)
> 
> I'm pretty sure this was already fixed in cvs with one of the 
> 'sqlexec succeeds too much' patches.
> 
> Have Fun!
> Peter

Yes and not...
Patch is really old (IMHO 0.61 is prehistoric compared to CVS...)
however problem is present.
Attached a patch that fix TDS_FAIL handling from
tds_process_result_tokens.

freddy77

Attachment: vedi.diff.gz
Description: vedi.diff.gz