[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20200227132231.457798739@linuxfoundation.org>
Date: Thu, 27 Feb 2020 14:35:55 +0100
From: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To: linux-kernel@...r.kernel.org
Cc: Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
stable@...r.kernel.org, Morumuri Srivalli <smorumu1@...ibm.com>,
Alexander Duyck <alexander.h.duyck@...ux.intel.com>,
David Dai <zdai@...ux.vnet.ibm.com>,
Aaron Brown <aaron.f.brown@...el.com>,
Kai-Heng Feng <kai.heng.feng@...onical.com>
Subject: [PATCH 5.4 015/135] e1000e: Use rtnl_lock to prevent race conditions between net and pci/pm
From: Alexander Duyck <alexander.h.duyck@...ux.intel.com>
commit a7023819404ac9bd2bb311a4fafd38515cfa71ec upstream.
This patch is meant to address possible race conditions that can exist
between network configuration and power management. A similar issue was
fixed for igb in commit 9474933caf21 ("igb: close/suspend race in
netif_device_detach").
In addition it consolidates the code so that the PCI error handling code
will essentially perform the power management freeze on the device prior to
attempting a reset, and will thaw the device afterwards if that is what it
is planning to do. Otherwise when we call close on the interface it should
see it is detached and not attempt to call the logic to down the interface
and free the IRQs again.
>>From what I can tell the check that was adding the check for __E1000_DOWN
in e1000e_close was added when runtime power management was added. However
it should not be relevant for us as we perform a call to
pm_runtime_get_sync before we call e1000_down/free_irq so it should always
be back up before we call into this anyway.
Reported-by: Morumuri Srivalli <smorumu1@...ibm.com>
Signed-off-by: Alexander Duyck <alexander.h.duyck@...ux.intel.com>
Tested-by: David Dai <zdai@...ux.vnet.ibm.com>
Tested-by: Aaron Brown <aaron.f.brown@...el.com>
Cc: Kai-Heng Feng <kai.heng.feng@...onical.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
---
drivers/net/ethernet/intel/e1000e/netdev.c | 68 ++++++++++++++---------------
1 file changed, 35 insertions(+), 33 deletions(-)
--- a/drivers/net/ethernet/intel/e1000e/netdev.c
+++ b/drivers/net/ethernet/intel/e1000e/netdev.c
@@ -4713,12 +4713,12 @@ int e1000e_close(struct net_device *netd
pm_runtime_get_sync(&pdev->dev);
- if (!test_bit(__E1000_DOWN, &adapter->state)) {
+ if (netif_device_present(netdev)) {
e1000e_down(adapter, true);
e1000_free_irq(adapter);
/* Link status message must follow this format */
- pr_info("%s NIC Link is Down\n", adapter->netdev->name);
+ pr_info("%s NIC Link is Down\n", netdev->name);
}
napi_disable(&adapter->napi);
@@ -6309,10 +6309,14 @@ static int e1000e_pm_freeze(struct devic
{
struct net_device *netdev = dev_get_drvdata(dev);
struct e1000_adapter *adapter = netdev_priv(netdev);
+ bool present;
+ rtnl_lock();
+
+ present = netif_device_present(netdev);
netif_device_detach(netdev);
- if (netif_running(netdev)) {
+ if (present && netif_running(netdev)) {
int count = E1000_CHECK_RESET_COUNT;
while (test_bit(__E1000_RESETTING, &adapter->state) && count--)
@@ -6324,6 +6328,8 @@ static int e1000e_pm_freeze(struct devic
e1000e_down(adapter, false);
e1000_free_irq(adapter);
}
+ rtnl_unlock();
+
e1000e_reset_interrupt_capability(adapter);
/* Allow time for pending master requests to run */
@@ -6571,6 +6577,30 @@ static void e1000e_disable_aspm_locked(s
__e1000e_disable_aspm(pdev, state, 1);
}
+static int e1000e_pm_thaw(struct device *dev)
+{
+ struct net_device *netdev = dev_get_drvdata(dev);
+ struct e1000_adapter *adapter = netdev_priv(netdev);
+ int rc = 0;
+
+ e1000e_set_interrupt_capability(adapter);
+
+ rtnl_lock();
+ if (netif_running(netdev)) {
+ rc = e1000_request_irq(adapter);
+ if (rc)
+ goto err_irq;
+
+ e1000e_up(adapter);
+ }
+
+ netif_device_attach(netdev);
+err_irq:
+ rtnl_unlock();
+
+ return rc;
+}
+
#ifdef CONFIG_PM
static int __e1000_resume(struct pci_dev *pdev)
{
@@ -6638,26 +6668,6 @@ static int __e1000_resume(struct pci_dev
}
#ifdef CONFIG_PM_SLEEP
-static int e1000e_pm_thaw(struct device *dev)
-{
- struct net_device *netdev = dev_get_drvdata(dev);
- struct e1000_adapter *adapter = netdev_priv(netdev);
-
- e1000e_set_interrupt_capability(adapter);
- if (netif_running(netdev)) {
- u32 err = e1000_request_irq(adapter);
-
- if (err)
- return err;
-
- e1000e_up(adapter);
- }
-
- netif_device_attach(netdev);
-
- return 0;
-}
-
static int e1000e_pm_suspend(struct device *dev)
{
struct pci_dev *pdev = to_pci_dev(dev);
@@ -6829,16 +6839,11 @@ static void e1000_netpoll(struct net_dev
static pci_ers_result_t e1000_io_error_detected(struct pci_dev *pdev,
pci_channel_state_t state)
{
- struct net_device *netdev = pci_get_drvdata(pdev);
- struct e1000_adapter *adapter = netdev_priv(netdev);
-
- netif_device_detach(netdev);
+ e1000e_pm_freeze(&pdev->dev);
if (state == pci_channel_io_perm_failure)
return PCI_ERS_RESULT_DISCONNECT;
- if (netif_running(netdev))
- e1000e_down(adapter, true);
pci_disable_device(pdev);
/* Request a slot slot reset. */
@@ -6904,10 +6909,7 @@ static void e1000_io_resume(struct pci_d
e1000_init_manageability_pt(adapter);
- if (netif_running(netdev))
- e1000e_up(adapter);
-
- netif_device_attach(netdev);
+ e1000e_pm_thaw(&pdev->dev);
/* If the controller has AMT, do not set DRV_LOAD until the interface
* is up. For all other cases, let the f/w know that the h/w is now
Powered by blists - more mailing lists