| Message ID | 1282690240-5301-1-git-send-email-leitao@linux.vnet.ibm.com |
|---|---|
| State | Accepted, archived |
| Delegated to: | David Miller |
| Headers | show |
On Tue, Aug 24, 2010 at 03:50:40PM -0700, leitao@linux.vnet.ibm.com wrote: > Currently qlge can deadlock when the interface is going > down, and the mpi_port_cfg_work() is executing on another > processor. It happens because unregister_netdev() holds > the rtnl lock, and the mpi_port_cfg_work() also request > this lock. > > Since unregiter_netdev() may wait mpi_port_cfg_work(), who > also request the holding lock, it can cause an deadlock, > displaying the following error: > Thanks Breno. Signed-off-by: Ron Mercer <ron.mercer@qlogic.com> -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
From: Ron Mercer <ron.mercer@qlogic.com> Date: Wed, 25 Aug 2010 07:42:30 -0700 > On Tue, Aug 24, 2010 at 03:50:40PM -0700, leitao@linux.vnet.ibm.com wrote: >> Currently qlge can deadlock when the interface is going >> down, and the mpi_port_cfg_work() is executing on another >> processor. It happens because unregister_netdev() holds >> the rtnl lock, and the mpi_port_cfg_work() also request >> this lock. >> >> Since unregiter_netdev() may wait mpi_port_cfg_work(), who >> also request the holding lock, it can cause an deadlock, >> displaying the following error: >> > > Thanks Breno. > > Signed-off-by: Ron Mercer <ron.mercer@qlogic.com> Applied. -- To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/drivers/net/qlge/qlge_main.c b/drivers/net/qlge/qlge_main.c index c9f9754..5a24521 100644 --- a/drivers/net/qlge/qlge_main.c +++ b/drivers/net/qlge/qlge_main.c @@ -3889,11 +3889,8 @@ int ql_wol(struct ql_adapter *qdev) return status; } -static int ql_adapter_down(struct ql_adapter *qdev) +static void ql_cancel_all_work_sync(struct ql_adapter *qdev) { - int i, status = 0; - - ql_link_off(qdev); /* Don't kill the reset worker thread if we * are in the process of recovery. @@ -3905,6 +3902,15 @@ static int ql_adapter_down(struct ql_adapter *qdev) cancel_delayed_work_sync(&qdev->mpi_idc_work); cancel_delayed_work_sync(&qdev->mpi_core_to_log); cancel_delayed_work_sync(&qdev->mpi_port_cfg_work); +} + +static int ql_adapter_down(struct ql_adapter *qdev) +{ + int i, status = 0; + + ql_link_off(qdev); + + ql_cancel_all_work_sync(qdev); for (i = 0; i < qdev->rss_ring_count; i++) napi_disable(&qdev->rx_ring[i].napi); @@ -4727,6 +4733,7 @@ static void __devexit qlge_remove(struct pci_dev *pdev) struct net_device *ndev = pci_get_drvdata(pdev); struct ql_adapter *qdev = netdev_priv(ndev); del_timer_sync(&qdev->timer); + ql_cancel_all_work_sync(qdev); unregister_netdev(ndev); ql_release_all(pdev); pci_disable_device(pdev); @@ -4746,13 +4753,7 @@ static void ql_eeh_close(struct net_device *ndev) /* Disabling the timer */ del_timer_sync(&qdev->timer); - if (test_bit(QL_ADAPTER_UP, &qdev->flags)) - cancel_delayed_work_sync(&qdev->asic_reset_work); - cancel_delayed_work_sync(&qdev->mpi_reset_work); - cancel_delayed_work_sync(&qdev->mpi_work); - cancel_delayed_work_sync(&qdev->mpi_idc_work); - cancel_delayed_work_sync(&qdev->mpi_core_to_log); - cancel_delayed_work_sync(&qdev->mpi_port_cfg_work); + ql_cancel_all_work_sync(qdev); for (i = 0; i < qdev->rss_ring_count; i++) netif_napi_del(&qdev->rx_ring[i].napi);
Currently qlge can deadlock when the interface is going down, and the mpi_port_cfg_work() is executing on another processor. It happens because unregister_netdev() holds the rtnl lock, and the mpi_port_cfg_work() also request this lock. Since unregiter_netdev() may wait mpi_port_cfg_work(), who also request the holding lock, it can cause an deadlock, displaying the following error: "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. rmmod D 00000080c6c1d190 0 3993 2081 0x00008080 Call Trace: [c000000975f56ee0] [c0000000000152a0] .__switch_to+0x100/0x1d0 [c000000975f56f70] [c0000000005781b4] .schedule+0x3a4/0x8c0 [c000000975f570c0] [c000000000578e8c] .schedule_timeout+0x24c/0x350 [c000000975f571e0] [c000000000578a88] .wait_for_common+0x198/0x210 [c000000975f572c0] [c0000000000abbb4] .__cancel_work_timer+0x2c4/0x2e0 [c000000975f57400] [d0000000078e7a20] .ql_adapter_down+0x80/0x260 [qlge] [c000000975f574b0] [d0000000078e7d80] .qlge_close+0x70/0x130 [qlge] [c000000975f57540] [c000000000497ef8] .__dev_close+0x98/0xf0 [c000000975f575d0] [c000000000497f74] .dev_close+0x24/0x60 [c000000975f57650] [c000000000498080] .rollback_registered_many+0xd0/0x2b0 [c000000975f576f0] [c000000000498338] .rollback_registered+0x38/0x50 [c000000975f57780] [c0000000004983d8] .unregister_netdevice_queue+0x88/0xe0 [c000000975f57810] [c000000000498574] .unregister_netdev+0x24/0x40 [c000000975f57890] [d0000000078f6f38] .qlge_remove+0x3c/0x78 [qlge] [c000000975f57920] [c0000000002d9298] .pci_device_remove+0x48/0x90 [c000000975f579a0] [c000000000372850] .__device_release_driver+0xa0/0x130 [c000000975f57a30] [c000000000372a08] .driver_detach+0x128/0x150 [c000000975f57ad0] [c000000000371134] .bus_remove_driver+0xc4/0x1a0 [c000000975f57b70] [c00000000037357c] .driver_unregister+0x8c/0xd0 [c000000975f57c00] [c0000000002d968c] .pci_unregister_driver+0x5c/0x110 [c000000975f57ca0] [d0000000078f6ee4] .qlge_exit+0x1c/0x34 [qlge] Signed-off-by: Breno Leitao <leitao@linux.vnet.ibm.com> --- drivers/net/qlge/qlge_main.c | 23 ++++++++++++----------- 1 files changed, 12 insertions(+), 11 deletions(-)