RHEL5 e1000 driver crashes when run as VMware guest

Issue

• e1000 driver tracebacks when running under VMware ESX4.
• RHEL5 guest network goes down after the log message "irq 59: nobody cared (try booting with the "irqpoll" option)".
• RHEL5 guest was not reachable via network, but it can be accessed via ESX console. A 'service network restart' restores the guest network connectivity.

• Following call trace logged in /var/log/messages during the issue.

  irq 59: nobody cared (try booting with the "irqpoll" option)
  
  Call Trace:
  <IRQ>  [<ffffffff800bc899>] __report_bad_irq+0x30/0x7d
  [<ffffffff800bcacc>] note_interrupt+0x1e6/0x227
  [<ffffffff800bbfc8>] __do_IRQ+0xbd/0x103
  [<ffffffff8006da2b>] do_IRQ+0xe7/0xf5
  [<ffffffff8005e615>] ret_from_intr+0x0/0xa
  [<ffffffff800123d1>] __do_softirq+0x51/0x133
  [<ffffffff8005f2fc>] call_softirq+0x1c/0x28
  [<ffffffff8006dba8>] do_softirq+0x2c/0x85
  [<ffffffff8005ec8e>] apic_timer_interrupt+0x66/0x6c
  <EOI>  [<ffffffff881f5d48>] :e1000:e1000_watchdog_task+0x574/0x5d8
  [<ffffffff881f57d4>] :e1000:e1000_watchdog_task+0x0/0x5d8
  [<ffffffff8004dc37>] run_workqueue+0x94/0xe4
  [<ffffffff8004a472>] worker_thread+0x0/0x122
  [<ffffffff8004a562>] worker_thread+0xf0/0x122
  [<ffffffff8008e16d>] default_wake_function+0x0/0xe
  [<ffffffff80032bdc>] kthread+0xfe/0x132
  [<ffffffff8005efb1>] child_rip+0xa/0x11
  [<ffffffff80172534>] vgacon_cursor+0x0/0x1a5
  [<ffffffff80032ade>] kthread+0x0/0x132
  [<ffffffff8005efa7>] child_rip+0x0/0x11
  
  handlers:
  [<ffffffff881f3d7b>] (e1000_intr+0x0/0x12c [e1000])
  Disabling IRQ #59
  NETDEV WATCHDOG: eth0: transmit timed out
  bonding: bond0: link status definitely down for interface eth0, disabling it
  bonding: bond0: now running without any active interface !