Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1642176
| From | Boris Ostrovsky <boris.ostrovsky@oracle.com> |
|---|---|
| Newsgroups | linux.kernel |
| Subject | Re: [PATCH 03/18] xen/pvcalls: initialize the module and register the xenbus backend |
| Date | 2017-05-16 03:50 +0200 |
| Message-ID | <tHzMC-7sL-7@gated-at.bofh.it> (permalink) |
| References | <tHuWC-4t8-7@gated-at.bofh.it> <tHuWC-4t8-9@gated-at.bofh.it> <tHuWC-4t8-25@gated-at.bofh.it> |
| Organization | linux.* mail to news gateway |
On 05/15/2017 04:35 PM, Stefano Stabellini wrote:
> The pvcalls backend has one ioworker per cpu: the ioworkers are
> implemented as a cpu bound workqueue, and will deal with the actual
> socket and data ring reads/writes.
>
> ioworkers are global: we only have one set for all the frontends. They
> process requests on their wqs list in order, once they are done with a
> request, they'll remove it from the list. A spinlock is used for
> protecting the list. Each ioworker is bound to a different cpu to
> maximize throughput.
>
> Signed-off-by: Stefano Stabellini <stefano@aporeto.com>
> CC: boris.ostrovsky@oracle.com
> CC: jgross@suse.com
> ---
> drivers/xen/pvcalls-back.c | 64 ++++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 64 insertions(+)
>
> diff --git a/drivers/xen/pvcalls-back.c b/drivers/xen/pvcalls-back.c
> index 2dbf7d8..46a889a 100644
> --- a/drivers/xen/pvcalls-back.c
> +++ b/drivers/xen/pvcalls-back.c
> @@ -25,6 +25,26 @@
> #include <xen/xenbus.h>
> #include <xen/interface/io/pvcalls.h>
>
> +struct pvcalls_ioworker {
> + struct work_struct register_work;
> + atomic_t io;
> + struct list_head wqs;
> + spinlock_t lock;
> + int num;
> +};
> +
> +struct pvcalls_back_global {
> + struct pvcalls_ioworker *ioworkers;
> + int nr_ioworkers;
> + struct workqueue_struct *wq;
> + struct list_head privs;
> + struct rw_semaphore privs_lock;
Is there a reason why these are called "privs"?
And why are you using a rw semaphore --- I only noticed two instances of
use and both are writes.
> +} pvcalls_back_global;
> +
> +static void pvcalls_back_ioworker(struct work_struct *work)
> +{
> +}
> +
> static int pvcalls_back_probe(struct xenbus_device *dev,
> const struct xenbus_device_id *id)
> {
> @@ -59,3 +79,47 @@ static int pvcalls_back_uevent(struct xenbus_device *xdev,
> .uevent = pvcalls_back_uevent,
> .otherend_changed = pvcalls_back_changed,
> };
> +
> +static int __init pvcalls_back_init(void)
> +{
> + int ret, i, cpu;
> +
> + if (!xen_domain())
> + return -ENODEV;
> +
> + ret = xenbus_register_backend(&pvcalls_back_driver);
> + if (ret < 0)
> + return ret;
> +
> + init_rwsem(&pvcalls_back_global.privs_lock);
> + INIT_LIST_HEAD(&pvcalls_back_global.privs);
> + pvcalls_back_global.wq = alloc_workqueue("pvcalls_io", 0, 0);
> + if (!pvcalls_back_global.wq)
> + goto error;
> + pvcalls_back_global.nr_ioworkers = num_online_cpus();
Should nr_ioworkers be updated on CPU hot(un)plug?
> + pvcalls_back_global.ioworkers = kzalloc(
> + sizeof(*pvcalls_back_global.ioworkers) *
> + pvcalls_back_global.nr_ioworkers, GFP_KERNEL);
> + if (!pvcalls_back_global.ioworkers)
> + goto error;
> + i = 0;
> + for_each_online_cpu(cpu) {
> + pvcalls_back_global.ioworkers[i].num = i;
> + atomic_set(&pvcalls_back_global.ioworkers[i].io, 1);
> + spin_lock_init(&pvcalls_back_global.ioworkers[i].lock);
> + INIT_LIST_HEAD(&pvcalls_back_global.ioworkers[i].wqs);
> + INIT_WORK(&pvcalls_back_global.ioworkers[i].register_work,
> + pvcalls_back_ioworker);
> + i++;
> + }
> + return 0;
> +
> +error:
> + if (pvcalls_back_global.wq)
> + destroy_workqueue(pvcalls_back_global.wq);
> + xenbus_unregister_driver(&pvcalls_back_driver);
> + kfree(pvcalls_back_global.ioworkers);
> + memset(&pvcalls_back_global, 0, sizeof(pvcalls_back_global));
> + return -ENOMEM;
This routine could use more newlines. (and in other patches too)
-boris
> +}
> +module_init(pvcalls_back_init);
>
Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
[PATCH 01/18] xen: introduce the pvcalls interface header Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
[PATCH 14/18] xen/pvcalls: disconnect and module_exit Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
[PATCH 02/18] xen/pvcalls: introduce the pvcalls xenbus backend Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
[PATCH 15/18] xen/pvcalls: introduce the ioworker Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
[PATCH 09/18] xen/pvcalls: implement bind command Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
[PATCH 03/18] xen/pvcalls: initialize the module and register the xenbus backend Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
Re: [PATCH 03/18] xen/pvcalls: initialize the module and register the xenbus backend Boris Ostrovsky <boris.ostrovsky@oracle.com> - 2017-05-16 03:50 +0200
Re: [PATCH 03/18] xen/pvcalls: initialize the module and register the xenbus backend Stefano Stabellini <sstabellini@kernel.org> - 2017-05-16 22:10 +0200
Re: [PATCH 03/18] xen/pvcalls: initialize the module and register the xenbus backend Stefano Stabellini <sstabellini@kernel.org> - 2017-05-16 22:30 +0200
Re: [PATCH 03/18] xen/pvcalls: initialize the module and register the xenbus backend Juergen Gross <jgross@suse.com> - 2017-05-16 08:50 +0200
Re: [PATCH 03/18] xen/pvcalls: initialize the module and register the xenbus backend Stefano Stabellini <sstabellini@kernel.org> - 2017-05-16 22:00 +0200
Re: [PATCH 03/18] xen/pvcalls: initialize the module and register the xenbus backend Juergen Gross <jgross@suse.com> - 2017-05-17 07:30 +0200
Re: [PATCH 03/18] xen/pvcalls: initialize the module and register the xenbus backend Stefano Stabellini <sstabellini@kernel.org> - 2017-05-18 23:20 +0200
Re: [PATCH 03/18] xen/pvcalls: initialize the module and register the xenbus backend Stefano Stabellini <sstabellini@kernel.org> - 2017-05-20 00:40 +0200
[PATCH 10/18] xen/pvcalls: implement listen command Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
[PATCH 13/18] xen/pvcalls: implement release command Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
[PATCH 06/18] xen/pvcalls: handle commands from the frontend Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
Re: [PATCH 06/18] xen/pvcalls: handle commands from the frontend Boris Ostrovsky <boris.ostrovsky@oracle.com> - 2017-05-16 04:20 +0200
Re: [PATCH 06/18] xen/pvcalls: handle commands from the frontend Stefano Stabellini <sstabellini@kernel.org> - 2017-05-16 23:00 +0200
[PATCH 07/18] xen/pvcalls: implement socket command Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
Re: [PATCH 07/18] xen/pvcalls: implement socket command Boris Ostrovsky <boris.ostrovsky@oracle.com> - 2017-05-16 04:30 +0200
Re: [PATCH 07/18] xen/pvcalls: implement socket command Stefano Stabellini <sstabellini@kernel.org> - 2017-05-16 22:50 +0200
[PATCH 12/18] xen/pvcalls: implement poll command Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
[PATCH 16/18] xen/pvcalls: implement read Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
[PATCH 08/18] xen/pvcalls: implement connect command Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
Re: [PATCH 08/18] xen/pvcalls: implement connect command Boris Ostrovsky <boris.ostrovsky@oracle.com> - 2017-05-16 04:50 +0200
Re: [PATCH 08/18] xen/pvcalls: implement connect command Stefano Stabellini <sstabellini@kernel.org> - 2017-05-16 23:10 +0200
Re: [PATCH 08/18] xen/pvcalls: implement connect command Boris Ostrovsky <boris.ostrovsky@oracle.com> - 2017-05-17 00:00 +0200
Re: [PATCH 08/18] xen/pvcalls: implement connect command Stefano Stabellini <sstabellini@kernel.org> - 2017-05-18 21:20 +0200
Re: [PATCH 08/18] xen/pvcalls: implement connect command Boris Ostrovsky <boris.ostrovsky@oracle.com> - 2017-05-18 22:30 +0200
[PATCH 11/18] xen/pvcalls: implement accept command Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:40 +0200
[PATCH 05/18] xen/pvcalls: connect to a frontend Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:50 +0200
Re: [PATCH 05/18] xen/pvcalls: connect to a frontend Boris Ostrovsky <boris.ostrovsky@oracle.com> - 2017-05-16 04:10 +0200
Re: [PATCH 05/18] xen/pvcalls: connect to a frontend Stefano Stabellini <sstabellini@kernel.org> - 2017-05-16 22:30 +0200
Re: [PATCH 05/18] xen/pvcalls: connect to a frontend Stefano Stabellini <sstabellini@kernel.org> - 2017-05-16 22:40 +0200
[PATCH 04/18] xen/pvcalls: xenbus state handling Stefano Stabellini <sstabellini@kernel.org> - 2017-05-15 22:50 +0200
Re: [PATCH 04/18] xen/pvcalls: xenbus state handling Boris Ostrovsky <boris.ostrovsky@oracle.com> - 2017-05-16 03:50 +0200
Re: [PATCH 04/18] xen/pvcalls: xenbus state handling Stefano Stabellini <sstabellini@kernel.org> - 2017-05-16 22:20 +0200
csiph-web