From: "Zhang, Helin" <helin.zhang@intel.com>
To: Dex Chen <dex.chen@ruckuswireless.com>
Cc: "dev@dpdk.org" <dev@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH] kni: allow per-net instances
Date: Thu, 22 Oct 2015 03:28:11 +0000 [thread overview]
Message-ID: <F35DEAC7BCE34641BA9FAC6BCA4A12E70A91AB4A@SHSMSX104.ccr.corp.intel.com> (raw)
In-Reply-To: <1435831933-13339-1-git-send-email-dex.chen@ruckuswireless.com>
Hi Dex
Two comments inlined. Thank you very much for the really good contribution to KNI!
Regards,
Helin
> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Dex Chen
> Sent: Thursday, July 2, 2015 6:12 PM
> To: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH] kni: allow per-net instances
>
> There is a global variable 'device_in_use' which is used to make sure only one
> instance is using /dev/kni device. If you were using LXC, you will find there is only
> one instance of KNI example could be run even differnt namespaces were
> created.
>
> In order to have /dev/kni used simultaneously in different namespaces, making
> all of global variables as per network namespace variables.
>
> With regard to single kernel thread mode, there will be one kernel thread for
> each of network namespace.
>
> Signed-off-by: Dex Chen <dex.chen@ruckuswireless.com>
> ---
> lib/librte_eal/linuxapp/kni/kni_misc.c | 129
> ++++++++++++++++++++++-----------
> 1 file changed, 85 insertions(+), 44 deletions(-)
>
> diff --git a/lib/librte_eal/linuxapp/kni/kni_misc.c
> b/lib/librte_eal/linuxapp/kni/kni_misc.c
> index 2e9fa89..5ba8ab8 100644
> --- a/lib/librte_eal/linuxapp/kni/kni_misc.c
> +++ b/lib/librte_eal/linuxapp/kni/kni_misc.c
> @@ -28,6 +28,9 @@
> #include <linux/pci.h>
> #include <linux/kthread.h>
> #include <linux/rwsem.h>
> +#include <linux/nsproxy.h>
> +#include <net/net_namespace.h>
> +#include <net/netns/generic.h>
>
> #include <exec-env/rte_kni_common.h>
> #include "kni_dev.h"
> @@ -90,18 +93,48 @@ static unsigned multiple_kthread_on = 0;
>
> #define KNI_DEV_IN_USE_BIT_NUM 0 /* Bit number for device in use */
>
> -static volatile unsigned long device_in_use; /* device in use flag */ -static struct
> task_struct *kni_kthread;
> +static int kni_net_id;
>
> -/* kni list lock */
> -static DECLARE_RWSEM(kni_list_lock);
> +struct kni_net {
> + volatile unsigned long device_in_use; /* device in use flag */
> + struct task_struct *kni_kthread;
> + struct rw_semaphore kni_list_lock;
> + struct list_head kni_list_head;
> +};
> +
> +static __net_init int kni_init_net(struct net *net) {
> + struct kni_net *knet = net_generic(net, kni_net_id);
>
> -/* kni list */
> -static struct list_head kni_list_head = LIST_HEAD_INIT(kni_list_head);
> + /* Clear the bit of device in use */
> + clear_bit(KNI_DEV_IN_USE_BIT_NUM, &knet->device_in_use);
> +
> + init_rwsem(&knet->kni_list_lock);
> + INIT_LIST_HEAD(&knet->kni_list_head);
> +
> + return 0;
> +}
> +
> +static __net_exit void kni_exit_net(struct net *net) {
> + /*
> + * Nothing to do here.
> + * Assuming all cleanup jobs were done in kni_release().
> + */
> +}
Agree with Stephen, kernel should handle it well.
> +
> +static struct pernet_operations kni_net_ops = {
> + .init = kni_init_net,
> + .exit = kni_exit_net,
> + .id = &kni_net_id,
> + .size = sizeof(struct kni_net),
> +};
>
> static int __init
> kni_init(void)
> {
> + int rc;
> +
> KNI_PRINT("######## DPDK kni module loading ########\n");
>
> if (kni_parse_kthread_mode() < 0) {
> @@ -114,8 +147,9 @@ kni_init(void)
> return -EPERM;
> }
>
> - /* Clear the bit of device in use */
> - clear_bit(KNI_DEV_IN_USE_BIT_NUM, &device_in_use);
> + rc = register_pernet_subsys(&kni_net_ops);
> + if (rc)
> + goto out;
>
> /* Configure the lo mode according to the input parameter */
> kni_net_config_lo_mode(lo_mode);
> @@ -123,11 +157,16 @@ kni_init(void)
> KNI_PRINT("######## DPDK kni module loaded ########\n");
>
> return 0;
> +
> +out:
> + misc_deregister(&kni_misc);
> + return rc;
> }
>
> static void __exit
> kni_exit(void)
> {
> + unregister_pernet_subsys(&kni_net_ops);
Should above 'unregsiter' be moved after 'misc_deregister()'?
> misc_deregister(&kni_misc);
> KNI_PRINT("####### DPDK kni module unloaded #######\n"); } @@
> -151,19 +190,22 @@ kni_parse_kthread_mode(void) static int
next prev parent reply other threads:[~2015-10-22 3:28 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-07-02 10:12 Dex Chen
2015-10-21 15:40 ` Thomas Monjalon
2015-10-21 16:22 ` Stephen Hemminger
2015-10-22 3:28 ` Zhang, Helin [this message]
2015-10-26 10:41 ` [dpdk-dev] [PATCH v2] " Dex Chen
2015-10-27 2:56 ` [dpdk-dev] [PATCH v3] " Dex Chen
2015-10-27 3:03 ` Zhang, Helin
2015-11-04 1:29 ` Thomas Monjalon
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=F35DEAC7BCE34641BA9FAC6BCA4A12E70A91AB4A@SHSMSX104.ccr.corp.intel.com \
--to=helin.zhang@intel.com \
--cc=dev@dpdk.org \
--cc=dex.chen@ruckuswireless.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).