Reboot Hangs When R8169 Loaded, Intel D945GCLF Atom Board With Anything in PCI Slot, 2.6.28-6 Vanilla Kernel & Debian Etch

(This is a newbie reporting the problem, so take it with a grain of salt, to be sure.)

Reboot hangs when R8169 loaded, Intel D945GCLF Atom MB, 2.6.28-6 vanilla kernel on Debian etch if any card is in PCI slot.

I do know truly that the r8169 module is the culprit, but since r8169 has fingerprints on the problem (see below), and I have had several issues with the D945GCLF before for me with this driver, I figured might as well start there, and report it as such. If I am filing this in the wrong place, please direct me to a better one.

Tested 2.6.28-5 and 2.6.28-6 on the D945GCLF. With no devices in the PCI slot, and r8169 enabled, unit will shutdown -h and showdown -r just fine, and r8169 working in practice.

However, when any PCI device, and I tried video capture cards, USB port expander cards, unrelated network cards using different chip sets, WiFi PCI cards, any card I tried, shutdown -h continued to work fine, but any attempt at reboot with either reboot command or shutdown -r resulted in a system that did not reboot, just displayed the message as follows:

<snip>
Will now restart
r8169 0000:00:00.0: PME# enable
Restarting system
machine restart

And then it just sits there.

I did not have this issue with this Motherboard D945GCLF with 2.6.26.

Here is what makes me think the r8169 might be involved:

* When I ifdown eth0, reboot still does not work;
* When I rmmod r8169 just prior to reboot, then reboot works fine (and there is then no r8169 in the Restart message) ;
* When I disable the r8169 in BIOS, the reboot function works fine again (obviously not the Ethernet port then)

I include an excerpted lspic -v and dmesg below, for reference. I can live with the rmmod r8169 just prior to reboot as a work-around, but it seems like there must be a cleaner fix somewhere.

Cheers,

- Michael

lspic -v:

<snip>
01:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8101E PCI Express Fast Ethernet controller (rev 02)
Subsystem: Intel Corporation Unknown device 0001
Flags: bus master, fast devsel, latency 0, IRQ 508
I/O ports at 1000 [size=256]
Memory at 30200000 (64-bit, non-prefetchable) [size=4K]
Memory at 30000000 (64-bit, prefetchable) [size=64K]
Expansion ROM at 30020000 [disabled] [size=128K]
Capabilities: [40] Power Management version 3
Capabilities: [50] Message Signalled Interrupts: Mask- 64bit+
Queue=0/0 Enable+
Capabilities: [70] Express Endpoint IRQ 1
Capabilities: [ac] MSI-X: Enable- Mask- TabSize=2
Capabilities: [cc] Vital Product Data
<snip>

dmesg:
<snip>
r8169 Gigabit Ethernet driver 2.3LK-NAPI loaded r8169 0000:01:00.0: found PCI INT A -> IRQ 11 r8169 0000:01:00.0: sharing IRQ 11 with 0000:00:02.0 r8169 0000:01:00.0: sharing IRQ 11 with 0000:00:1d.3 r8169 0000:01:00.0: setting latency timer to 64 r8169 0000:01:00.0: irq 508 for MSI/MSI-X <snip>

Michael Grollman <[email protected]> :
[...]

Your report is fine.

Can you send a complete 'dmesg' from boot before the problem happens then try the patch below ?

index 0771eb6..fe5b99c 100644
@@ -3847,11 +3847,6 @@ out:
 	return 0;
 }
 
-static void rtl_shutdown(struct pci_dev *pdev)
-{
-	rtl8169_suspend(pdev, PMSG_SUSPEND);
-}
-
 #endif /* CONFIG_PM */
 
 static struct pci_driver rtl8169_pci_driver = {

@@ -3862,7 +3857,6 @@ static struct pci_driver rtl8169_pci_driver = {
 #ifdef CONFIG_PM
 	.suspend	= rtl8169_suspend,
 	.resume		= rtl8169_resume,
-	.shutdown	= rtl_shutdown,
 #endif
 };
 

Ueimor