[PATCH v3] net: Cavium: Fix MAC address setting in shutdown state

[PATCH v3] net: Cavium: Fix MAC address setting in shutdown state

[Pavel Fedin]

This bug pops up with NetworkManager on Fedora 21. NetworkManager tends to
stop the interface (nicvf_stop() is called) before changing settings. In
stopped state MAC cannot be sent to a PF. However, when the interface is
restarted (nicvf_open() is called), we ping the PF using NIC_MBOX_MSG_READY
message, and the PF replies back with old MAC address, overriding what we
had after MAC setting from userspace. As a result, we cannot set MAC
address using NetworkManager.

This patch introduces special tracking of MAC change in stopped state so
that the correct new MAC address is sent to a PF when interface is reopen.

Signed-off-by: Pavel Fedin <p.fedin@samsung.com>
---
Changes for v3:
- Fixed (actually hand-edited) file permissions to be 644. I occasionally pushed them as 755 to my
local repo.
Changes for v2:
- Fixed a couple of checkpatch.pl reports (was forgotten in v1)
- Adjusted email client prefs, hopefully will not corrupt patches any more.
---
 drivers/net/ethernet/cavium/thunder/nic.h        |  1 +
 drivers/net/ethernet/cavium/thunder/nicvf_main.c | 14 ++++++++++++--
 2 files changed, 13 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/cavium/thunder/nic.h b/drivers/net/ethernet/cavium/thunder/nic.h
index a3b43e5..dda8a02 100644
--- a/drivers/net/ethernet/cavium/thunder/nic.h
+++ b/drivers/net/ethernet/cavium/thunder/nic.h
@@ -263,6 +263,7 @@ struct nicvf {
 	bool			pf_acked;
 	bool			pf_nacked;
 	bool			bgx_stats_acked;
+	bool			set_mac_pending;
 } ____cacheline_aligned_in_smp;
 
 /* PF <--> VF Mailbox communication
diff --git a/drivers/net/ethernet/cavium/thunder/nicvf_main.c
b/drivers/net/ethernet/cavium/thunder/nicvf_main.c
index 02da802..8b119a0 100644
--- a/drivers/net/ethernet/cavium/thunder/nicvf_main.c
+++ b/drivers/net/ethernet/cavium/thunder/nicvf_main.c
@@ -201,7 +201,9 @@ static void  nicvf_handle_mbx_intr(struct nicvf *nic)
 		nic->vf_id = mbx.nic_cfg.vf_id & 0x7F;
 		nic->tns_mode = mbx.nic_cfg.tns_mode & 0x7F;
 		nic->node = mbx.nic_cfg.node_id;
-		ether_addr_copy(nic->netdev->dev_addr, mbx.nic_cfg.mac_addr);
+		if (!nic->set_mac_pending)
+			ether_addr_copy(nic->netdev->dev_addr,
+					mbx.nic_cfg.mac_addr);
 		nic->link_up = false;
 		nic->duplex = 0;
 		nic->speed = 0;
@@ -941,6 +943,11 @@ int nicvf_open(struct net_device *netdev)
 		nicvf_hw_set_mac_addr(nic, netdev);
 	}
 
+	if (nic->set_mac_pending) {
+		nic->set_mac_pending = false;
+		nicvf_hw_set_mac_addr(nic, netdev);
+	}
+
 	/* Init tasklet for handling Qset err interrupt */
 	tasklet_init(&nic->qs_err_task, nicvf_handle_qs_err,
 		     (unsigned long)nic);
@@ -1040,9 +1047,12 @@ static int nicvf_set_mac_address(struct net_device *netdev, void *p)
 
 	memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len);
 
-	if (nic->msix_enabled)
+	if (nic->msix_enabled) {
 		if (nicvf_hw_set_mac_addr(nic, netdev))
 			return -EBUSY;
+	} else {
+		nic->set_mac_pending = true;
+	}
 
 	return 0;
 }
-- 
1.9.5.msysgit.0


--
To unsubscribe from this list: send the line "unsubscribe netdev" in

Re: [PATCH v3] net: Cavium: Fix MAC address setting in shutdown state

[David Miller]

From: Pavel Fedin <p.fedin@samsung.com>
Date: Tue, 23 Jun 2015 17:51:06 +0300

> This bug pops up with NetworkManager on Fedora 21. NetworkManager tends to
> stop the interface (nicvf_stop() is called) before changing settings. In
> stopped state MAC cannot be sent to a PF. However, when the interface is
> restarted (nicvf_open() is called), we ping the PF using NIC_MBOX_MSG_READY
> message, and the PF replies back with old MAC address, overriding what we
> had after MAC setting from userspace. As a result, we cannot set MAC
> address using NetworkManager.
> 
> This patch introduces special tracking of MAC change in stopped state so
> that the correct new MAC address is sent to a PF when interface is reopen.
> 
> Signed-off-by: Pavel Fedin <p.fedin@samsung.com>

Applied, thanks.

> - Fixed (actually hand-edited) file permissions to be 644. I occasionally pushed them as 755 to my
> local repo.

You should be sending patches against my repo, or something derived from it,
which therefore has the correct file permissions for these files.