On Sunday 14 June 2009, Rafael J. Wysocki wrote:

On Saturday 13 June 2009, Andreas Mohr wrote:

quoted

Hi all,

after having added non-MII PHY card support to e100, I noticed that
the suspend handler rejects power-management non-capable PCI cards,

Well, that means we have a bug somewhere in the PCI PM core.

quoted

causing a S2R request to immediately get back up to the desktop,
losing network access in the process (rtnl mutex deadlock).

That's bad.

quoted

ChangeLog:
Support PCI cards which are lacking power management capability
in the e100 suspend handler.


Frankly I was unsure how to best add this to the driver in a clean way.
Usually drivers use pci_set_power_state(..., pci_choose_state(...))
in order to avoid the rejection of an open-coded
pci_set_power_state(..., PCI_D3hot) in case of a non-PM card,
however pci_choose_state() depends on the _pm-internal_ pm_message_t type,
which was doable in .suspend directly but not at the other e100
driver locations where it was used.

Next attempt was to extend __e100_power_off() with a pci_power_t parameter,
however since __e100_power_off() is called by two locations,
that meant that I'd have to use pci_choose_state() at _both_ call sites.

Thus I simply resorted to do a brute-force yet most simple
pci_find_capability() check in the __e100_power_off() function.


Tested on 2.6.30-rc8 and suspending/resuming fine, checkpatch.pl:ed.
Patch against 2.6.30-rc8 with my original non-MII support patch applied.
(should apply fine in any case, I'd think).
Intended for testing in -mmotm or so.

Thanks!

Signed-off-by: Andreas Mohr <redacted>

--- linux-2.6.30-rc8.e100/drivers/net/e100.c.my_patch_orig	2009-06-13 18:47:53.000000000 +0200
+++ linux-2.6.30-rc8.e100/drivers/net/e100.c	2009-06-13 20:27:46.000000000 +0200

@@ -2897,6 +2897,13 @@ static void __e100_shutdown(struct pci_d
 
 static int __e100_power_off(struct pci_dev *pdev, bool wake)
 {
+	/* some older devices don't support PCI PM
+	 * (e.g. mac_82557_D100_B combo card with 80c24 PHY)
+	 * - skip those! (they most likely won't support WoL either)
+	 */
+	if (!pci_find_capability(pdev, PCI_CAP_ID_PM))
+		return 0;

Devices without PCI_CAP_ID_PM may still be power-manageable by ACPI, so
returning 0 at this point is not a general solution.

quoted

+
 	if (wake) {
 		return pci_prepare_to_sleep(pdev);

pci_prepare_to_sleep() is supposed to return 0 for your device.  I'll have a
look at it.

Please check if the appended patch helps.

Best,
Rafael

---
 drivers/pci/pci.c |    5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

Index: linux-2.6/drivers/pci/pci.c
===================================================================

--- linux-2.6.orig/drivers/pci/pci.c
+++ linux-2.6/drivers/pci/pci.c

@@ -1284,15 +1284,14 @@ pci_power_t pci_target_state(struct pci_
 		default:
 			target_state = state;
 		}
+	} else if (!dev->pm_cap) {
+		target_state = PCI_D0;
 	} else if (device_may_wakeup(&dev->dev)) {
 		/*
 		 * Find the deepest state from which the device can generate
 		 * wake-up events, make it the target state and enable device
 		 * to generate PME#.
 		 */
-		if (!dev->pm_cap)
-			return PCI_POWER_ERROR;
-
 		if (dev->pme_support) {
 			while (target_state
 			      && !(dev->pme_support & (1 << target_state)))

------------------------------------------------------------------------------

Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables unlimited
royalty-free distribution of the report engine for externally facing 
server and web deployment.
http://p.sf.net/sfu/businessobjects