2018-06-06 21:36:12

by Dexuan Cui

Subject: [PATCH v2] hv_netvsc: Fix a network regression after ifdown/ifup

Recently people reported the NIC stops working after
"ifdown eth0; ifup eth0". It turns out in this case the TX queues are not
enabled, after the refactoring of the common detach logic: when the NIC
has sub-channels, usually we enable all the TX queues after all
sub-channels are set up: see rndis_set_subchannel() ->
netif_device_attach(), but in the case of "ifdown eth0; ifup eth0" where
the number of channels doesn't change, we also must make sure the TX queues
are enabled. The patch fixes the regression.

Fixes: 7b2ee50c0cd5 ("hv_netvsc: common detach logic")
Signed-off-by: Dexuan Cui <[email protected]>
Cc: [email protected]
Cc: Stephen Hemminger <[email protected]>
Cc: K. Y. Srinivasan <[email protected]>
Cc: Haiyang Zhang <[email protected]>
---

Changes since v1:
Removed the check and the comment from the code [Stephen Hemminger]

drivers/net/hyperv/netvsc_drv.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/net/hyperv/netvsc_drv.c b/drivers/net/hyperv/netvsc_drv.c
index da07ccd..eb8dccd 100644
--- a/drivers/net/hyperv/netvsc_drv.c
+++ b/drivers/net/hyperv/netvsc_drv.c
@@ -126,8 +126,10 @@ static int netvsc_open(struct net_device *net)
}

rdev = nvdev->extension;
- if (!rdev->link_state)
+ if (!rdev->link_state) {
netif_carrier_on(net);
+ netif_tx_wake_all_queues(net);
+ }

if (vf_netdev) {
/* Setting synthetic device up transparently sets
--
2.7.4

2018-06-07 20:25:52

by David Miller

Subject: Re: [PATCH v2] hv_netvsc: Fix a network regression after ifdown/ifup

From: Dexuan Cui <[email protected]>
Date: Wed, 6 Jun 2018 21:32:51 +0000

>
> Recently people reported the NIC stops working after
> "ifdown eth0; ifup eth0". It turns out in this case the TX queues are not
> enabled, after the refactoring of the common detach logic: when the NIC
> has sub-channels, usually we enable all the TX queues after all
> sub-channels are set up: see rndis_set_subchannel() ->
> netif_device_attach(), but in the case of "ifdown eth0; ifup eth0" where
> the number of channels doesn't change, we also must make sure the TX queues
> are enabled. The patch fixes the regression.
>
> Fixes: 7b2ee50c0cd5 ("hv_netvsc: common detach logic")
> Signed-off-by: Dexuan Cui <[email protected]>

Applied and queued up for -stable.