Re: [PATCH] headers_check fix: linux/netfilter/xt_osf.h

[Sam Ravnborg <sam@ravnborg.org>]

On Sat, Jun 27, 2009 at 09:52:09PM +0400, Evgeniy Polyakov wrote:
> Hi.
> 
> On Sat, Jun 27, 2009 at 10:01:04PM +0530, Jaswinder Singh Rajput (jaswinder@kernel.org) wrote:
> > fix the following 'make headers_check' warnings:
> > 
> >   usr/include/linux/netfilter/xt_osf.h:40: found __[us]{8,16,32,64} type without #include <linux/types.h>
> > 
> > Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com>
> 
> Looks good, thank you.
> Patrick, please apply.

What about the other potential issues in the same file?
struct xt_osf_opt {
        __u16                   kind, length;
        struct xt_osf_wc        wc;
};

Do we know that struct xt_osf_wc is always aligned at a two byte
address also on 64 bit?
Do we know that sizeof(struct xt_osf_opt) is the same
on all platforms?

struct xt_osf_user_finger {
        struct xt_osf_wc        wss;

        __u8                    ttl, df;
        __u16                   ss, mss;
        __u16                   opt_num;

        char                    genre[MAXGENRELEN];
        char                    version[MAXGENRELEN];
        char                    subtype[MAXGENRELEN];

        /* MAX_IPOPTLEN is maximum if all options are NOPs or EOLs */
        struct xt_osf_opt       opt[MAX_IPOPTLEN];
};

Do we know that opt[MAX_IPOPTLEN] always start at the same offset
with different architectures?

struct xt_osf_nlmsg {
        struct xt_osf_user_finger       f;
        struct iphdr            ip;
        struct tcphdr           tcp;
};

We do not knwo struct iphdr/tcphdr - missing include.
Do we know the alignment of the above structs?
Are they always the same on all archs?

	Sam