Patchwork [net-next] net: flow_dissector.c missing include linux/export.h

login
register
mail settings
Submitter Jesper Dangaard Brouer
Date Jan. 24, 2012, 3:57 p.m.
Message ID <1327420643.4549.113.camel@probook>
Download mbox | patch
Permalink /patch/137580/
State Accepted
Delegated to: David Miller
Headers show

Comments

Jesper Dangaard Brouer - Jan. 24, 2012, 3:57 p.m.
The file net/core/flow_dissector.c seems to be missing
including linux/export.h.

Signed-off-by: Jesper Dangaard Brouer <hawk@comx.dk>
---
 net/core/flow_dissector.c |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)
Eric Dumazet - Jan. 24, 2012, 4:15 p.m.
Le mardi 24 janvier 2012 à 16:57 +0100, Jesper Dangaard Brouer a écrit :
> The file net/core/flow_dissector.c seems to be missing
> including linux/export.h.
> 
> Signed-off-by: Jesper Dangaard Brouer <hawk@comx.dk>
> ---
>  net/core/flow_dissector.c |    1 +
>  1 files changed, 1 insertions(+), 0 deletions(-)
> 
> diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c
> index 0985b9b..a225089 100644
> --- a/net/core/flow_dissector.c
> +++ b/net/core/flow_dissector.c
> @@ -1,4 +1,5 @@
>  #include <linux/skbuff.h>
> +#include <linux/export.h>
>  #include <linux/ip.h>
>  #include <linux/ipv6.h>
>  #include <linux/if_vlan.h>


Acked-by: Eric Dumazet <eric.dumazet@gmail.com>

Hmm, not sure how you found this, since export.h is included from
linux/netdevice.h (via net/netprio_cgroup.h & linux/module.h)

Or maybe you want to remove linux/module.h inclusion from
net/netprio_cgroup.h ???



--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Paul Gortmaker - Jan. 24, 2012, 7:44 p.m.
On Tue, Jan 24, 2012 at 11:15 AM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> Le mardi 24 janvier 2012 à 16:57 +0100, Jesper Dangaard Brouer a écrit :
>> The file net/core/flow_dissector.c seems to be missing
>> including linux/export.h.
>>
>> Signed-off-by: Jesper Dangaard Brouer <hawk@comx.dk>
>> ---
>>  net/core/flow_dissector.c |    1 +
>>  1 files changed, 1 insertions(+), 0 deletions(-)
>>
>> diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c
>> index 0985b9b..a225089 100644
>> --- a/net/core/flow_dissector.c
>> +++ b/net/core/flow_dissector.c
>> @@ -1,4 +1,5 @@
>>  #include <linux/skbuff.h>
>> +#include <linux/export.h>
>>  #include <linux/ip.h>
>>  #include <linux/ipv6.h>
>>  #include <linux/if_vlan.h>
>
>
> Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
>
> Hmm, not sure how you found this, since export.h is included from
> linux/netdevice.h (via net/netprio_cgroup.h & linux/module.h)
>
> Or maybe you want to remove linux/module.h inclusion from
> net/netprio_cgroup.h ???

That would be the better solution.  I had them all killed off and it
would be nice to try and keep it that way as long as possible.

I'll look into this if that wasn't part of Jesper's original intention.

Paul.

>
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
David Miller - Jan. 24, 2012, 8:51 p.m.
From: Paul Gortmaker <paul.gortmaker@windriver.com>
Date: Tue, 24 Jan 2012 14:44:21 -0500

> On Tue, Jan 24, 2012 at 11:15 AM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
>> Le mardi 24 janvier 2012 à 16:57 +0100, Jesper Dangaard Brouer a écrit :
>>> The file net/core/flow_dissector.c seems to be missing
>>> including linux/export.h.
>>>
>>> Signed-off-by: Jesper Dangaard Brouer <hawk@comx.dk>
>>> ---
>>>  net/core/flow_dissector.c |    1 +
>>>  1 files changed, 1 insertions(+), 0 deletions(-)
>>>
>>> diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c
>>> index 0985b9b..a225089 100644
>>> --- a/net/core/flow_dissector.c
>>> +++ b/net/core/flow_dissector.c
>>> @@ -1,4 +1,5 @@
>>>  #include <linux/skbuff.h>
>>> +#include <linux/export.h>
>>>  #include <linux/ip.h>
>>>  #include <linux/ipv6.h>
>>>  #include <linux/if_vlan.h>
>>
>>
>> Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
>>
>> Hmm, not sure how you found this, since export.h is included from
>> linux/netdevice.h (via net/netprio_cgroup.h & linux/module.h)
>>
>> Or maybe you want to remove linux/module.h inclusion from
>> net/netprio_cgroup.h ???
> 
> That would be the better solution.  I had them all killed off and it
> would be nice to try and keep it that way as long as possible.
> 
> I'll look into this if that wasn't part of Jesper's original intention.

Jesper did you actually hit this problem and get a real build
failure?
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Paul Gortmaker - Jan. 24, 2012, 8:59 p.m.
On Tue, Jan 24, 2012 at 3:51 PM, David Miller <davem@davemloft.net> wrote:
> From: Paul Gortmaker <paul.gortmaker@windriver.com>
> Date: Tue, 24 Jan 2012 14:44:21 -0500
>
>> On Tue, Jan 24, 2012 at 11:15 AM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
>>> Le mardi 24 janvier 2012 à 16:57 +0100, Jesper Dangaard Brouer a écrit :
>>>> The file net/core/flow_dissector.c seems to be missing
>>>> including linux/export.h.
>>>>
>>>> Signed-off-by: Jesper Dangaard Brouer <hawk@comx.dk>
>>>> ---
>>>>  net/core/flow_dissector.c |    1 +
>>>>  1 files changed, 1 insertions(+), 0 deletions(-)
>>>>
>>>> diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c
>>>> index 0985b9b..a225089 100644
>>>> --- a/net/core/flow_dissector.c
>>>> +++ b/net/core/flow_dissector.c
>>>> @@ -1,4 +1,5 @@
>>>>  #include <linux/skbuff.h>
>>>> +#include <linux/export.h>
>>>>  #include <linux/ip.h>
>>>>  #include <linux/ipv6.h>
>>>>  #include <linux/if_vlan.h>
>>>
>>>
>>> Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
>>>
>>> Hmm, not sure how you found this, since export.h is included from
>>> linux/netdevice.h (via net/netprio_cgroup.h & linux/module.h)
>>>
>>> Or maybe you want to remove linux/module.h inclusion from
>>> net/netprio_cgroup.h ???
>>
>> That would be the better solution.  I had them all killed off and it
>> would be nice to try and keep it that way as long as possible.
>>
>> I'll look into this if that wasn't part of Jesper's original intention.
>
> Jesper did you actually hit this problem and get a real build
> failure?

I've reproduced it locally, and verified it as a fix, but only *after*
I've deleted the module.h from netprio_cgroup.h

Doing builds to see what else (aside from the dsa/mv88e6xxx)
shakes loose.  Hopefully that is it.  If so, I'll send a patch
shortly to delete the module.h

Paul.

> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
David Miller - Jan. 24, 2012, 9:03 p.m.
From: Paul Gortmaker <paul.gortmaker@windriver.com>
Date: Tue, 24 Jan 2012 15:59:00 -0500

> I've reproduced it locally, and verified it as a fix, but only *after*
> I've deleted the module.h from netprio_cgroup.h
> 
> Doing builds to see what else (aside from the dsa/mv88e6xxx)
> shakes loose.  Hopefully that is it.  If so, I'll send a patch
> shortly to delete the module.h

Ok, I'll apply Jesper's patch to 'net' then.

Thanks.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Jesper Dangaard Brouer - Jan. 24, 2012, 9:42 p.m.
On Tue, 2012-01-24 at 15:51 -0500, David Miller wrote:
> Jesper did you actually hit this problem and get a real build
> failure? 

Yes, but I didn't explain the circumstances, sorry.

I was backporting Erics SFQ bug fix (sch_sfq: dont put new flow at the
end of flows / d47a0ac7b66) to stable kernel 3.2.1.

But to avoid conflicts I also cherry-picked the other SFQ changes, plus
the skb_flow_dissect changes, and cherry-picked:

 # sch_sfq: use skb_flow_dissect()
 # (Need flow_dissect APIs)
 git cherry-pick 11fca931d35a34d01ce209eb8d51ff667c9f5e7c

 # net: introduce skb_flow_dissect()
 git cherry-pick 0744dd00c1b1be99a25b62b1b48df440e82e57e0

 # flow_dissector: use a 64bit load/store
 git cherry-pick 4d77d2b567ec66a443792d99e96ac760991d80d0

 # sch_sfq: rehash queues in perturb timer
 git cherry-pick 225d9b89c937633dfeec502741a174fe0bab5b9f

 # sch_sfq: dont put new flow at the end of flows
 git cherry-pick  d47a0ac7b66883987275598d6039f902f4410ca9

 # net_sched: sfq: fix mem alloc error recovery
 git cherry-pick bd16a6cce2a7f169b559abc5672fd2c66e91fb36

So, I guess the "problem" is because I didn't have the patch, which adds
the include net/netprio_cgroup.h in linux/netdevice.h, and have the side
effect of including module.h, the patch:
 net: add network priority cgroup infrastructure (v4) / 5bc1421e34ecfe

git describe --contains 5bc1421e34ecfe
v3.3-rc1~182^2~519

Still think it more correct to include linux/export.h in
flow_dissector.c.

Patch

diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c
index 0985b9b..a225089 100644
--- a/net/core/flow_dissector.c
+++ b/net/core/flow_dissector.c
@@ -1,4 +1,5 @@ 
 #include <linux/skbuff.h>
+#include <linux/export.h>
 #include <linux/ip.h>
 #include <linux/ipv6.h>
 #include <linux/if_vlan.h>