OSDN Git Service

iwlwifi: pcie: disable L1 Active after pci_enable_device
authorEmmanuel Grumbach <emmanuel.grumbach@intel.com>
Mon, 29 Jul 2013 20:05:18 +0000 (23:05 +0300)
committerJohannes Berg <johannes.berg@intel.com>
Wed, 31 Jul 2013 06:56:34 +0000 (08:56 +0200)
As Arjan pointed out, we mustn't do anything related to PCI
configuration until the device is properly enabled with
pci_enable_device().

Cc: stable@vger.kernel.org
Reported-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
drivers/net/wireless/iwlwifi/pcie/trans.c

index 96cfcdd..390e2f0 100644 (file)
@@ -1502,16 +1502,16 @@ struct iwl_trans *iwl_trans_pcie_alloc(struct pci_dev *pdev,
        spin_lock_init(&trans_pcie->reg_lock);
        init_waitqueue_head(&trans_pcie->ucode_write_waitq);
 
-       /* W/A - seems to solve weird behavior. We need to remove this if we
-        * don't want to stay in L1 all the time. This wastes a lot of power */
-       pci_disable_link_state(pdev, PCIE_LINK_STATE_L0S | PCIE_LINK_STATE_L1 |
-                              PCIE_LINK_STATE_CLKPM);
-
        if (pci_enable_device(pdev)) {
                err = -ENODEV;
                goto out_no_pci;
        }
 
+       /* W/A - seems to solve weird behavior. We need to remove this if we
+        * don't want to stay in L1 all the time. This wastes a lot of power */
+       pci_disable_link_state(pdev, PCIE_LINK_STATE_L0S | PCIE_LINK_STATE_L1 |
+                              PCIE_LINK_STATE_CLKPM);
+
        pci_set_master(pdev);
 
        err = pci_set_dma_mask(pdev, DMA_BIT_MASK(36));