malloc(M_WAITOK) of "Mbuf", forcing M_NOWAIT with the following non-sleepable locks held:

Hello,

with current sources from today i get a whole lot of these on an nfs 
server when building world over nfs.

malloc(M_WAITOK) of "Mbuf", forcing M_NOWAIT with the following 
non-sleepable locks held:
exclusive sleep mutex inp (tcpinp) r = 0 (0xc13e5b1c) locked _at_ 
/storage/src/sys/netinet/tcp_usrreq.c:616
Stack backtrace:
backtrace(c0744604,cbb53a4c,1,179,c13f5700) at backtrace+0x17
witness_warn(5,0,c06d383e,c06bdebe,c06be99e) at witness_warn+0x1b5
uma_zalloc_arg(c0c21b00,cbb53a80,2,fa,0) at uma_zalloc_arg+0x4c
nfs_realign(c1650424,28,c06ccf37,29d,cbb53ac0) at nfs_realign+0xb4
nfsrv_getstream(c1650400,1,c06ccf37,1d5,0) at nfsrv_getstream+0x36d
nfsrv_rcv(c144813c,c1650400,1,19a,c13e745c) at nfsrv_rcv+0x19d
sowakeup(c144813c,c144818c,c06cacef,4e2,108) at sowakeup+0x101
tcp_input(c13f5700,14,c11a8000,1,0) at tcp_input+0x1120
ip_input(c13f5700,0,c06c8872,95,c0745638) at ip_input+0x945
netisr_processqueue(c0745638,0,c06c8872,fd,c1107240) at 
netisr_processqueue+0x8e
swi_net(0,0,c06bcf17,270,0) at swi_net+0xa3
ithread_loop(c10dad80,cbb53d48,c06bccfa,328,0) at ithread_loop+0x182
fork_exit(c0502590,c10dad80,cbb53d48) at fork_exit+0xc2
fork_trampoline() at fork_trampoline+0x8
--- trap 0x1, eip = 0, esp = 0xcbb53d7c, ebp = 0 ---

Regards,
flo