Hi Mikado, On Sun, 8 Jan 2006 01:25:05 -0800 (PST) Mikado <Mikado <[email protected]>> wrote: | Is there anything wrong in my codes? I think the problem is the | netlink_unicast(), because when I didn't call it, everything work well. Check whether nl_sk equals NULL. I don't know whether there's an assert() macro available for you, so try this: /* [...] */ if (nl_sk == NULL) { printk(KERN_ALERT "nltest: nl_sk is NULL!"); goto return_free; } netlink_unicast(nl_sk, nl_skb, pid, 0); return_free: kfree_skb(nl_skb); } A gdb backtrace of the user space part won't give you much, since you found out the problem lies in kernel space. What does netlink_unicast() do? Leslie -- gpg --keyserver pgp.mit.edu --recv-keys 0x52D70289
Attachment:
pgp5rMQ656k0P.pgp
Description: PGP signature
- References:
- Resend: Netlink socket problem
- From: Mikado <[email protected]>
- Resend: Netlink socket problem
- Prev by Date: Re: Why is 2.4.32 four times faster than 2.6.14.6??
- Next by Date: Re: [PATCH/RFC] Shared page tables
- Previous by thread: Resend: Netlink socket problem
- Next by thread: [2.6.15] running tcpdump on 3c905b causes freeze (reproducable)
- Index(es):
 |