PCI: pciehp: fix possible endless loop in pcie_isr
authorKenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Tue, 3 Feb 2009 06:06:13 +0000 (15:06 +0900)
committerJesse Barnes <jbarnes@hobbes.lan>
Fri, 20 Mar 2009 02:29:28 +0000 (19:29 -0700)
Fix possible endless loop in pcie_isr.

Currently, pcie_isr() (interrupt service routine of pciehp) can end up in an
endless loop if the Slot Status register is set again immediately after being
cleared. According to the past discussion (see below URL) this case can happen
if the power fault detected bit is set during handling.

http://sourceforge.net/mailarchive/message.php?msg_id=20051130135409.A14918%40unix-os.sc.intel.com

Signed-off-by: Kenji Kaneshige <kaneshige.kenji@jp.fujitsu.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
drivers/pci/hotplug/pciehp_hpc.c

index 7a16c6897bb92698abc8edfb9af1d909f235453d..c1a312f240600ebe576834d3394d65e39c51dacf 100644 (file)
@@ -672,10 +672,11 @@ static irqreturn_t pcie_isr(int irq, void *dev_id)
                detected &= (PCI_EXP_SLTSTA_ABP | PCI_EXP_SLTSTA_PFD |
                             PCI_EXP_SLTSTA_MRLSC | PCI_EXP_SLTSTA_PDC |
                             PCI_EXP_SLTSTA_CC);
+               detected &= ~intr_loc;
                intr_loc |= detected;
                if (!intr_loc)
                        return IRQ_NONE;
-               if (detected && pciehp_writew(ctrl, PCI_EXP_SLTSTA, detected)) {
+               if (detected && pciehp_writew(ctrl, PCI_EXP_SLTSTA, intr_loc)) {
                        ctrl_err(ctrl, "%s: Cannot write to SLOTSTATUS\n",
                                 __func__);
                        return IRQ_NONE;