On Fri, 2 Oct 2009, Christian Borntraeger wrote:

since commit 02b51df1b07b4e9ca823c89284e704cadb323cd1 (proc connector: add event 
for process becoming session leader) we have the following warning:
Badness at kernel/softirq.c:143
[...]
Krnl PSW : 0404c00180000000 00000000001481d4 (local_bh_enable+0xb0/0xe0)
[...]
Call Trace:
([<000000013fe04100>] 0x13fe04100)
 [<000000000048a946>] sk_filter+0x9a/0xd0
 [<000000000049d938>] netlink_broadcast+0x2c0/0x53c
 [<00000000003ba9ae>] cn_netlink_send+0x272/0x2b0
 [<00000000003baef0>] proc_sid_connector+0xc4/0xd4
 [<0000000000142604>] __set_special_pids+0x58/0x90
 [<0000000000159938>] sys_setsid+0xb4/0xd8
 [<00000000001187fe>] sysc_noemu+0x10/0x16
 [<00000041616cb266>] 0x41616cb266

The warning is
--->    WARN_ON_ONCE(in_irq() || irqs_disabled());

The network code must not be called with disabled interrupts but
sys_setsid holds the tasklist_lock with spinlock_irq while calling
the connector. 
After a discussion we agreed that we can move proc_sid_connector
from __set_special_pids to sys_setsid.
We also agreed that it is sufficient to change the check from
task_session(curr) != pid into err > 0, since if we don't change the
session, this means we were already the leader and return -EPERM.

One last thing:
There is also daemonize(), and some people might want to get a
notification in that case. Since daemonize() is only needed if a user
space does kernel_thread this does not look important (and there seems
to be no consensus if this connector should be called in daemonize). If
we really want this, we can add proc_sid_connector to daemonize() in an
additional patch (Scott?)

Signed-off-by: Christian Borntraeger <redacted>
CCed: Scott James Remnant [off-list ref]
CCed: Matt Helsley [off-list ref]
CCed: David S. Miller [off-list ref]
Acked-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Evgeniy Polyakov <redacted>

Acked-by: David Rientjes <rientjes@google.com>

I was getting the same softirq warnings and later in slub in slab_alloc() 
at might_sleep_if(gfpflags & __GFP_WAIT) from __alloc_skb(..., GFP_KERNEL) 
in this context that would infinitely spam my log.