* [PATCH 2/3] netns: add dummy struct inside "struct net_generic"
@ 2016-12-02 1:12 Alexey Dobriyan
2016-12-02 5:42 ` Cong Wang
0 siblings, 1 reply; 3+ messages in thread
From: Alexey Dobriyan @ 2016-12-02 1:12 UTC (permalink / raw)
To: davem; +Cc: netdev, xemul
This is precursor to fixing "[id - 1]" bloat inside net_generic().
Name "s" is chosen to complement name "u" often used for dummy unions.
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
---
include/net/netns/generic.h | 6 ++++--
net/core/net_namespace.c | 8 ++++----
2 files changed, 8 insertions(+), 6 deletions(-)
--- a/include/net/netns/generic.h
+++ b/include/net/netns/generic.h
@@ -25,8 +25,10 @@
*/
struct net_generic {
- unsigned int len;
- struct rcu_head rcu;
+ struct {
+ unsigned int len;
+ struct rcu_head rcu;
+ } s;
void *ptr[0];
};
--- a/net/core/net_namespace.c
+++ b/net/core/net_namespace.c
@@ -50,7 +50,7 @@ static struct net_generic *net_alloc_generic(void)
ng = kzalloc(generic_size, GFP_KERNEL);
if (ng)
- ng->len = max_gen_ptrs;
+ ng->s.len = max_gen_ptrs;
return ng;
}
@@ -64,7 +64,7 @@ static int net_assign_generic(struct net *net, unsigned int id, void *data)
old_ng = rcu_dereference_protected(net->gen,
lockdep_is_held(&net_mutex));
- if (old_ng->len >= id) {
+ if (old_ng->s.len >= id) {
old_ng->ptr[id - 1] = data;
return 0;
}
@@ -84,11 +84,11 @@ static int net_assign_generic(struct net *net, unsigned int id, void *data)
* the old copy for kfree after a grace period.
*/
- memcpy(&ng->ptr, &old_ng->ptr, old_ng->len * sizeof(void*));
+ memcpy(&ng->ptr, &old_ng->ptr, old_ng->s.len * sizeof(void*));
ng->ptr[id - 1] = data;
rcu_assign_pointer(net->gen, ng);
- kfree_rcu(old_ng, rcu);
+ kfree_rcu(old_ng, s.rcu);
return 0;
}
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH 2/3] netns: add dummy struct inside "struct net_generic"
2016-12-02 1:12 [PATCH 2/3] netns: add dummy struct inside "struct net_generic" Alexey Dobriyan
@ 2016-12-02 5:42 ` Cong Wang
2016-12-02 6:53 ` Cong Wang
0 siblings, 1 reply; 3+ messages in thread
From: Cong Wang @ 2016-12-02 5:42 UTC (permalink / raw)
To: Alexey Dobriyan
Cc: David Miller, Linux Kernel Network Developers, Pavel Emelyanov
On Thu, Dec 1, 2016 at 5:12 PM, Alexey Dobriyan <adobriyan@gmail.com> wrote:
> struct net_generic {
> - unsigned int len;
> - struct rcu_head rcu;
> + struct {
> + unsigned int len;
> + struct rcu_head rcu;
> + } s;
>
> void *ptr[0];
> };
I think you can put them in a union, since rcu is only used
for kfree_rcu() where len is not needed and rcu_head doesn't
need to be initialized by callers.
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH 2/3] netns: add dummy struct inside "struct net_generic"
2016-12-02 5:42 ` Cong Wang
@ 2016-12-02 6:53 ` Cong Wang
0 siblings, 0 replies; 3+ messages in thread
From: Cong Wang @ 2016-12-02 6:53 UTC (permalink / raw)
To: Alexey Dobriyan
Cc: David Miller, Linux Kernel Network Developers, Pavel Emelyanov
On Thu, Dec 1, 2016 at 9:42 PM, Cong Wang <xiyou.wangcong@gmail.com> wrote:
> On Thu, Dec 1, 2016 at 5:12 PM, Alexey Dobriyan <adobriyan@gmail.com> wrote:
>> struct net_generic {
>> - unsigned int len;
>> - struct rcu_head rcu;
>> + struct {
>> + unsigned int len;
>> + struct rcu_head rcu;
>> + } s;
>>
>> void *ptr[0];
>> };
>
> I think you can put them in a union, since rcu is only used
> for kfree_rcu() where len is not needed and rcu_head doesn't
> need to be initialized by callers.
Never mind, readers could be still reading ->len while we modify
->rcu. So they can't be in a union.
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2016-12-02 6:53 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-12-02 1:12 [PATCH 2/3] netns: add dummy struct inside "struct net_generic" Alexey Dobriyan
2016-12-02 5:42 ` Cong Wang
2016-12-02 6:53 ` Cong Wang
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.