Hi all,
Today's linux-next merge of the ipsec-next tree got conflicts in:
net/ipv4/xfrm4_output.c
net/ipv6/xfrm6_output.c
between commit:
0c922a4850eb ("xfrm: Always set XFRM_TRANSFORMED in xfrm{4,6}_output_finish")
from Linus' tree and commit:
2ab6096db2f1 ("xfrm: remove output_finish indirection from xfrm_state_afinfo")
from the ipsec-next tree.
I fixed it up (I used the latter versions of these files and then added
the following patch) and can carry the fix as necessary. This is now fixed
as far as linux-next is concerned, but any non trivial conflicts should
be mentioned to your upstream maintainer when your tree is submitted for
merging. You may also want to consider cooperating with the maintainer
of the conflicting tree to minimise any particularly complex conflicts.
From: Stephen Rothwell <[email protected]>
Date: Mon, 11 May 2020 12:57:24 +1000
Subject: [PATCH] xfrm: merge fixup for "remove output_finish indirection from xfrm_state_afinfo"
Signed-off-by: Stephen Rothwell <[email protected]>
---
net/xfrm/xfrm_output.c | 4 ----
1 file changed, 4 deletions(-)
diff --git a/net/xfrm/xfrm_output.c b/net/xfrm/xfrm_output.c
index 886a9b284b3a..0f4b3a5e02ba 100644
--- a/net/xfrm/xfrm_output.c
+++ b/net/xfrm/xfrm_output.c
@@ -574,16 +574,12 @@ int xfrm_output(struct sock *sk, struct sk_buff *skb)
switch (x->outer_mode.family) {
case AF_INET:
memset(IPCB(skb), 0, sizeof(*IPCB(skb)));
-#ifdef CONFIG_NETFILTER
IPCB(skb)->flags |= IPSKB_XFRM_TRANSFORMED;
-#endif
break;
case AF_INET6:
memset(IP6CB(skb), 0, sizeof(*IP6CB(skb)));
-#ifdef CONFIG_NETFILTER
IP6CB(skb)->flags |= IP6SKB_XFRM_TRANSFORMED;
-#endif
break;
}
--
2.26.2
--
Cheers,
Stephen Rothwell
Hi Dave,
On Mon, 11 May 2020 13:00:15 +1000 Stephen Rothwell <[email protected]> wrote:
>
> Today's linux-next merge of the ipsec-next tree got conflicts in:
>
> net/ipv4/xfrm4_output.c
> net/ipv6/xfrm6_output.c
>
> between commit:
>
> 0c922a4850eb ("xfrm: Always set XFRM_TRANSFORMED in xfrm{4,6}_output_finish")
>
> from Linus' tree and commit:
>
> 2ab6096db2f1 ("xfrm: remove output_finish indirection from xfrm_state_afinfo")
>
> from the ipsec-next tree.
>
> I fixed it up (I used the latter versions of these files and then added
> the following patch) and can carry the fix as necessary. This is now fixed
> as far as linux-next is concerned, but any non trivial conflicts should
> be mentioned to your upstream maintainer when your tree is submitted for
> merging. You may also want to consider cooperating with the maintainer
> of the conflicting tree to minimise any particularly complex conflicts.
>
> From: Stephen Rothwell <[email protected]>
> Date: Mon, 11 May 2020 12:57:24 +1000
> Subject: [PATCH] xfrm: merge fixup for "remove output_finish indirection from xfrm_state_afinfo"
>
> Signed-off-by: Stephen Rothwell <[email protected]>
> ---
> net/xfrm/xfrm_output.c | 4 ----
> 1 file changed, 4 deletions(-)
>
> diff --git a/net/xfrm/xfrm_output.c b/net/xfrm/xfrm_output.c
> index 886a9b284b3a..0f4b3a5e02ba 100644
> --- a/net/xfrm/xfrm_output.c
> +++ b/net/xfrm/xfrm_output.c
> @@ -574,16 +574,12 @@ int xfrm_output(struct sock *sk, struct sk_buff *skb)
> switch (x->outer_mode.family) {
> case AF_INET:
> memset(IPCB(skb), 0, sizeof(*IPCB(skb)));
> -#ifdef CONFIG_NETFILTER
> IPCB(skb)->flags |= IPSKB_XFRM_TRANSFORMED;
> -#endif
> break;
> case AF_INET6:
> memset(IP6CB(skb), 0, sizeof(*IP6CB(skb)));
>
> -#ifdef CONFIG_NETFILTER
> IP6CB(skb)->flags |= IP6SKB_XFRM_TRANSFORMED;
> -#endif
> break;
> }
>
It looks like this merge resolution fix up was missed when the
ipsec-next tree was merged into the net-next tree.
--
Cheers,
Stephen Rothwell
On 6/1/20 5:20 PM, Stephen Rothwell wrote:
> Hi Dave,
>
> On Mon, 11 May 2020 13:00:15 +1000 Stephen Rothwell <[email protected]> wrote:
>>
>> Today's linux-next merge of the ipsec-next tree got conflicts in:
>>
>> net/ipv4/xfrm4_output.c
>> net/ipv6/xfrm6_output.c
>>
>> between commit:
>>
>> 0c922a4850eb ("xfrm: Always set XFRM_TRANSFORMED in xfrm{4,6}_output_finish")
>>
>> from Linus' tree and commit:
>>
>> 2ab6096db2f1 ("xfrm: remove output_finish indirection from xfrm_state_afinfo")
>>
>> from the ipsec-next tree.
>>
>> I fixed it up (I used the latter versions of these files and then added
>> the following patch) and can carry the fix as necessary. This is now fixed
>> as far as linux-next is concerned, but any non trivial conflicts should
>> be mentioned to your upstream maintainer when your tree is submitted for
>> merging. You may also want to consider cooperating with the maintainer
>> of the conflicting tree to minimise any particularly complex conflicts.
>>
>> From: Stephen Rothwell <[email protected]>
>> Date: Mon, 11 May 2020 12:57:24 +1000
>> Subject: [PATCH] xfrm: merge fixup for "remove output_finish indirection from xfrm_state_afinfo"
>>
>> Signed-off-by: Stephen Rothwell <[email protected]>
>> ---
>> net/xfrm/xfrm_output.c | 4 ----
>> 1 file changed, 4 deletions(-)
>>
>> diff --git a/net/xfrm/xfrm_output.c b/net/xfrm/xfrm_output.c
>> index 886a9b284b3a..0f4b3a5e02ba 100644
>> --- a/net/xfrm/xfrm_output.c
>> +++ b/net/xfrm/xfrm_output.c
>> @@ -574,16 +574,12 @@ int xfrm_output(struct sock *sk, struct sk_buff *skb)
>> switch (x->outer_mode.family) {
>> case AF_INET:
>> memset(IPCB(skb), 0, sizeof(*IPCB(skb)));
>> -#ifdef CONFIG_NETFILTER
>> IPCB(skb)->flags |= IPSKB_XFRM_TRANSFORMED;
>> -#endif
>> break;
>> case AF_INET6:
>> memset(IP6CB(skb), 0, sizeof(*IP6CB(skb)));
>>
>> -#ifdef CONFIG_NETFILTER
>> IP6CB(skb)->flags |= IP6SKB_XFRM_TRANSFORMED;
>> -#endif
>> break;
>> }
>>
>
> It looks like this merge resolution fix up was missed when the
> ipsec-next tree was merged into the net-next tree.
>
Thanks for catching that, Stephen. Important fix in 0c922a4850eb.
Hi David,
On Mon, 1 Jun 2020 20:30:34 -0600 David Ahern <[email protected]> wrote:
>
> On 6/1/20 5:20 PM, Stephen Rothwell wrote:
> > Hi Dave,
> >
> > On Mon, 11 May 2020 13:00:15 +1000 Stephen Rothwell <[email protected]> wrote:
> >>
> >> Today's linux-next merge of the ipsec-next tree got conflicts in:
> >>
> >> net/ipv4/xfrm4_output.c
> >> net/ipv6/xfrm6_output.c
> >>
> >> between commit:
> >>
> >> 0c922a4850eb ("xfrm: Always set XFRM_TRANSFORMED in xfrm{4,6}_output_finish")
> >>
> >> from Linus' tree and commit:
> >>
> >> 2ab6096db2f1 ("xfrm: remove output_finish indirection from xfrm_state_afinfo")
> >>
> >> from the ipsec-next tree.
> >>
> >> I fixed it up (I used the latter versions of these files and then added
> >> the following patch) and can carry the fix as necessary. This is now fixed
> >> as far as linux-next is concerned, but any non trivial conflicts should
> >> be mentioned to your upstream maintainer when your tree is submitted for
> >> merging. You may also want to consider cooperating with the maintainer
> >> of the conflicting tree to minimise any particularly complex conflicts.
> >>
> >> From: Stephen Rothwell <[email protected]>
> >> Date: Mon, 11 May 2020 12:57:24 +1000
> >> Subject: [PATCH] xfrm: merge fixup for "remove output_finish indirection from xfrm_state_afinfo"
> >>
> >> Signed-off-by: Stephen Rothwell <[email protected]>
> >> ---
> >> net/xfrm/xfrm_output.c | 4 ----
> >> 1 file changed, 4 deletions(-)
> >>
> >> diff --git a/net/xfrm/xfrm_output.c b/net/xfrm/xfrm_output.c
> >> index 886a9b284b3a..0f4b3a5e02ba 100644
> >> --- a/net/xfrm/xfrm_output.c
> >> +++ b/net/xfrm/xfrm_output.c
> >> @@ -574,16 +574,12 @@ int xfrm_output(struct sock *sk, struct sk_buff *skb)
> >> switch (x->outer_mode.family) {
> >> case AF_INET:
> >> memset(IPCB(skb), 0, sizeof(*IPCB(skb)));
> >> -#ifdef CONFIG_NETFILTER
> >> IPCB(skb)->flags |= IPSKB_XFRM_TRANSFORMED;
> >> -#endif
> >> break;
> >> case AF_INET6:
> >> memset(IP6CB(skb), 0, sizeof(*IP6CB(skb)));
> >>
> >> -#ifdef CONFIG_NETFILTER
> >> IP6CB(skb)->flags |= IP6SKB_XFRM_TRANSFORMED;
> >> -#endif
> >> break;
> >> }
> >>
> >
> > It looks like this merge resolution fix up was missed when the
> > ipsec-next tree was merged into the net-next tree.
> >
>
> Thanks for catching that, Stephen. Important fix in 0c922a4850eb.
>
And now the net-next tree has been merged into Linus' tree without this fix :-(
--
Cheers,
Stephen Rothwell
On 6/3/20 7:26 PM, Stephen Rothwell wrote:
>
> And now the net-next tree has been merged into Linus' tree without this fix :-(
>
I took a look earlier and I think it is fine. Some code was moved around
in ipsec-next and I think the merge is good. I'll run the test cases
later this week and double check. Thanks for the reminder
On Wed, Jun 03, 2020 at 08:55:01PM -0600, David Ahern wrote:
> On 6/3/20 7:26 PM, Stephen Rothwell wrote:
> >
> > And now the net-next tree has been merged into Linus' tree without this fix :-(
> >
>
> I took a look earlier and I think it is fine. Some code was moved around
> in ipsec-next and I think the merge is good. I'll run the test cases
> later this week and double check. Thanks for the reminder
The setting of XFRM_TRANSFORMED moved to xfrm_output() and depends
on CONFIG_NETFILTER. So I think the fix is needed. After the merge
of the net tree today, I have both conflicting patches patches in
the ipsec tree. I'd apply the fix from Stephen unless you say
it is not needed.
On 6/4/20 12:41 AM, Steffen Klassert wrote:
> On Wed, Jun 03, 2020 at 08:55:01PM -0600, David Ahern wrote:
>> On 6/3/20 7:26 PM, Stephen Rothwell wrote:
>>>
>>> And now the net-next tree has been merged into Linus' tree without this fix :-(
>>>
>>
>> I took a look earlier and I think it is fine. Some code was moved around
>> in ipsec-next and I think the merge is good. I'll run the test cases
>> later this week and double check. Thanks for the reminder
>
> The setting of XFRM_TRANSFORMED moved to xfrm_output() and depends
> on CONFIG_NETFILTER. So I think the fix is needed. After the merge
> of the net tree today, I have both conflicting patches patches in
> the ipsec tree. I'd apply the fix from Stephen unless you say
> it is not needed.
>
Indeed. I must have been looking at -net. Both -net and -net-next have
it conditional, so yes a fixup patch is needed.
On Thu, Jun 04, 2020 at 06:44:10AM -0600, David Ahern wrote:
> On 6/4/20 12:41 AM, Steffen Klassert wrote:
> > On Wed, Jun 03, 2020 at 08:55:01PM -0600, David Ahern wrote:
> >> On 6/3/20 7:26 PM, Stephen Rothwell wrote:
> >>>
> >>> And now the net-next tree has been merged into Linus' tree without this fix :-(
> >>>
> >>
> >> I took a look earlier and I think it is fine. Some code was moved around
> >> in ipsec-next and I think the merge is good. I'll run the test cases
> >> later this week and double check. Thanks for the reminder
> >
> > The setting of XFRM_TRANSFORMED moved to xfrm_output() and depends
> > on CONFIG_NETFILTER. So I think the fix is needed. After the merge
> > of the net tree today, I have both conflicting patches patches in
> > the ipsec tree. I'd apply the fix from Stephen unless you say
> > it is not needed.
> >
>
> Indeed. I must have been looking at -net. Both -net and -net-next have
> it conditional, so yes a fixup patch is needed.
The fixup patch from Stephen is now applied to the ipsec tree.
On 6/4/20 6:44 AM, David Ahern wrote:
> On 6/4/20 12:41 AM, Steffen Klassert wrote:
>> On Wed, Jun 03, 2020 at 08:55:01PM -0600, David Ahern wrote:
>>> On 6/3/20 7:26 PM, Stephen Rothwell wrote:
>>>>
>>>> And now the net-next tree has been merged into Linus' tree without this fix :-(
>>>>
>>>
>>> I took a look earlier and I think it is fine. Some code was moved around
>>> in ipsec-next and I think the merge is good. I'll run the test cases
>>> later this week and double check. Thanks for the reminder
>>
>> The setting of XFRM_TRANSFORMED moved to xfrm_output() and depends
>> on CONFIG_NETFILTER. So I think the fix is needed. After the merge
>> of the net tree today, I have both conflicting patches patches in
>> the ipsec tree. I'd apply the fix from Stephen unless you say
>> it is not needed.
>>
>
> Indeed. I must have been looking at -net. Both -net and -net-next have
> it conditional, so yes a fixup patch is needed.
>
I see that both net and net-next still have the conditional in xfrm_output:
#ifdef CONFIG_NETFILTER
IPCB(skb)->flags |= IPSKB_XFRM_TRANSFORMED;
#endif
break;
case AF_INET6:
memset(IP6CB(skb), 0, sizeof(*IP6CB(skb)));
#ifdef CONFIG_NETFILTER
IP6CB(skb)->flags |= IP6SKB_XFRM_TRANSFORMED;
#endif
Did you submit the merge fix? If not, I can do it today.
On Tue, Jun 16, 2020 at 07:39:30AM -0600, David Ahern wrote:
> >
> > Indeed. I must have been looking at -net. Both -net and -net-next have
> > it conditional, so yes a fixup patch is needed.
> >
>
> I see that both net and net-next still have the conditional in xfrm_output:
>
> #ifdef CONFIG_NETFILTER
> IPCB(skb)->flags |= IPSKB_XFRM_TRANSFORMED;
> #endif
> break;
> case AF_INET6:
> memset(IP6CB(skb), 0, sizeof(*IP6CB(skb)));
>
> #ifdef CONFIG_NETFILTER
> IP6CB(skb)->flags |= IP6SKB_XFRM_TRANSFORMED;
> #endif
>
> Did you submit the merge fix? If not, I can do it today.
I still have it in the ipsec tree, I'll do a pull request
this week. The fixup will go to the net tree then. It should
be already in linux-next.