Hi !

Again a little problem....

I try to compile a source in C langage.
but esql return me a core.
while including .h in my code....

i think that the file created is too big ...
Under WinXP it compile and run fine.

But under AIX , it fails ....
INFORMIX-ESQL Version 9.51.UC3

I can't cut the file in differrent files, cos' it's generated by another
executable...


I know this is not an AIX newsgroup but still the problem may have
appear.... Is there a way for increasing the memory for the esql executable
?

in fact we've got 1Go Ram and 2 Go swap empty and ready to be used, but
the esql seems to indicates that esql use around 130Mo. (Neither that the
problem is 10ko nor 500ko)
and never more.


Maybe it's an esql issue (but i don't have answer from my support) so if
you've got info about this...

It will be very helpful =)

Thx by advance !!

hobbes wrote:
> Hi !
>
> Again a little problem....
>
> I try to compile a source in C langage.
> but esql return me a core.
> while including .h in my code....
>
> i think that the file created is too big ...
> Under WinXP it compile and run fine.
>
> But under AIX , it fails ....
> INFORMIX-ESQL Version 9.51.UC3
>
> I can't cut the file in differrent files, cos' it's generated by another
> executable...
>
>
> I know this is not an AIX newsgroup but still the problem may have
> appear.... Is there a way for increasing the memory for the esql executable
> ?
>
> in fact we've got 1Go Ram and 2 Go swap empty and ready to be used, but
> the esql seems to indicates that esql use around 130Mo. (Neither that the
> problem is 10ko nor 500ko)
> and never more.
>
>
> Maybe it's an esql issue (but i don't have answer from my support) so if
> you've got info about this...
>
> It will be very helpful =)
>
> Thx by advance !!
>
>
Core dumps are -- typically -- caused by code errors, and not by much
else. If you have the lint utility on the box, try this:

esql -e -I. -I.. file.ec
lint -mux -I${INFORMIXDIR}/incl/esql -I. -I.. file.c

Replace "-I. -I.." with wherever your .h file is. Your arguments to line
may be different than "-mux," read the man page.

Fix everything lint yammers at you about.

If you don't have lint, try compiling with gcc (if you have it) and fix
whatever it yammers at you about.

Thank you for the answer...

but i'm pretty sure that it is an esql issue...

> Core dumps are -- typically -- caused by code errors, and not by much
> else. If you have the lint utility on the box, try this:
>

Code seems OK.
I know, that's maybe not a reference, but it compile fine and finally run
under any Windows Env'.(and also many Unix ...)
But this version (for the moment) as only been compiled under Win ... but
it's the same source code at this time.

> esql -e -I. -I.. file.ec

that is the syntax i actually use.

i can't execute gcc, since no .c were generated after core dump.




If I remove some #include directive, and some code, i can finally build it
up....
But i can't split this file. It must stay as one and only one file.





ps:another thing .. esql is consumig a lot of memory under MS windows ...
Have you experienced the same thing ?

> Are you sure you're not trying to include a header that doesn't exist?

No, they all exists and are valid .

> Have you got a bunch of #ifdefs in your .h and are they correct?
Yes , but i removed all of them when i got my first core dump

> Have you (carefully) checked your include file? Can you make a tiny ESQL/C
> program (see below) that includes the .h and does it blow up (if so, the
> problem is in the .h file)?

Yes checked and double check.
the thing is that i include all the structs of the database....
nearly 300 structs in seperate includes file.

if i remove the one onto esqlc make the core, it do the same things on the
next one ... and on and on .....

All is OK with the include file ... other programs that use this structure
build and run well

> Remember, ESQL/C is just a preprocessor; all it does is read your source
> code and translate it to C for the compiler. Bad source? Core dump.

Yes i know that.... already having problem with SQLserver Preprocessor some
times ago.
And Oracle Preprocessor's doing the same problem with some sources....
but theses source were Huge... and manually created...


Thk you for all your indications, and your help...
If you've got some more idea do not hesitate !

Thk you once again !!

hobbes wrote:

>
>
> Yes i know that.... already having problem with SQLserver Preprocessor some
> times ago.
> And Oracle Preprocessor's doing the same problem with some sources....
> but theses source were Huge... and manually created...
>
>
> Thk you for all your indications, and your help...
> If you've got some more idea do not hesitate !
>
> Thk you once again !!
>
>
>
The only thing you have left is to start by commenting-out everything in
the header and then uncomment section by section until you find the one
that kills it, alas.

Good luck!