According to Guillaume Nault RT_TOS should never be used for IPv6.

Fixes: 3a56f86f1be6a ("geneve: handle ipv6 priority like ipv4 tos")
Signed-off-by: Matthias May <matthias.may@westermo.com>
---
v1 -> v2:
 - Fix spacing of "Fixes" tag.
 - Add missing CCs
---
 drivers/net/geneve.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c
index 4c380c06f178..e1a4480e6f17 100644
--- a/drivers/net/geneve.c
+++ b/drivers/net/geneve.c
@@ -877,8 +877,7 @@ static struct dst_entry *geneve_get_v6_dst(struct sk_buff *skb,
 		use_cache = false;
 	}
 
-	fl6->flowlabel = ip6_make_flowinfo(RT_TOS(prio),
-					   info->key.label);
+	fl6->flowlabel = ip6_make_flowinfo(prio, info->key.label);
 	dst_cache = (struct dst_cache *)&info->dst_cache;
 	if (use_cache) {
 		dst = dst_cache_get_ip6(dst_cache, &fl6->saddr);
-- 
2.35.1

On 8/2/22 6:09 AM, Matthias May wrote:
> According to Guillaume Nault RT_TOS should never be used for IPv6.
> 
> Fixes: 3a56f86f1be6a ("geneve: handle ipv6 priority like ipv4 tos")
> Signed-off-by: Matthias May <matthias.may@westermo.com>
> ---
> v1 -> v2:
>  - Fix spacing of "Fixes" tag.
>  - Add missing CCs
> ---
>  drivers/net/geneve.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 

Reviewed-by: David Ahern <dsahern@kernel.org>

On Tue, Aug 02, 2022 at 02:09:32PM +0200, Matthias May wrote:
> According to Guillaume Nault RT_TOS should never be used for IPv6.

While I obviously agree with this statement, I don't think it brings
much information to other people reading the commit message. Even
though the cover letter has more information, it won't be immediately
available to people reading the patch. Also the cover letter will be
lost if the patch gets picked up for the -stable trees.

Appart from that, I'm fine with the patch of course.

> Fixes: 3a56f86f1be6a ("geneve: handle ipv6 priority like ipv4 tos")
> Signed-off-by: Matthias May <matthias.may@westermo.com>
> ---
> v1 -> v2:
>  - Fix spacing of "Fixes" tag.
>  - Add missing CCs
> ---
>  drivers/net/geneve.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c
> index 4c380c06f178..e1a4480e6f17 100644
> --- a/drivers/net/geneve.c
> +++ b/drivers/net/geneve.c
> @@ -877,8 +877,7 @@ static struct dst_entry *geneve_get_v6_dst(struct sk_buff *skb,
>  		use_cache = false;
>  	}
>  
> -	fl6->flowlabel = ip6_make_flowinfo(RT_TOS(prio),
> -					   info->key.label);
> +	fl6->flowlabel = ip6_make_flowinfo(prio, info->key.label);
>  	dst_cache = (struct dst_cache *)&info->dst_cache;
>  	if (use_cache) {
>  		dst = dst_cache_get_ip6(dst_cache, &fl6->saddr);
> -- 
> 2.35.1
>

On Tue, Aug 02, 2022 at 02:09:32PM +0200, Matthias May wrote:
> According to Guillaume Nault RT_TOS should never be used for IPv6.
> 
> Fixes: 3a56f86f1be6a ("geneve: handle ipv6 priority like ipv4 tos")

While I'm at it, the SHA1 is normally truncated to 12 charaters, not 13.

> Signed-off-by: Matthias May <matthias.may@westermo.com>
> ---
> v1 -> v2:
>  - Fix spacing of "Fixes" tag.
>  - Add missing CCs
> ---
>  drivers/net/geneve.c | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c
> index 4c380c06f178..e1a4480e6f17 100644
> --- a/drivers/net/geneve.c
> +++ b/drivers/net/geneve.c
> @@ -877,8 +877,7 @@ static struct dst_entry *geneve_get_v6_dst(struct sk_buff *skb,
>  		use_cache = false;
>  	}
>  
> -	fl6->flowlabel = ip6_make_flowinfo(RT_TOS(prio),
> -					   info->key.label);
> +	fl6->flowlabel = ip6_make_flowinfo(prio, info->key.label);
>  	dst_cache = (struct dst_cache *)&info->dst_cache;
>  	if (use_cache) {
>  		dst = dst_cache_get_ip6(dst_cache, &fl6->saddr);
> -- 
> 2.35.1
>

On 03/08/2022 13:30, Guillaume Nault wrote:
> On Tue, Aug 02, 2022 at 02:09:32PM +0200, Matthias May wrote:
>> According to Guillaume Nault RT_TOS should never be used for IPv6.
>>
>> Fixes: 3a56f86f1be6a ("geneve: handle ipv6 priority like ipv4 tos")
> 
> While I'm at it, the SHA1 is normally truncated to 12 charaters, not 13.
> 

Hi Guillaume
Ups, missed to remove the additional character in this patch.
For some reason git blame shows the SHA1 with 13 characters, not sure why.

BR
Matthias