> Hello, all.
>
> When I do a build of the backends under Linux, I have to leave the UMAX
> backend out because for some weird reason, it starts to eat a very large
> chunk of memory. So large, it even ran me out of my fifty megabytes's
> worth of swap space. Adding more swap space does NOT help either.
>
> Yet when I compile the backends under FreeBSD, that problem does not
> manifest itself.
>
> I most recently did this with gcc 2.7.2.3 under Slackware 3.4 on one
> machine, and PGCC (based on EGCS 1.0.3) on another Slackware 3.4 system,
> same results.
>
> The FreeBSD system in question is 2.2.7 RELEASE, under which I do most of
> my heavy-duty debugging runs; Linux puts too much useless crap in my
> debugging logs, especially the recent work I'd done in locating my
> end-of-scan problem.
>
Hi Ian,
That is a problem of the gcc. If you turn off optimization like described in
the PROBELMS-file, you are able to compile the umax-backend without any
errors!
Bye
Oliver
-- Source code, list archive, and docs: http://www.mostang.com/sane/ To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com