2.5.69, aar_rx_intr, irq 30: nobody cared!

May  6 13:15:37 test kernel: handlers:
May  6 13:15:37 test kernel: [aac_rx_intr+0/212]
May  6 13:15:37 test kernel: irq 30: nobody cared!
May  6 13:15:37 test kernel: Call Trace: [handle_IRQ_event+147/232]  [do_IRQ+194/304]
  [  common_interrupt+24/32]  [try_to_unmap_one+275/368]  [bio_put+24/48]  
  [end_bio_bh_io_sync+33/48]  [bio_endio+70/76]  [__end_that_request_first+256/464] 
  [end_that_request_first+23/28]  [scsi_end_request+36/160]  [scsi_io_completion+561/1100]
  [sd_rw_intr+470/480]  [scsi_finish_command+208/216]  [scsi_softirq+490/516]  
  [do_softirq+106/200]  [do_IRQ+288/304]  [default_idle+0/52]  [rest_init+0/72]  
  [common_interrupt+24/32]  [default_idle+0/52]  [rest_init+0/72]  [default_idle+41/52]  
  [cpu_idle+55/72]  [rest_init+69/72]  [start_kernel+379/384]

How spurious are these? I just looked at the OIMR programming and i can't see how this managed to slip by, perhaps someone really doesn't honour their interrupt disable bit.

function.linuxpower.ca