Skip to contentSkip to conversation

ip: rework the fix for dflt addr selection for connected nexthop"

PATCH Applied Nicolas Dichtel 3 participants 6 messages Oct 20, 2022 – Oct 27, 2022

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

cover ip: rework the fix for dflt addr selection for connected nexthop"

This series reworks the fix that is reverted in the second commit. As Julian explained, nhc_scope is related to nhc_gw, it's not the scope of the route.

 net/ipv4/fib_frontend.c  | 4 ++--
 net/ipv4/fib_semantics.c | 2 +-
 net/ipv4/nexthop.c       | 2 +-
 3 files changed, 4 insertions(+), 4 deletions(-)

Comments are welcome.

Regards,
Nicolas

#1 lore raw

PATCH 1/3 Revert "ip: fix triggering of 'icmp redirect'"

This reverts commit eb55dc09b5dd040232d5de32812cc83001a23da6.

The patch that introduces this bug is reverted right after this one.

CC Julian Anastasov <[email protected]>
---
 net/ipv4/fib_frontend.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
net/ipv4/fib_frontend.c +2 -2
index 943edf4ad4db..f361d3d56be2 100644
@@ -389,7 +389,7 @@ static int __fib_validate_source(struct sk_buff *skb, __be32 src, __be32 dst,
 	dev_match = dev_match || (res.type == RTN_LOCAL &&
 				  dev == net->loopback_dev);
 	if (dev_match) {
-		ret = FIB_RES_NHC(res)->nhc_scope >= RT_SCOPE_LINK;
+		ret = FIB_RES_NHC(res)->nhc_scope >= RT_SCOPE_HOST;
 		return ret;
 	}
 	if (no_addr)
@@ -401,7 +401,7 @@ static int __fib_validate_source(struct sk_buff *skb, __be32 src, __be32 dst,
 	ret = 0;
 	if (fib_lookup(net, &fl4, &res, FIB_LOOKUP_IGNORE_LINKSTATE) == 0) {
 		if (res.type == RTN_UNICAST)
-			ret = FIB_RES_NHC(res)->nhc_scope >= RT_SCOPE_LINK;
+			ret = FIB_RES_NHC(res)->nhc_scope >= RT_SCOPE_HOST;
 	}
 	return ret;
 
Signature · 1 line
2.37.3
#2 lore raw

PATCH 2/3 Revert "ip: fix dflt addr selection for connected nexthop"

In reply to Nicolas Dichtel

This reverts commit 747c14307214b55dbd8250e1ab44cad8305756f1.

As explained by Julian, nhc_scope is related to nhc_gw, not to the route. Revert the original patch. The initial problem is fixed differently in the next commit.

---
 net/ipv4/fib_semantics.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/ipv4/fib_semantics.c +1 -1
index e9a7f70a54df..f721c308248b 100644
@@ -1231,7 +1231,7 @@ static int fib_check_nh_nongw(struct net *net, struct fib_nh *nh,
 
 	nh->fib_nh_dev = in_dev->dev;
 	netdev_hold(nh->fib_nh_dev, &nh->fib_nh_dev_tracker, GFP_ATOMIC);
-	nh->fib_nh_scope = RT_SCOPE_LINK;
+	nh->fib_nh_scope = RT_SCOPE_HOST;
 	if (!netif_carrier_ok(nh->fib_nh_dev))
 		nh->fib_nh_flags |= RTNH_F_LINKDOWN;
 	err = 0;
Signature · 1 line
2.37.3
#3 lore raw

PATCH 3/3 nh: fix scope used to find saddr when adding non gw nh

In reply to Nicolas Dichtel

As explained by Julian, fib_nh_scope is related to fib_nh_gw4, but fib_info_update_nhc_saddr() needs the scope of the route, which is the scope "before" fib_nh_scope, ie fib_nh_scope - 1.

This patch fixes the problem described in commit 747c14307214 ("ip: fix dflt addr selection for connected nexthop").

Fixes 597cfe4fc339 ("nexthop: Add support for IPv4 nexthops")
CC Julian Anastasov <[email protected]>
---
 net/ipv4/nexthop.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/ipv4/nexthop.c +1 -1
index 853a75a8fbaf..d8ef05347fd9 100644
@@ -2534,7 +2534,7 @@ static int nh_create_ipv4(struct net *net, struct nexthop *nh,
 	if (!err) {
 		nh->nh_flags = fib_nh->fib_nh_flags;
 		fib_info_update_nhc_saddr(net, &fib_nh->nh_common,
-					  fib_nh->fib_nh_scope);
+					  !fib_nh->fib_nh_scope ? 0 : fib_nh->fib_nh_scope - 1);
 	} else {
 		fib_nh_release(net, fib_nh);
 	}
Signature · 1 line
2.37.3
Reviewed-by on the v1 series
#4 lore raw
In reply to Nicolas Dichtel

Hello,

Quoted 10 lines · Nicolas Dichtel

This series reworks the fix that is reverted in the second commit. As Julian explained, nhc_scope is related to nhc_gw, it's not the scope of the route.

 net/ipv4/fib_frontend.c  | 4 ++--
 net/ipv4/fib_semantics.c | 2 +-
 net/ipv4/nexthop.c       | 2 +-
 3 files changed, 4 insertions(+), 4 deletions(-)

Comments are welcome.

Patchset looks good to me, thanks!

Regards

--
Julian Anastasov <[email protected]>

#5 lore raw
In reply to Nicolas Dichtel
This series was applied to netdev/net.git (master) by Jakub Kicinski <[email protected]>:

Hello:

This series was applied to netdev/net.git (master) by Jakub Kicinski <[email protected]>:

Quoted 10 lines · On Thu, 20 Oct 2022 12:09:49 +0200 you wrote:

This series reworks the fix that is reverted in the second commit. As Julian explained, nhc_scope is related to nhc_gw, it's not the scope of the route.

 net/ipv4/fib_frontend.c  | 4 ++--
 net/ipv4/fib_semantics.c | 2 +-
 net/ipv4/nexthop.c       | 2 +-
 3 files changed, 4 insertions(+), 4 deletions(-)

[...]

Here is the summary with links:
- [net,1/3] Revert "ip: fix triggering of 'icmp redirect'"

    https://git.kernel.org/netdev/net/c/745b913a5994
  - [net,2/3] Revert "ip: fix dflt addr selection for connected nexthop"
    https://git.kernel.org/netdev/net/c/e021c329ee19
  - [net,3/3] nh: fix scope used to find saddr when adding non gw nh
    https://git.kernel.org/netdev/net/c/bac0f937c343

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html