[Monetdb-developers] overflow checks in the Pathfinder shredder?

Stefan Manegold Stefan.Manegold at cwi.nl
Wed Jun 21 14:02:00 CEST 2006


Jan,

sounds great. thanks!

btw: 

watch out that tests like (x > GDK_int_max) only work properly if x has a
larger domain than int!
otherwise, either the compiler already complains, or the result should
always be false...

to be on the save side, we should better use (x < GDK_int_max) or (x >=
GDK_int_max).

likewise for the other types.

Stefan


On Wed, Jun 21, 2006 at 01:47:48PM +0200, Jan Flokstra wrote:
> Stefan,
> 
> We check the main PRE number against INT_MAX and we check if the the size of a 
> BAT does not try grow beyond INT_MAX.
>  We do not check the 'level' and 'size' values against the max values you 
> mentioned. I will add these checks and change the max value for PRE numbers 
> from INT_MAX into GDK_oid_max,
> 
> Jan.
> 
> 
> On Wednesday 21 June 2006 13:39, Stefan Manegold wrote:
> > Jan F. (and/or others),
> >
> > did I understand you correctly on Tuesday that there are checks
> > in the pathfinder shredder (XQuery_0-12 release branch)
> > that detect overflows in case of huge documents, i.e.,
> >
> > 	level > GDK_chr_max
> > 	size  > GDK_int_max
> > 	OIDs  > GDK_oid_max
> >
> > and generate the respective errors?
> >
> > If not, shouldn't we add them, and trigger a proper error in case they
> > occur?
> >
> > The same question arises also for (intermediate) results during query
> > processing...
> >
> > Stefan
> 
> 

-- 
| Dr. Stefan Manegold | mailto:Stefan.Manegold at cwi.nl |
| CWI,  P.O.Box 94079 | http://www.cwi.nl/~manegold/  |
| 1090 GB Amsterdam   | Tel.: +31 (20) 592-4212       |
| The Netherlands     | Fax : +31 (20) 592-4312       |




More information about the developers-list mailing list