It makes less sense to remove define of ipv6 variable from ipv6 file
and put it in ipv4 file and declare it in ipv6 file, which was done
in 3d010c8031e3 ("udp: do not accept non-tunnel GSO skbs landing in
a tunnel").

So let's move it back to ipv6 file. It also makes the code similar -
the key is defined right above the respective enable function.

Signed-off-by: Siddh Raman Pant <siddh.raman.pant@oracle.com>
---
I'm not sure why ipv4 key is exported using EXPORT_IPV6_MOD?

 net/ipv4/udp.c | 5 -----
 net/ipv6/udp.c | 4 +++-
 2 files changed, 3 insertions(+), 6 deletions(-)

diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index 30dfbf73729d..44b08ede7133 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -810,11 +810,6 @@ static inline bool __udp_is_mcast_sock(struct net *net, const struct sock *sk,
 DEFINE_STATIC_KEY_FALSE(udp_encap_needed_key);
 EXPORT_IPV6_MOD(udp_encap_needed_key);
 
-#if IS_ENABLED(CONFIG_IPV6)
-DEFINE_STATIC_KEY_FALSE(udpv6_encap_needed_key);
-EXPORT_IPV6_MOD(udpv6_encap_needed_key);
-#endif
-
 void udp_encap_enable(void)
 {
 	static_branch_inc(&udp_encap_needed_key);
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 813a2ba75824..6b92d4f466d5 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -596,7 +596,9 @@ int udpv6_recvmsg(struct sock *sk, struct msghdr *msg, size_t len,
 	goto try_again;
 }
 
-DECLARE_STATIC_KEY_FALSE(udpv6_encap_needed_key);
+DEFINE_STATIC_KEY_FALSE(udpv6_encap_needed_key);
+EXPORT_IPV6_MOD(udpv6_encap_needed_key);
+
 void udpv6_encap_enable(void)
 {
 	static_branch_inc(&udpv6_encap_needed_key);
-- 
2.51.0

Hi Siddh,

kernel test robot noticed the following build errors:

[auto build test ERROR on net-next/main]
[also build test ERROR on net/main linus/master v6.18-rc2 next-20251024]
[cannot apply to horms-ipvs/master]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Siddh-Raman-Pant/udp-Move-back-definition-of-udpv6_encap_needed_key-to-ipv6-file/20251023-171157
base:   net-next/main
patch link:    https://lore.kernel.org/r/20251023090736.99644-1-siddh.raman.pant%40oracle.com
patch subject: [PATCH] udp: Move back definition of udpv6_encap_needed_key to ipv6 file.
config: arm64-defconfig (https://download.01.org/0day-ci/archive/20251024/202510242214.RjDUfVB7-lkp@intel.com/config)
compiler: aarch64-linux-gcc (GCC) 15.1.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20251024/202510242214.RjDUfVB7-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202510242214.RjDUfVB7-lkp@intel.com/

All errors (new ones prefixed by >>):

>> aarch64-linux-ld: net/ipv4/udp.o:(__jump_table+0xe8): undefined reference to `udpv6_encap_needed_key'

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

2025-10-23, 14:37:36 +0530, Siddh Raman Pant wrote:
> It makes less sense to remove define of ipv6 variable from ipv6 file
> and put it in ipv4 file and declare it in ipv6 file, which was done
> in 3d010c8031e3 ("udp: do not accept non-tunnel GSO skbs landing in
> a tunnel").

It would be good to CC all the authors and reviewers of the patch(es)
you're mentioning.

> So let's move it back to ipv6 file. It also makes the code similar -
> the key is defined right above the respective enable function.

I don't think that works with CONFIG_IPV6=m. The ipv4 code will need
to access the key via udp_encap_needed().


> Signed-off-by: Siddh Raman Pant <siddh.raman.pant@oracle.com>
> ---
> I'm not sure why ipv4 key is exported using EXPORT_IPV6_MOD?

So that the ipv6 code can use it via
udp_encap_needed()/udp_unexpected_gso() when it's built as a module.

-- 
Sabrina

On Thu, Oct 23 2025 at 15:47:37 +0530, Sabrina Dubroca wrote:
> 2025-10-23, 14:37:36 +0530, Siddh Raman Pant wrote:
> > It makes less sense to remove define of ipv6 variable from ipv6 file
> > and put it in ipv4 file and declare it in ipv6 file, which was done
> > in 3d010c8031e3 ("udp: do not accept non-tunnel GSO skbs landing in
> > a tunnel").
> 
> It would be good to CC all the authors and reviewers of the patch(es)
> you're mentioning.
> 
> > So let's move it back to ipv6 file. It also makes the code similar -
> > the key is defined right above the respective enable function.
> 
> I don't think that works with CONFIG_IPV6=m. The ipv4 code will need
> to access the key via udp_encap_needed().
> 
> 
> > Signed-off-by: Siddh Raman Pant <siddh.raman.pant@oracle.com>
> > ---
> > I'm not sure why ipv4 key is exported using EXPORT_IPV6_MOD?
> 
> So that the ipv6 code can use it via
> udp_encap_needed()/udp_unexpected_gso() when it's built as a module.

Cool, thank you very much for the explanation! That was what I was
missing.

Please ignore this patch then.

Thanks,
Siddh