2010-06-17 14:17:31

by Fred Isaman

Subject: [PATCH 0/1] fix delay on mount and umount

This fixes the delay issues we've been having.

2010-06-17 14:42:28

by Timo Aaltonen

Subject: Re: [PATCH 0/1] fix delay on mount and umount

On Wed, 16 Jun 2010, Fred Isaman wrote:

> This fixes the delay issues we've been having.

What kind of delays are we talking about and since which kernel version?
With 2.6.32 it takes several minutes to umount a 100 NFSv4 shares.
Mounting them is much faster but not instant.

Probably will end up using the automounter anyway, but it's nice to know
if this fixes the issue.

--
Timo Aaltonen
Systems Specialist
Aalto University IT

2010-06-17 14:17:31

by Fred Isaman

Subject: [PATCH 1/1] net: fix deliver_no_wcard regression on loopback device

From: John Fastabend <[email protected]>

deliver_no_wcard is not being set in skb_copy_header.
In the skb_cloned case it is not being cleared and
may cause the skb to be dropped when the loopback device
pushes it back up the stack.

Signed-off-by: John Fastabend <[email protected]>
---
net/core/skbuff.c | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 9f07e74..bcf2fa3 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -532,6 +532,7 @@ static void __copy_skb_header(struct sk_buff *new, const struct sk_buff *old)
new->ip_summed = old->ip_summed;
skb_copy_queue_mapping(new, old);
new->priority = old->priority;
+ new->deliver_no_wcard = old->deliver_no_wcard;
#if defined(CONFIG_IP_VS) || defined(CONFIG_IP_VS_MODULE)
new->ipvs_property = old->ipvs_property;
#endif
--
1.6.6.1

2010-06-17 14:51:50

by Fred Isaman

Subject: Re: [PATCH 0/1] fix delay on mount and umount

On Thu, Jun 17, 2010 at 10:41 AM, Timo Aaltonen <[email protected]> wrote:
> On Wed, 16 Jun 2010, Fred Isaman wrote:
>
>> This fixes the delay issues we've been having.
>
> What kind of delays are we talking about and since which kernel version?
> With 2.6.32 it takes several minutes to umount a 100 NFSv4 shares. Mounting
> them is much faster but not instant.
>
> Probably will end up using the automounter anyway, but it's nice to know if
> this fixes the issue.
>

Sorry, I should have been more specific. This fixes a regression that
occurred between 2.6.35-rc2 and 2.6.35-rc3.
Mounts would hang for approx 2 minutes.

Fred

> --
> Timo Aaltonen
> Systems Specialist
> Aalto University IT
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to [email protected]
> More majordomo info at ?http://vger.kernel.org/majordomo-info.html
>

2010-06-17 16:01:17

by Benny Halevy

Subject: Re: [PATCH 1/1] net: fix deliver_no_wcard regression on loopback device

Took in at pnfs-all-2.6.35-rc3-2010-06-17

Thanks!

Benny


On Jun. 16, 2010, 9:02 -0400, Fred Isaman <[email protected]> wrote:
> From: John Fastabend <[email protected]>
>
> deliver_no_wcard is not being set in skb_copy_header.
> In the skb_cloned case it is not being cleared and
> may cause the skb to be dropped when the loopback device
> pushes it back up the stack.
>
> Signed-off-by: John Fastabend <[email protected]>
> ---
> net/core/skbuff.c | 1 +
> 1 files changed, 1 insertions(+), 0 deletions(-)
>
> diff --git a/net/core/skbuff.c b/net/core/skbuff.c
> index 9f07e74..bcf2fa3 100644
> --- a/net/core/skbuff.c
> +++ b/net/core/skbuff.c
> @@ -532,6 +532,7 @@ static void __copy_skb_header(struct sk_buff *new, const struct sk_buff *old)
> new->ip_summed = old->ip_summed;
> skb_copy_queue_mapping(new, old);
> new->priority = old->priority;
> + new->deliver_no_wcard = old->deliver_no_wcard;
> #if defined(CONFIG_IP_VS) || defined(CONFIG_IP_VS_MODULE)
> new->ipvs_property = old->ipvs_property;
> #endif