ASPM: Fix pcie devices with non-pcie children
authorMatthew Garrett <mjg@redhat.com>
Tue, 27 Mar 2012 14:17:41 +0000 (10:17 -0400)
committerLinus Torvalds <torvalds@linux-foundation.org>
Sat, 31 Mar 2012 19:49:56 +0000 (12:49 -0700)
Since 3.2.12 and 3.3, some systems are failing to boot with a BUG_ON.
Some other systems using the pata_jmicron driver fail to boot because no
disks are detected.  Passing pcie_aspm=force on the kernel command line
works around it.

The cause: commit 4949be16822e ("PCI: ignore pre-1.1 ASPM quirking when
ASPM is disabled") changed the behaviour of pcie_aspm_sanity_check() to
always return 0 if aspm is disabled, in order to avoid cases where we
changed ASPM state on pre-PCIe 1.1 devices.

This skipped the secondary function of pcie_aspm_sanity_check which was
to avoid us enabling ASPM on devices that had non-PCIe children, causing
trouble later on.  Move the aspm_disabled check so we continue to honour
that scenario.

Addresses https://bugzilla.kernel.org/show_bug.cgi?id=42979 and
          http://bugs.debian.org/665420

Reported-by: Romain Francoise <romain@orebokech.com> # kernel panic
Reported-by: Chris Holland <bandidoirlandes@gmail.com> # disk detection trouble
Signed-off-by: Matthew Garrett <mjg@redhat.com>
Cc: stable@vger.kernel.org
Tested-by: Hatem Masmoudi <hatem.masmoudi@gmail.com> # Dell Latitude E5520
Tested-by: janek <jan0x6c@gmail.com> # pata_jmicron with JMB362/JMB363
[jn: with more symptoms in log message]
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
drivers/pci/pcie/aspm.c

index 4bdef24..b500840 100644 (file)
@@ -508,9 +508,6 @@ static int pcie_aspm_sanity_check(struct pci_dev *pdev)
        int pos;
        u32 reg32;
 
-       if (aspm_disabled)
-               return 0;
-
        /*
         * Some functions in a slot might not all be PCIe functions,
         * very strange. Disable ASPM for the whole slot
@@ -519,6 +516,16 @@ static int pcie_aspm_sanity_check(struct pci_dev *pdev)
                pos = pci_pcie_cap(child);
                if (!pos)
                        return -EINVAL;
+
+               /*
+                * If ASPM is disabled then we're not going to change
+                * the BIOS state. It's safe to continue even if it's a
+                * pre-1.1 device
+                */
+
+               if (aspm_disabled)
+                       continue;
+
                /*
                 * Disable ASPM for pre-1.1 PCIe device, we follow MS to use
                 * RBER bit to determine if a function is 1.1 version device