Hi all
I am attempting to port the cheops network mapping/diagnostic program
from Linux to FreeBSD (see www.marko.net/cheops). One of the first snags
I have hit comes in using SIOCGIFCONF to queries the network interface
names and addresses.
The cheops code assumes that the buffer returned will
Just a follow up on this: on Stevens vol 2 pg 117, code line 299, is the
implication that the returned buffer DOES hold an array of struct ifreq
elements. So this does seem to indicate that something may be broken on
FreeBSD. At the very least there is some ambiguity - is this an array of
struct
On Tue, Jan 09, 2001 at 12:21:11PM +0200, Graham Wheeler wrote:
I am attempting to port the cheops network mapping/diagnostic program
from Linux to FreeBSD (see www.marko.net/cheops). One of the first snags
I have hit comes in using SIOCGIFCONF to queries the network interface
names and
On Tue, Jan 09, 2001 at 12:30:44PM +0200, Graham Wheeler wrote:
Just a follow up on this: on Stevens vol 2 pg 117, code line 299, is the
implication that the returned buffer DOES hold an array of struct ifreq
elements. So this does seem to indicate that something may be broken on
FreeBSD. At
David Malone wrote:
On Tue, Jan 09, 2001 at 12:21:11PM +0200, Graham Wheeler wrote:
I am attempting to port the cheops network mapping/diagnostic program
from Linux to FreeBSD (see www.marko.net/cheops). One of the first snags
I have hit comes in using SIOCGIFCONF to queries the
David Malone wrote:
If you read the paragraph below that code, it notes that the ifreq
structures are of variable length. The spacing depends on the size
of the returned info.
That's true. In which case the cheops code is wrong, as it iterates
through the list by incrementing a pointer to a
David Malone wrote:
If you read the paragraph below that code, it notes that the ifreq
structures are of variable length. The spacing depends on the size
of the returned info.
David.
Something that isn't clear to me - do you know (Mark for Linux, Dave or
someone else for FreeBSD)
Something that isn't clear to me - do you know (Mark for Linux, Dave or
someone else for FreeBSD) whether it is reasonable to assume the
ifr_name if the struct ifreq will be NUL terminated? I know that the
name in a struct sockaddr_dl is not necessarily so terminated, but for
the ifr_name
Marko - does Linux have getifaddrs()? I somehow doubt it...
Linux should have getifaddrs() if it has support for IPv6 in
userland libraries. There is an implimentation of it at:
http://www.linux-ipv6.org/cvsweb/libinet6/?cvsroot=usagi-libc
David.
To Unsubscribe: send mail to [EMAIL
On Tue, Jan 09, 2001 at 03:39:22PM +0200, Graham Wheeler wrote:
David Malone wrote:
If you read the paragraph below that code, it notes that the ifreq
structures are of variable length. The spacing depends on the size
of the returned info.
That's true. In which case the cheops code
At 10:51 AM 01/09/2001, David Malone wrote:
Marko - does Linux have getifaddrs()? I somehow doubt it...
Linux should have getifaddrs() if it has support for IPv6 in
userland libraries. There is an implimentation of it at:
http://www.linux-ipv6.org/cvsweb/libinet6/?cvsroot=usagi-libc
11 matches
Mail list logo
