[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: compilation error in kdesupport
From:       "Raul Fernandes" <rgfernandes () gmail ! com>
Date:       2008-12-08 17:31:57
Message-ID: fa597860812080931m41e3bb48gaa1ff40166340f62 () mail ! gmail ! com
[Download RAW message or body]

> Please fix your system. The code is correct both in Qt and in KDE:
> http://www.opengroup.org/onlinepubs/000095399/functions/if_nametoindex.html
> That's the POSIX.1 definition. It says if_nametoindex is in net/if.h.

AFAIK the linux doesn't follow posix strictly. I don't know why but
they have moved the header.

> Reinforcing Thiago's comment here: Don't just blindly update your headers by
> copying the headers from the Kernel source tree. There are sanitized versions
> of the linux headers that must be used instead of the normal ones from the
> kernel. If you don't use them, you seriously risk this kind of problem since
> the upstream unsantized headers can and will change under your feet.

I didn't blindly update the headers. I have made a backup.
And I'm testing it. I know that I'm compiling a unstable version, so I
want to find the bugs.
And it is not a old and unmaintained version of a secondary
library/application, it is the most recent version of kernel.
If the code doesn't compiles with the new version of kernel, we can't
say that it is linux compatible, can we??
About the unsanitized headers, the glibc 2.9 compiles cleanly here
with these headers as other programs do.
Unfortunately, now I receive a "won't fix", but some time from now, we
will see a commit saying "Fix compile with recent kernel headers". It
happened before.
I quit. I won't try to make a patch that won't go upstream. Someone
else will do in the future.

Raul Fernandes
rgfernandes@gmail.com
 
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<
[prev in list] [next in list] [prev in thread] [next in thread]