I've just installed 3.2.2. I'm using a config that was working fine with 3.1 and all seems ok.
I get one error that I don't get on 3.1 at startup.
This is repeated for all SUBST statements in the config.
[1293000.484026] INFO: task kamailio:19912 blocked for more than 120 seconds.
[1293000.484081] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[1293000.484169] kamailio D c4608e58 0 19912 19899 0x00000000
[1293000.484173] f6b5d980 00000082 40827cc6 c4608e58 000006cf c1441e20 c1441e20 000000fd
[1293000.484177] f6b5db3c c4608e20 00000000 4085cc46 00049753 ffe3a684 ffffffff f7077bac
[1293000.484180] f6b5d9ac f6b5db3c f6b5d980 c102b2b7 00000001 c4608e20 00000004 00000005
[1293000.484184] Call Trace:
[1293000.484191] [<c102b2b7>] ? check_preempt_wakeup+0x18c/0x202
[1293000.484195] [<c127f55f>] ? __mutex_lock_common+0xe8/0x13b
[1293000.484198] [<c127f5c1>] ? __mutex_lock_slowpath+0xf/0x11
[1293000.484200] [<c127f652>] ? mutex_lock+0x17/0x24
[1293000.484202] [<c127f652>] ? mutex_lock+0x17/0x24
[1293000.484205] [<c123f68b>] ? unix_dgram_recvmsg+0x49/0x1e4
[1293000.484207] [<c10268bd>] ? __wake_up_common+0x34/0x59
[1293000.484210] [<c102a97c>] ? __wake_up_sync_key+0x33/0x49
[1293000.484213] [<c11dd3b0>] ? __sock_recvmsg+0x50/0x58
[1293000.484215] [<c11dd8e7>] ? sock_recvmsg+0x9d/0xb6
[1293000.484218] [<c104a652>] ? autoremove_wake_function+0x0/0x2d
[1293000.484221] [<c1092950>] ? __alloc_pages_nodemask+0xf3/0x4d9
[1293000.484224] [<c108e16b>] ? find_lock_page+0x1b/0x36
[1293000.484227] [<c1025769>] ? kmap_atomic_prot+0xd7/0xfc
[1293000.484230] [<c113cd18>] ? cpumask_any_but+0x20/0x2b
[1293000.484232] [<c11de830>] ? sys_recvfrom+0xc0/0x120
[1293000.484235] [<c1143b63>] ? copy_from_user+0x27/0x10e
[1293000.484237] [<c1025769>] ? kmap_atomic_prot+0xd7/0xfc
[1293000.484240] [<c10a2deb>] ? handle_mm_fault+0x982/0xa22
[1293000.484243] [<c104a652>] ? autoremove_wake_function+0x0/0x2d
[1293000.484246] [<c10baa1d>] ? fsnotify_access+0x5a/0x61
[1293000.484248] [<c11decee>] ? sys_socketcall+0x123/0x1aa
[1293000.484251] [<c100813b>] ? sysenter_do_call+0x12/0x28
19912 is MI datagram. I have been getting these errors since 1.x and what I normally do is echo 0 > /proc/sys/kernel/hung_task_timeout_secs to get rid of it.
It must be a module that I am loading. Is this possible to debug? I can again get rid of the messages but would love to know whats causing it.