Skip to main content.
home | support | download

Back to List Archive

Re: Assembler errors

From: Andrew Lord <andrewlord(at)not-real.internode.on.net>
Date: Tue Jul 22 2003 - 04:21:13 GMT
Hi Bill,

Thankyou for your response.

> > # # # # # # # # gcc -v
> > Reading specs from /usr/lib/gcc-lib/i386-redhat-linux/2.96/specs
> > gcc version 2.96 20000731 (Red Hat Linux 7.2 2.96-108.7.2)
>
> Please see:
>
>   http://gcc.gnu.org/gcc-2.96.html
>
> You might be wise to upgrade your verison of gcc.  I don't see any of
> the warning you are seeing.  I'm able to generate some of the warnings
> by adding additional compiler flags, though.

Unfortunately an upgrade is not an option as the site I'm compiling on is 
remote and the server operators are not permitted by Ensim to change any 
facet of their system :-(.  So I'm at the mercy of the server operator as to 
when they purchase an upgraded package from Ensim.

> Also, it would be more helpful to see these warnings on a currnet cvs
> version of swish.  Makes it easier to track down the problems by line
> numbers.

Yes I agree.  Just that I have the Swish::Indexes.pm module that you put 
together a while ago and have not had any success getting it to work with 
swish 2.3.  I did try it some time ago with Swish 2.3 & the other module that 
was written to allow Swish 2.3 to work with older modules, but no success.  
That work is archived now & Swish 2.2.3 works fine (well, on my own computer 
anyway) so I'll stick with it for the time being.

>    make >/dev/null
>
> will make the output easier to parse in your emails, too.
>
> > Do you have any further suggestions for how the 'incompatible pointer
> > type' errors can be overcome ?
>
> Those are warnings not errors.  They shouldn't keep swish-e from
> building.  Yes, they should be fixed -- over the years (and large number
> of programmers) have used different compilers and were lax in casting,
> and likely there's been an overuse of casting.

No they haven't stopped it from building and the tests all run fine.  I was 
simply unaware of whether the warnings might be indicative of particular 
situations in which the binary would not operate correctly/at all.  But if I 
understand you correctly, it seems that those warnings can probably be 
overlooked in running swish.  I'll give it a try as it is, then.

Thankyou for your help.

Cheers,

Andrew
Received on Tue Jul 22 04:21:23 2003