[Monetdb-developers] ! Testing 'Current' sql: Results DIFFER from 2007.04.28_04-03-41's (GDKdebug=10) !

Ying Zhang Y.Zhang at cwi.nl
Sun Apr 29 14:58:59 CEST 2007


Hi Romulo,

I had the same problem as you some time ago.  gcc is by default less
picky than icc:(  I think you can add the option "--enable-strict" to
'configure', which should make your gcc more strict.

Regards,

Jennie



On Sun, Apr 29, 2007 at 12:14:55PM +0200, Romulo Goncalves wrote:
> Stefan Manegold wrote:
> > Romulo,
> > 
> > you most probably compiled with gcc (the default), right.
> > That's fine.
> > Moreover, gcc also works fine with testing.
> > 
> > However, icc (the Intel compiler) is a bit more "picky" (read: *accurate*)
> > and points out that in function create_func(), variable "sub_f" is set, but
> > never used --- looking at the code, I must say icc is (as always) very right
> > so:
> > 
> > In fucntion create_func() in sql/src/server/sql_psm.mx, variable "sub_f" is
> > declared and initialized (set) to NULL in line 508.
> > Then, in line 525, variable "sub_f" is set to 
> > "sql_bind_func_(s, fname, type_list)".
> > However, there is no further occurance of variable "sub_f" in function
> > create_func(); hence, it is set (see above) but *never* *used*...
> > 
> > ... either you "forgot" to used (the content of) variable "sub_f"
> > "somewhere" in function create_func(), or you don't need it.
> > In the former case, icc helped you to find a BUG in your code.
> > In the latter case, you should remove variable "sub_f" from function
> > create_func(), i.e., remove line 508 and change line 525
> > from
> >         if (!instantiate && (sub_f = sql_bind_func_(s, fname, type_list))) {
> > to
> >         if (!instantiate && sql_bind_func_(s, fname, type_list)) {
> > 
> Yes I know, I just like to detect these things before my commit, but if 
> the compiler is icc I prefer to wait for the compilation results during 
> the night  ;)
> 
> The fix is already check in.
> 
> Regards,
> Romulo
> > 
> > Stefan
> > 
> > 
> > On Sun, Apr 29, 2007 at 11:57:03AM +0200, Romulo Goncalves wrote:
> >> Project account wrote:
> >>> Testing results of  sql  in  /export/scratch2/monet/Testing/Current  with GDKdebug=10  DIFFER from those of  2007.04.28_04-03-41 !
> >>>
> >>> Maybe  YOUR checkins  since 28.04.2007, 04:03:41  caused some of the problems:
> >>>
> >>> buildtools		 
> >>> MonetDB		 
> >>> clients		 
> >>> MonetDB4		 
> >>> MonetDB5		mk at cwi.nl 
> >>> sql		mk at cwi.nl niels at cwi.nl romulog at users.sourceforge.net 
> >>>
> >>> Check  http://monetdb.cwi.nl/Development/TestWeb/Current/sql/  for details !
> >>>
> >>>
> >>> Old file: Mtest10..2007.04.28_04-03-41..out
> >>> New file: Mtest10..2007.04.29_04-16-39..out
> >>> ===========================================
> >>>
> >>>
> >>> Summary:
> >>> ========
> >>>
> >>>  * 1 New systems
> >>> !! 3 Systems now DON'T produce output !!
> >> I compiled my sources before the commit with optimization on:
> >> --enable-strict --enable-assert --enable-optimize --enable-bits=64
> >>
> >> Why the variable initialized, but never used is not detected by this 
> >> compilation? Which compilation must be done to detect these kind of errors?
> >>
> >> Regards,
> >> Romulo
> >>
> >>
> >> -------------------------------------------------------------------------
> >> This SF.net email is sponsored by DB2 Express
> >> Download DB2 Express C - the FREE version of DB2 express and take
> >> control of your XML. No limits. Just data. Click to get it now.
> >> http://sourceforge.net/powerbar/db2/
> >> _______________________________________________
> >> Monetdb-developers mailing list
> >> Monetdb-developers at lists.sourceforge.net
> >> https://lists.sourceforge.net/lists/listinfo/monetdb-developers
> >>
> > 
> 
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
> _______________________________________________
> Monetdb-developers mailing list
> Monetdb-developers at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/monetdb-developers




More information about the developers-list mailing list