forcedeth MCP55 msi problems

forcedeth MCP55 msi problems

[Andreas Petlund]

[-- Attachment #1: Type: text/plain, Size: 984 bytes --]

I have a machine with an Asus motherboard (P5N32-SLI Premium - NVidia  
MCP55 chipset) where the forcedeth module does not work with msi. This  
seems to bother quite some people ( see https://bugs.launchpad.net/ubuntu/+source/linux/+bug/136836 
  ).

Reading various forum posts makes me inclined to believe this is a  
firmware problem, but Asus have no fix with their latest BIOS-upgrade  
(and I am starting to believe that we will never see one as the last  
update came in March this year).

Is it an option to workaround this problem using a hardware-specific  
quirk that disables msi/msix for the device, or will that affect a lot  
of users that have no issues with this? On my system the NIC  
identifies as "cb84". I don't know if this is a common factor for all  
who has this problem, or only for my motherboard.
Maybe the benefit of making it work with all MCP55-NICs is worth  
disabling msi/msix also for unaffected ones?

Does anybody have alternative suggestions?


[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 2213 bytes --]

[PATCH] pci: Added quirk to disable msi for MCP55 NIC on Asus P5N32-SLI Premium

[Andreas Petlund]

Andreas Petlund wrote:
> I have a machine with an Asus motherboard (P5N32-SLI Premium - NVidia 
> MCP55 chipset) where the forcedeth module does not work with msi. This 
> seems to bother quite some people ( see 
> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/136836 ).
>
> Reading various forum posts makes me inclined to believe this is a 
> firmware problem, but Asus have no fix with their latest BIOS-upgrade 
> (and I am starting to believe that we will never see one as the last 
> update came in March this year).
>
> Is it an option to workaround this problem using a hardware-specific 
> quirk that disables msi/msix for the device, or will that affect a lot 
> of users that have no issues with this? On my system the NIC 
> identifies as "cb84". I don't know if this is a common factor for all 
> who has this problem, or only for my motherboard.
> Maybe the benefit of making it work with all MCP55-NICs is worth 
> disabling msi/msix also for unaffected ones?
>
> Does anybody have alternative suggestions?
>
This patch detects the known affected motherboard and disables msi for 
the MCP55 NIC devices, allowing normal netwrk operation without manually 
having to reload forcedeth module with msi=0 and msix=0.

Signed-off-by: Andreas Petlund <andreas@petlund.no>
---
 drivers/pci/quirks.c |   16 ++++++++++++++++
 1 files changed, 16 insertions(+), 0 deletions(-)

diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
index 5049a47..1bbd912 100644
--- a/drivers/pci/quirks.c
+++ b/drivers/pci/quirks.c
@@ -1828,6 +1828,22 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_SERVERWORKS,
                         PCI_DEVICE_ID_SERVERWORKS_HT1000_PXB,
                         ht_enable_msi_mapping);
 
+/* The P5N32-SLI Premium motherboard from Asus has a problem with msi
+ * for the MCP55 NIC. It is not yet determined whether the msi problem
+ * also affects other devices. As for now, turn off msi for this device.
+ */
+static void __devinit nvenet_msi_disable(struct pci_dev *dev)
+{
+       if (dmi_name_in_vendors("P5N32-SLI PREMIUM")) {
+               dev_info(&dev->dev,
+                        "Disabling msi for MCP55 NIC on P5N32-SLI 
Premium\n");
+               dev->no_msi = 1;
+       }
+}
+DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_NVIDIA,
+                       PCI_DEVICE_ID_NVIDIA_NVENET_15,
+                       nvenet_msi_disable);
+
 static void __devinit nv_msi_ht_cap_quirk(struct pci_dev *dev)
 {
        struct pci_dev *host_bridge;
--

[PATCH] pci: Added quirk to disable msi for MCP55 NIC on Asus P5N32-SLI Premium

[Andreas Petlund]

Andreas Petlund wrote:
> I have a machine with an Asus motherboard (P5N32-SLI Premium - NVidia
> MCP55 chipset) where the forcedeth module does not work with msi. This
> seems to bother quite some people ( see
> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/136836 ).
> 
> Reading various forum posts makes me inclined to believe this is a
> firmware problem, but Asus have no fix with their latest BIOS-upgrade
> (and I am starting to believe that we will never see one as the last
> update came in March this year).
> 
> Is it an option to workaround this problem using a hardware-specific
> quirk that disables msi/msix for the device, or will that affect a lot
> of users that have no issues with this? On my system the NIC identifies
> as "cb84". I don't know if this is a common factor for all who has this
> problem, or only for my motherboard.
> Maybe the benefit of making it work with all MCP55-NICs is worth
> disabling msi/msix also for unaffected ones?
> 
> Does anybody have alternative suggestions?
> 

This patch detects the known affected motherboard and disables msi for the MCP55 NIC devices, allowing normal network operation without manually having to reload forcedeth module with msi=0 and msix=0.


Signed-off-by: Andreas Petlund <andreas@petlund.no>
---
 drivers/pci/quirks.c |   16 ++++++++++++++++
 1 files changed, 16 insertions(+), 0 deletions(-)

diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
index 5049a47..1bbd912 100644
--- a/drivers/pci/quirks.c
+++ b/drivers/pci/quirks.c
@@ -1828,6 +1828,22 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_SERVERWORKS,
 			 PCI_DEVICE_ID_SERVERWORKS_HT1000_PXB,
 			 ht_enable_msi_mapping);
 
+/* The P5N32-SLI Premium motherboard from Asus has a problem with msi
+ * for the MCP55 NIC. It is not yet determined whether the msi problem
+ * also affects other devices. As for now, turn off msi for this device.
+ */
+static void __devinit nvenet_msi_disable(struct pci_dev *dev)
+{
+	if (dmi_name_in_vendors("P5N32-SLI PREMIUM")) {
+		dev_info(&dev->dev,
+			 "Disabling msi for MCP55 NIC on P5N32-SLI Premium\n");
+		dev->no_msi = 1;
+	}
+}
+DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_NVIDIA,
+			PCI_DEVICE_ID_NVIDIA_NVENET_15,
+			nvenet_msi_disable);
+
 static void __devinit nv_msi_ht_cap_quirk(struct pci_dev *dev)
 {
 	struct pci_dev *host_bridge;
-- 
1.5.6.3