2023-11-17 15:28:09

by Kunwu Chan

[permalink] [raw]
Subject: [PATCH] ipv4: Correct/silence an endian warning in __ip_do_redirect

net/ipv4/route.c:783:46: warning: incorrect type in argument 2 (different base types)
net/ipv4/route.c:783:46: expected unsigned int [usertype] key
net/ipv4/route.c:783:46: got restricted __be32 [usertype] new_gw

Signed-off-by: Kunwu Chan <[email protected]>
---
net/ipv4/route.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 3290a4442b4a..e8a542c6b031 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -780,7 +780,7 @@ static void __ip_do_redirect(struct rtable *rt, struct sk_buff *skb, struct flow
goto reject_redirect;
}

- n = __ipv4_neigh_lookup(rt->dst.dev, new_gw);
+ n = __ipv4_neigh_lookup(rt->dst.dev, be32_to_cpu(new_gw));
if (!n)
n = neigh_create(&arp_tbl, &new_gw, rt->dst.dev);
if (!IS_ERR(n)) {
--
2.34.1


2023-11-17 17:16:01

by Eric Dumazet

[permalink] [raw]
Subject: Re: [PATCH] ipv4: Correct/silence an endian warning in __ip_do_redirect

On Fri, Nov 17, 2023 at 6:07 PM Kunwu Chan <[email protected]> wrote:
>
> net/ipv4/route.c:783:46: warning: incorrect type in argument 2 (different base types)
> net/ipv4/route.c:783:46: expected unsigned int [usertype] key
> net/ipv4/route.c:783:46: got restricted __be32 [usertype] new_gw
>
> Signed-off-by: Kunwu Chan <[email protected]>

We need Fixes: tag for networking patches.

> ---
> net/ipv4/route.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/net/ipv4/route.c b/net/ipv4/route.c
> index 3290a4442b4a..e8a542c6b031 100644
> --- a/net/ipv4/route.c
> +++ b/net/ipv4/route.c
> @@ -780,7 +780,7 @@ static void __ip_do_redirect(struct rtable *rt, struct sk_buff *skb, struct flow
> goto reject_redirect;
> }
>
> - n = __ipv4_neigh_lookup(rt->dst.dev, new_gw);
> + n = __ipv4_neigh_lookup(rt->dst.dev, be32_to_cpu(new_gw));
> if (!n)
> n = neigh_create(&arp_tbl, &new_gw, rt->dst.dev);
> if (!IS_ERR(n)) {
> --
> 2.34.1
>

How was this patch tested ?

You are 'fixing' sparse warnings by replacing them with real bugs.

be32_to_cpu() is going to swap bytes on x86, so the lookup will fail horribly.

Here, if you must silence sparse, you want (__force u32)new_gw

Look at this commit for a template.

commit 3c42b2019863b327caa233072c50739d4144dd16

2023-11-19 14:08:33

by Kunwu Chan

[permalink] [raw]
Subject: Re: [PATCH] ipv4: Correct/silence an endian warning in __ip_do_redirect

Hi Eric,
Thank you so much for taking the time to guide me, I'm a rookie who
really wants to do my part for the kernel, and I can't get started, so I
thought about eliminating some of the sparse warnings. I've looked at
some other commits and thought I could resolve the alert this way, sorry
for the trouble.
Follow your suggestion:
1. I will add a 'Fixes' tag as follows:
'Fixes: 969447f226b4 ("ipv4: use new_gw for redirect neigh lookup")'

2. Refer to the modification method of commit
3c42b2019863b327caa233072c50739d4144dd16, and modify the patch to:
'n = __ipv4_neigh_lookup(rt->dst.dev, (__force u32)new_gw); '

On 2023/11/18 01:15, Eric Dumazet wrote:
> On Fri, Nov 17, 2023 at 6:07 PM Kunwu Chan <[email protected]> wrote:
>>
>> net/ipv4/route.c:783:46: warning: incorrect type in argument 2 (different base types)
>> net/ipv4/route.c:783:46: expected unsigned int [usertype] key
>> net/ipv4/route.c:783:46: got restricted __be32 [usertype] new_gw
>>
>> Signed-off-by: Kunwu Chan <[email protected]>
>
> We need Fixes: tag for networking patches.
>
>> ---
>> net/ipv4/route.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/net/ipv4/route.c b/net/ipv4/route.c
>> index 3290a4442b4a..e8a542c6b031 100644
>> --- a/net/ipv4/route.c
>> +++ b/net/ipv4/route.c
>> @@ -780,7 +780,7 @@ static void __ip_do_redirect(struct rtable *rt, struct sk_buff *skb, struct flow
>> goto reject_redirect;
>> }
>>
>> - n = __ipv4_neigh_lookup(rt->dst.dev, new_gw);
>> + n = __ipv4_neigh_lookup(rt->dst.dev, be32_to_cpu(new_gw));
>> if (!n)
>> n = neigh_create(&arp_tbl, &new_gw, rt->dst.dev);
>> if (!IS_ERR(n)) {
>> --
>> 2.34.1
>>
>
> How was this patch tested ?
>
> You are 'fixing' sparse warnings by replacing them with real bugs.
>
> be32_to_cpu() is going to swap bytes on x86, so the lookup will fail horribly.
>
> Here, if you must silence sparse, you want (__force u32)new_gw
>
> Look at this commit for a template.
>
> commit 3c42b2019863b327caa233072c50739d4144dd16

2023-11-19 14:18:27

by Kunwu Chan

[permalink] [raw]
Subject: [PATCH v2] ipv4: Correct/silence an endian warning in __ip_do_redirect

net/ipv4/route.c:783:46: warning: incorrect type in argument 2 (different base types)
net/ipv4/route.c:783:46: expected unsigned int [usertype] key
net/ipv4/route.c:783:46: got restricted __be32 [usertype] new_gw

Fixes: 969447f226b4 ("ipv4: use new_gw for redirect neigh lookup")
Suggested-by: Eric Dumazet <[email protected]>
Signed-off-by: Kunwu Chan <[email protected]>
---
net/ipv4/route.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 3290a4442b4a..16615d107cf0 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -780,7 +780,7 @@ static void __ip_do_redirect(struct rtable *rt, struct sk_buff *skb, struct flow
goto reject_redirect;
}

- n = __ipv4_neigh_lookup(rt->dst.dev, new_gw);
+ n = __ipv4_neigh_lookup(rt->dst.dev, (__force u32)new_gw);
if (!n)
n = neigh_create(&arp_tbl, &new_gw, rt->dst.dev);
if (!IS_ERR(n)) {
--
2.34.1

2023-11-21 12:11:25

by patchwork-bot+netdevbpf

[permalink] [raw]
Subject: Re: [PATCH v2] ipv4: Correct/silence an endian warning in __ip_do_redirect

Hello:

This patch was applied to netdev/net.git (main)
by Paolo Abeni <[email protected]>:

On Sun, 19 Nov 2023 22:17:59 +0800 you wrote:
> net/ipv4/route.c:783:46: warning: incorrect type in argument 2 (different base types)
> net/ipv4/route.c:783:46: expected unsigned int [usertype] key
> net/ipv4/route.c:783:46: got restricted __be32 [usertype] new_gw
>
> Fixes: 969447f226b4 ("ipv4: use new_gw for redirect neigh lookup")
> Suggested-by: Eric Dumazet <[email protected]>
> Signed-off-by: Kunwu Chan <[email protected]>
>
> [...]

Here is the summary with links:
- [v2] ipv4: Correct/silence an endian warning in __ip_do_redirect
https://git.kernel.org/netdev/net/c/c0e2926266af

You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html