On Wed, Dec 14, 2011 at 2:47 PM, Ansis Atteka <aatt...@nicira.com> wrote:
>
> On Wed, Dec 14, 2011 at 2:08 PM, Pravin B Shelar <pshe...@nicira.com> wrote:
>>
>> genl_lock is not exported from older kernel. Following patch add
>> genl_exec() which can run any function (passed as arg) with
>> genl_lock held.
>>
>> Signed-off-by: Pravin B Shelar <pshe...@nicira.com>
>> ---
>> datapath/datapath.c | 9 ++-
>> datapath/linux/compat/genetlink.inc | 102
>> +++++++++++++++++++++++++
>> datapath/linux/compat/include/net/genetlink.h | 6 ++
>> 3 files changed, 116 insertions(+), 1 deletions(-)
>>
>> diff --git a/datapath/datapath.c b/datapath/datapath.c
>> index c86c20b..025d932 100644
>> --- a/datapath/datapath.c
>> +++ b/datapath/datapath.c
>> @@ -2049,10 +2049,14 @@ static int __init dp_init(void)
>> pr_info("Open vSwitch switching datapath %s, built "__DATE__"
>> "__TIME__"\n",
>> VERSION BUILDNR);
>>
>> - err = ovs_tnl_init();
>> + err = genl_exec_init();
>> if (err)
>> goto error;
>>
>> + err = ovs_tnl_init();
>> + if (err)
>> + goto error_genl_exec;
>> +
>> err = ovs_flow_init();
>> if (err)
>> goto error_tnl_exit;
>> @@ -2079,6 +2083,8 @@ error_flow_exit:
>> ovs_flow_exit();
>> error_tnl_exit:
>> ovs_tnl_exit();
>> +error_genl_exec:
>> + genl_exec_exit();
>> error:
>> return err;
>> }
>> @@ -2091,6 +2097,7 @@ static void dp_cleanup(void)
>> ovs_vport_exit();
>> ovs_flow_exit();
>> ovs_tnl_exit();
>> + genl_exec_exit();
>> }
>>
>> module_init(dp_init);
>> diff --git a/datapath/linux/compat/genetlink.inc
>> b/datapath/linux/compat/genetlink.inc
>> index bf96980..b4de70e 100644
>> --- a/datapath/linux/compat/genetlink.inc
>> +++ b/datapath/linux/compat/genetlink.inc
>> @@ -146,3 +146,105 @@ void netlink_set_err(struct sock *ssk, u32 pid, u32
>> group, int code)
>> {
>> }
>> #endif
>> +
>> +#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,35)
>> +
>> +static DEFINE_MUTEX(genl_exec_lock);
>> +
>> +static genl_exec_func_t genl_exec_function;
>> +static int genl_exec_function_ret;
>> +static void *genl_exec_data;
>> +static struct completion done;
>> +
>> +
>> +static int genl_exec_cmd(struct sk_buff *dummy, struct genl_info *dummy2)
>> +{
>> + if (genl_exec_function)
>> + genl_exec_function_ret =
>> genl_exec_function(genl_exec_data);
>> + complete(&done);
>> + return 0;
>> +}
>> +
>> +enum genl_exec_attr {
>> + GENL_EXEC_UNSPEC,
>> + GENL_EXEC_RUN,
>> + GENL_EXEC_MAX,
>> +};
>> +
>> +static struct genl_family genl_exec_family = {
>> + .id = GENL_ID_GENERATE,
>> + .hdrsize = 0,
>> + .name = "GENL EXEC FAMILY",
>> + .version = 1,
>> + .maxattr = GENL_EXEC_MAX,
>> +};
>> +
>> +static struct genl_ops genl_exec_ops[] = {
>> + {
>> + .cmd = GENL_EXEC_RUN,
>> + .doit = genl_exec_cmd,
>> + .flags = CAP_NET_ADMIN,
>> + },
>> +};
>> +
>> +int genl_exec_init(void)
>> +{
>> + int err;
>> +
>> + err = genl_register_family_with_ops(&genl_exec_family,
>> + genl_exec_ops,
>> + ARRAY_SIZE(genl_exec_ops));
>> + return err;
>> +}
>> +
>> +void genl_exec_exit(void)
>> +{
>> + genl_unregister_family(&genl_exec_family);
>> +}
>> +
>> +/* genl_lock() is not exported from older kernel.
>> + * Following function allows any function to be executed with
>> + * genl_lock held. */
>
> You meant with "genl_exec_lock" instead of "genl_lock", right?
not really. this whole function is written to execute a function with
genl_lock() (genl_mutex) which is in kernel lock. only way to get that
lock is to execute code in genetlink callback.
I will update the comment to genl_mutex which is less confusing.
>>
>> +
>> +int genl_exec(genl_exec_func_t func, void *data)
>> +{
>> + struct sk_buff *skb;
>> +
>> + skb = genlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
>> + if (!skb)
>> + return -ENOMEM;
>> +
>> + genlmsg_put(skb, 0, 0, &genl_exec_family, NLM_F_REQUEST,
>> GENL_EXEC_RUN);
>> +
>> + mutex_lock(&genl_exec_lock);
>> + genl_exec_function = func;
>> + genl_exec_data = data;
>> + init_completion(&done);
>> + genlmsg_unicast(&init_net, skb, 0);
>> + wait_for_completion(&done);
>> + genl_exec_function = NULL;
>
> Shouldn't you prepare here a copy of genl_exec_function_ret before actually
> returning it? I sense a small-probability of a race condition here.
>>
right, I need to do it inside lock. Even with lock it is not
completely safe as callback is not taking any lock. But I think that
is not issue for now.
>> + mutex_unlock(&genl_exec_lock);
>> +
>> + return genl_exec_function_ret;
>> +}
>> +
>> +#else
>> +
>> +int genl_exec(genl_exec_func_t func, void *data)
>> +{
>> + int ret;
>> +
>> + genl_lock();
>> + ret = func(data);
>> + genl_unlock();
>> + return ret;
>> +}
>> +int genl_exec_init(void)
>> +{
>> + return 0;
>> +}
>> +void genl_exec_exit(void)
>> +{
>> +}
>> +
>> +#endif
>> diff --git a/datapath/linux/compat/include/net/genetlink.h
>> b/datapath/linux/compat/include/net/genetlink.h
>> index a1ff7c1..1bbb559 100644
>> --- a/datapath/linux/compat/include/net/genetlink.h
>> +++ b/datapath/linux/compat/include/net/genetlink.h
>> @@ -170,4 +170,10 @@ static inline struct net *genl_info_net(struct
>> genl_info *info)
>> #if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,32)
>> #define genlmsg_unicast(ignore_net, skb, pid) genlmsg_unicast(skb, pid)
>> #endif
>> +
>> +typedef int (*genl_exec_func_t)(void *data);
>> +int genl_exec(genl_exec_func_t func, void *data);
>> +int genl_exec_init(void);
>> +void genl_exec_exit(void);
>> +
>> #endif /* genetlink.h */
>> --
>> 1.7.1
>>
>> _______________________________________________
>> dev mailing list
>> dev@openvswitch.org
>> http://openvswitch.org/mailman/listinfo/dev
>
>
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
