[PATCH 3.19 and earlier] fib_rules: Fix dump_rules() not to exit early

From: Roland Dreier <redacted>

Backports of 41fc014332d9 ("fib_rules: fix fib rule dumps across
multiple skbs") introduced a regression in "ip rule show" - it ends up
dumping the first rule over and over and never exiting, because 3.19
and earlier are missing commit 053c095a82cf ("netlink: make
nlmsg_end() and genlmsg_end() void"), so fib_nl_fill_rule() ends up
returning skb->len (i.e. > 0) in the success case.

Fix this by checking the return code for < 0 instead of != 0.

Signed-off-by: Roland Dreier <redacted>
---
Hi, this is needed for all stable trees earlier than 4.0 that have
picked up 41fc014332d9; so far looks like at least 3.10.y and 3.14.y
have made such releases.

 net/core/fib_rules.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/core/fib_rules.c b/net/core/fib_rules.c
index 627e517077e4..84340a2605ed 100644
--- a/net/core/fib_rules.c
+++ b/net/core/fib_rules.c

@@ -606,7 +606,7 @@ static int dump_rules(struct sk_buff *skb, struct netlink_callback *cb,
 		err = fib_nl_fill_rule(skb, rule, NETLINK_CB(cb->skb).portid,
 				       cb->nlh->nlmsg_seq, RTM_NEWRULE,
 				       NLM_F_MULTI, ops);
-		if (err)
+		if (err < 0)
 			break;
 skip:
 		idx++;

-- 
2.5.0

On 10/05/2015, 07:29 PM, Roland Dreier wrote:

I integrated the fix into the backport of 41fc014332d9 in 3.12 and made
a note in there. Thanks!

-- 
js
suse labs

On Mon, 2015-10-05 at 10:29 -0700, Roland Dreier wrote:

From: Roland Dreier <redacted>

Backports of 41fc014332d9 ("fib_rules: fix fib rule dumps across
multiple skbs") introduced a regression in "ip rule show" - it ends up
dumping the first rule over and over and never exiting, because 3.19
and earlier are missing commit 053c095a82cf ("netlink: make
nlmsg_end() and genlmsg_end() void"), so fib_nl_fill_rule() ends up
returning skb->len (i.e. > 0) in the success case.

Fix this by checking the return code for < 0 instead of != 0.

Signed-off-by: Roland Dreier <redacted>
---
Hi, this is needed for all stable trees earlier than 4.0 that have
picked up 41fc014332d9; so far looks like at least 3.10.y and 3.14.y
have made such releases.

[...]

Thanks, I'm folding this into the backported patch in 3.2.

Ben.

-- 
Ben Hutchings
Once a job is fouled up, anything done to improve it makes it worse.

Hi Roland,

On Monday, 5. October 2015 10:29:28 Roland Dreier wrote:

From: Roland Dreier <redacted>

Backports of 41fc014332d9 ("fib_rules: fix fib rule dumps across
multiple skbs") introduced a regression in "ip rule show" - it ends up
dumping the first rule over and over and never exiting, because 3.19
and earlier are missing commit 053c095a82cf ("netlink: make
nlmsg_end() and genlmsg_end() void"), so fib_nl_fill_rule() ends up
returning skb->len (i.e. > 0) in the success case.

Fix this by checking the return code for < 0 instead of != 0.

thanks for this fix. You just saved me an afternoon of bisecting :)

I can confirm that this fixes the mentioned issue introduced in 3.14.54.
We have an automated ipsec VPN test that failed after the upgrade:
The "ip rule list" command was hanging forever.

Cheers,
Thomas

On Mon, Oct 05, 2015 at 10:29:28AM -0700, Roland Dreier wrote:

From: Roland Dreier <redacted>

Backports of 41fc014332d9 ("fib_rules: fix fib rule dumps across
multiple skbs") introduced a regression in "ip rule show" - it ends up
dumping the first rule over and over and never exiting, because 3.19
and earlier are missing commit 053c095a82cf ("netlink: make
nlmsg_end() and genlmsg_end() void"), so fib_nl_fill_rule() ends up
returning skb->len (i.e. > 0) in the success case.

Fix this by checking the return code for < 0 instead of != 0.

Signed-off-by: Roland Dreier <redacted>
---
Hi, this is needed for all stable trees earlier than 4.0 that have
picked up 41fc014332d9; so far looks like at least 3.10.y and 3.14.y
have made such releases.

Thanks for this, now applied to 3.10 and 3.14-stable trees.

greg k-h

On Mon, Oct 05, 2015 at 10:29:28AM -0700, Roland Dreier wrote:

From: Roland Dreier <redacted>

Backports of 41fc014332d9 ("fib_rules: fix fib rule dumps across
multiple skbs") introduced a regression in "ip rule show" - it ends up
dumping the first rule over and over and never exiting, because 3.19
and earlier are missing commit 053c095a82cf ("netlink: make
nlmsg_end() and genlmsg_end() void"), so fib_nl_fill_rule() ends up
returning skb->len (i.e. > 0) in the success case.

Fix this by checking the return code for < 0 instead of != 0.

Signed-off-by: Roland Dreier <redacted>
---
Hi, this is needed for all stable trees earlier than 4.0 that have
picked up 41fc014332d9; so far looks like at least 3.10.y and 3.14.y
have made such releases.

Thanks Roland, I'm queuing this fix for the 3.16 kernel as well.

Cheers,
--
Luís