RHEL6.4 nfsd periodically becomes blocked for 120 seconds, triggers hung task warnings

Issue

• nfsd locks up on RHEL 6.4 with the following backtrace:

INFO: task nfsd:43340 blocked for more than 240 seconds.
      Tainted: P           ---------------    2.6.32-431.20.3.el6.x86_64 #1
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
nfsd          D 0000000000000005     0 43340      2 0x00000080
 ffff880310a91c40 0000000000000046 0000000000000000 0000000000000000
 0000000000000000 0000000000000000 0000000000000000 0000000000000000
 ffff88082deff058 ffff880310a91fd8 000000000000fbc8 ffff88082deff058
Call Trace:
 [<ffffffff81529f9e>] __mutex_lock_slowpath+0x13e/0x180
 [<ffffffff81529e3b>] mutex_lock+0x2b/0x50
 [<ffffffffa067b15d>] nfsd_lookup_dentry+0xdd/0x500 [nfsd]
 [<ffffffffa067b5b3>] nfsd_lookup+0x33/0xd0 [nfsd]
 [<ffffffffa0684282>] nfsd3_proc_lookup+0x92/0xf0 [nfsd]
 [<ffffffffa0675425>] nfsd_dispatch+0xe5/0x230 [nfsd]
 [<ffffffffa0642844>] svc_process_common+0x344/0x640 [sunrpc]
 [<ffffffff81061d00>] ? default_wake_function+0x0/0x20
 [<ffffffffa0642e80>] svc_process+0x110/0x160 [sunrpc]
 [<ffffffffa0675b52>] nfsd+0xc2/0x160 [nfsd]
 [<ffffffffa0675a90>] ? nfsd+0x0/0x160 [nfsd]
 [<ffffffff8109abf6>] kthread+0x96/0xa0
 [<ffffffff8100c20a>] child_rip+0xa/0x20
 [<ffffffff8109ab60>] ? kthread+0x0/0xa0
 [<ffffffff8100c200>] ? child_rip+0x0/0x20