[PATCH] net: bcmgenet: Don't claim WOL when its not available

Some of the bcmgenet platforms don't correctly support WOL, yet
ethtool returns:

"Supports Wake-on: gsf"

which is false.

Ideally if there isn't a wol_irq, or there is something else that
keeps the device from being able to wakeup it should display:

"Supports Wake-on: d"

This patch checks whether the device can wakup, before using the
hard-coded supported flags. This corrects the ethtool reporting, as
well as the WOL configuration because ethtool verifies that the mode
is supported before attempting it.

Signed-off-by: Jeremy Linton <redacted>
---
 drivers/net/ethernet/broadcom/genet/bcmgenet_wol.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet_wol.c b/drivers/net/ethernet/broadcom/genet/bcmgenet_wol.c
index e31a5a397f11..f55d9d9c01a8 100644
--- a/drivers/net/ethernet/broadcom/genet/bcmgenet_wol.c
+++ b/drivers/net/ethernet/broadcom/genet/bcmgenet_wol.c

@@ -40,6 +40,13 @@
 void bcmgenet_get_wol(struct net_device *dev, struct ethtool_wolinfo *wol)
 {
 	struct bcmgenet_priv *priv = netdev_priv(dev);
+	struct device *kdev = &priv->pdev->dev;
+
+	if (!device_can_wakeup(kdev)) {
+		wol->supported = 0;
+		wol->wolopts = 0;
+		return;
+	}
 
 	wol->supported = WAKE_MAGIC | WAKE_MAGICSECURE | WAKE_FILTER;
 	wol->wolopts = priv->wolopts;

-- 
2.35.1

On Thu, Mar 10, 2022 at 4:55 AM Jeremy Linton [off-list ref] wrote:

Some of the bcmgenet platforms don't correctly support WOL, yet
ethtool returns:

"Supports Wake-on: gsf"

which is false.

Ideally if there isn't a wol_irq, or there is something else that
keeps the device from being able to wakeup it should display:

"Supports Wake-on: d"

This patch checks whether the device can wakup, before using the
hard-coded supported flags. This corrects the ethtool reporting, as
well as the WOL configuration because ethtool verifies that the mode
is supported before attempting it.

Signed-off-by: Jeremy Linton <redacted>

Tested-by: Peter Robinson <redacted>

This fixes the reporting of the WOL capabilities on the Raspberry Pi 4.

On 3/9/22 8:55 PM, Jeremy Linton wrote:

Some of the bcmgenet platforms don't correctly support WOL, yet
ethtool returns:

"Supports Wake-on: gsf"

which is false.

Ideally if there isn't a wol_irq, or there is something else that
keeps the device from being able to wakeup it should display:

"Supports Wake-on: d"

This patch checks whether the device can wakup, before using the
hard-coded supported flags. This corrects the ethtool reporting, as
well as the WOL configuration because ethtool verifies that the mode
is supported before attempting it.

Signed-off-by: Jeremy Linton <redacted>

Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Fixes: c51de7f3976b ("net: bcmgenet: add Wake-on-LAN support code")

Thanks Jeremy!
-- 
Florian

Hello:

This patch was applied to netdev/net.git (master)
by Jakub Kicinski [off-list ref]:

On Wed,  9 Mar 2022 22:55:35 -0600 you wrote:

Some of the bcmgenet platforms don't correctly support WOL, yet
ethtool returns:

"Supports Wake-on: gsf"

which is false.

[...]

Here is the summary with links:
  - net: bcmgenet: Don't claim WOL when its not available
    https://git.kernel.org/netdev/net/c/00b022f8f876

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html