| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <alpine.DEB.2.10.1703071650320.17906@sstabellini-ThinkPad-X260>
Date: Tue, 7 Mar 2017 16:55:38 -0800 (PST)
From: Stefano Stabellini <sstabellini@...nel.org>
To: Boris Ostrovsky <boris.ostrovsky@...cle.com>
cc: Stefano Stabellini <sstabellini@...nel.org>,
xen-devel@...ts.xenproject.org, linux-kernel@...r.kernel.org,
Stefano Stabellini <stefano@...reto.com>, jgross@...e.com,
Eric Van Hensbergen <ericvh@...il.com>,
Ron Minnich <rminnich@...dia.gov>,
Latchesar Ionkov <lucho@...kov.net>,
v9fs-developer@...ts.sourceforge.net
Subject: Re: [PATCH 5/7] xen/9pfs: send requests to the backend
On Tue, 7 Mar 2017, Boris Ostrovsky wrote:
> On 03/06/2017 03:01 PM, Stefano Stabellini wrote:
> > Implement struct p9_trans_module create and close functions by looking
> > at the available Xen 9pfs frontend-backend connections. We don't expect
> > many frontend-backend connections, thus walking a list is OK.
> >
> > Send requests to the backend by copying each request to one of the
> > available rings (each frontend-backend connection comes with multiple
> > rings). Handle the ring and notifications following the 9pfs
> > specification. If there are not enough free bytes on the ring for the
> > request, wait on the wait_queue: the backend will send a notification
> > after consuming more requests.
> >
> > Signed-off-by: Stefano Stabellini <stefano@...reto.com>
> > CC: boris.ostrovsky@...cle.com
> > CC: jgross@...e.com
> > CC: Eric Van Hensbergen <ericvh@...il.com>
> > CC: Ron Minnich <rminnich@...dia.gov>
> > CC: Latchesar Ionkov <lucho@...kov.net>
> > CC: v9fs-developer@...ts.sourceforge.net
> > ---
> > net/9p/trans_xen.c | 83 +++++++++++++++++++++++++++++++++++++++++++++++++++++-
> > 1 file changed, 82 insertions(+), 1 deletion(-)
> >
> > diff --git a/net/9p/trans_xen.c b/net/9p/trans_xen.c
> > index 9f6cf8d..4e26556 100644
> > --- a/net/9p/trans_xen.c
> > +++ b/net/9p/trans_xen.c
> > @@ -47,22 +47,103 @@ struct xen_9pfs_front_priv {
> > };
> > static LIST_HEAD(xen_9pfs_devs);
> >
> > +/* We don't currently allow canceling of requests */
> > static int p9_xen_cancel(struct p9_client *client, struct p9_req_t *req)
> > {
> > - return 0;
> > + return 1;
> > }
> >
> > static int p9_xen_create(struct p9_client *client, const char *addr, char *args)
> > {
> > + struct xen_9pfs_front_priv *priv = NULL;
> > +
> > + list_for_each_entry(priv, &xen_9pfs_devs, list) {
> > + if (!strcmp(priv->tag, addr))
> > + break;
> > + }
>
>
> You could simplify this (and p9_xen_close()) but assigning client and
> returning from inside the 'if' statement.
I'll do that.
> I am also not sure you need to initialize priv.
With the new changes, I won't need to.
> > + if (!priv || strcmp(priv->tag, addr))
> > + return -EINVAL;
> > +
> > + priv->client = client;
> > return 0;
> > }
> >
> > static void p9_xen_close(struct p9_client *client)
> > {
> > + struct xen_9pfs_front_priv *priv = NULL;
> > +
> > + list_for_each_entry(priv, &xen_9pfs_devs, list) {
> > + if (priv->client == client)
> > + break;
> > + }
> > + if (!priv || priv->client != client)
> > + return;
> > +
> > + priv->client = NULL;
> > + return;
> > +}
> > +
> > +static int p9_xen_write_todo(struct xen_9pfs_dataring *ring, RING_IDX size)
> > +{
> > + RING_IDX cons, prod;
> > +
> > + cons = ring->intf->out_cons;
> > + prod = ring->intf->out_prod;
> > + mb();
> > +
> > + if (XEN_9PFS_RING_SIZE - xen_9pfs_queued(prod, cons, XEN_9PFS_RING_SIZE) >= size)
> > + return 1;
> > + else
> > + return 0;
> > }
> >
> > static int p9_xen_request(struct p9_client *client, struct p9_req_t *p9_req)
> > {
> > + struct xen_9pfs_front_priv *priv = NULL;
> > + RING_IDX cons, prod, masked_cons, masked_prod;
> > + unsigned long flags;
> > + uint32_t size = p9_req->tc->size;
> > + struct xen_9pfs_dataring *ring;
> > + int num;
> > +
> > + list_for_each_entry(priv, &xen_9pfs_devs, list) {
> > + if (priv->client == client)
> > + break;
> > + }
> > + if (priv == NULL || priv->client != client)
> > + return -EINVAL;
> > +
> > + num = p9_req->tc->tag % priv->num_rings;
> > + ring = &priv->rings[num];
> > +
> > +again:
> > + while (wait_event_interruptible(ring->wq,
> > + p9_xen_write_todo(ring, size) > 0) != 0);
> > +
> > + spin_lock_irqsave(&ring->lock, flags);
> > + cons = ring->intf->out_cons;
> > + prod = ring->intf->out_prod;
> > + mb();
> > +
> > + if (XEN_9PFS_RING_SIZE - xen_9pfs_queued(prod, cons, XEN_9PFS_RING_SIZE) < size) {
>
>
> This looks like p9_xen_write_todo().
p9_xen_write_todo is just a wrapper around xen_9pfs_queued to provide
a return value that works well with wait_event_interruptible.
I would prefer not to call p9_xen_write_todo here, because it's simpler
if we don't read prod and cons twice.
> BTW, where is xen_9pfs_queued()
> defined? I couldn't find it. Same for xen_9pfs_mask() and
> xen_9pfs_write_packet().
They are provided by the new ring macros, see
include/xen/interface/io/ring.h (the first patch).
> > + spin_unlock_irqrestore(&ring->lock, flags);
> > + goto again;
> > + }
> > +
> > + masked_prod = xen_9pfs_mask(prod, XEN_9PFS_RING_SIZE);
> > + masked_cons = xen_9pfs_mask(cons, XEN_9PFS_RING_SIZE);
> > +
> > + xen_9pfs_write_packet(ring->ring.out,
> > + &masked_prod, masked_cons,
> > + XEN_9PFS_RING_SIZE, p9_req->tc->sdata, size);
> > +
> > + p9_req->status = REQ_STATUS_SENT;
> > + wmb(); /* write ring before updating pointer */
> > + prod += size;
> > + ring->intf->out_prod = prod;
> > + spin_unlock_irqrestore(&ring->lock, flags);
> > + notify_remote_via_irq(ring->irq);
> > +
> > return 0;
> > }
> >
> >
>
Powered by blists - more mailing lists