From: Dmitry Torokhov <dtor@vmware.com>
To: Greg KH <gregkh@linuxfoundation.org>
Cc: George Zhang <georgezhang@vmware.com>,
pv-drivers@vmware.com, linux-kernel@vger.kernel.org,
virtualization@lists.linux-foundation.org
Subject: Re: [Pv-drivers] [PATCH 01/12] VMCI: context implementation.
Date: Tue, 30 Oct 2012 08:56:47 -0700 [thread overview]
Message-ID: <6866362.eNDmFOn8qX@dtor-d630.eng.vmware.com> (raw)
In-Reply-To: <20121030154652.GB14167@kroah.com>
On Tuesday, October 30, 2012 08:46:52 AM Greg KH wrote:
> On Mon, Oct 29, 2012 at 09:01:40PM -0700, Dmitry Torokhov wrote:
> > Hi Greg,
> >
> > On Mon, Oct 29, 2012 at 07:10:58PM -0700, Greg KH wrote:
> > > On Mon, Oct 29, 2012 at 06:03:42PM -0700, George Zhang wrote:
> > > > +/*
> > > > + * Releases the VMCI context. If this is the last reference to
> > > > + * the context it will be deallocated. A context is created with
> > > > + * a reference count of one, and on destroy, it is removed from
> > > > + * the context list before its reference count is
> > > > + * decremented. Thus, if we reach zero, we are sure that nobody
> > > > + * else are about to increment it (they need the entry in the
> > > > + * context list for that). This function musn't be called with a
> > > > + * lock held.
> > > > + */
> > > > +void vmci_ctx_release(struct vmci_ctx *context)
> > > > +{
> > > > + ASSERT(context);
> > > > + kref_put(&context->kref, ctx_free_ctx);
> > > > +}
> > > > +
> > >
> > > Hm, are you _sure_ you should be calling this without a lock held?
> > > That's usually kref-101, you MUST hold a lock when calling put,
> > > otherwise you can race a kref_get() call, and all hell can break loose.
> > >
> > > Because of this, some saner people (like Al Viro), have suggested that I
> > > force the kref_put() and kref_get() calls pass in a spinlock just to
> > > enforce this.
> > >
> > > So, tell me what I'm missing here, and why you put the comment here
> > > saying that it really is supposed to be called without a lock held? How
> > > is that safe?
> >
> > Contexts are created/registered in vmci_ctx_init_ctx() and unregistered in
> > vmci_ctx_release_ctx() and these operations are protected by
> > ctx_list.lock spinlock. Context lookup (vmci_ctx_get) also uses spinlock
> > to traverse list of registered contexts and then grabs reference to the
> > [valid] context. The use of kref_put() without additional locking in
> > vmci_ctx_release() is fine as there is no chance of another thread
> > bumping count from 0 to 1.
>
> As I didn't see all callers of this holding that spinlock, it was
> confusing. You should put this type of description somewhere so that
> other reviewers don't have the same questions.
Fair enough, we'll add better comments to this code.
Thanks,
Dmitry
next prev parent reply other threads:[~2012-10-30 15:56 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-10-30 1:03 [PATCH 00/12] VMCI for Linux upstreaming George Zhang
2012-10-30 1:03 ` [PATCH 01/12] VMCI: context implementation George Zhang
2012-10-30 2:07 ` Greg KH
2012-10-30 2:10 ` Greg KH
2012-10-30 4:01 ` [Pv-drivers] " Dmitry Torokhov
2012-10-30 15:46 ` Greg KH
2012-10-30 15:56 ` Dmitry Torokhov [this message]
2012-10-30 1:03 ` [PATCH 02/12] VMCI: datagram implementation George Zhang
2012-10-30 1:04 ` [PATCH 03/12] VMCI: doorbell implementation George Zhang
2012-10-30 1:04 ` [PATCH 04/12] VMCI: device driver implementaton George Zhang
2012-10-30 2:21 ` Greg KH
2012-10-30 2:23 ` Greg KH
2012-10-30 4:15 ` [Pv-drivers] " Dmitry Torokhov
2012-10-30 15:49 ` Greg KH
2012-10-30 1:04 ` [PATCH 05/12] VMCI: event handling implementation George Zhang
2012-10-30 2:24 ` Greg KH
2012-10-30 4:58 ` [Pv-drivers] " Dmitry Torokhov
2012-10-30 15:50 ` Greg KH
2012-10-30 2:26 ` Greg KH
2012-10-30 5:01 ` [Pv-drivers] " Dmitry Torokhov
2012-10-30 15:50 ` Greg KH
2012-10-30 15:59 ` Dmitry Torokhov
2012-10-30 1:04 ` [PATCH 06/12] VMCI: handle array implementation George Zhang
2012-10-30 1:04 ` [PATCH 07/12] VMCI: queue pairs implementation George Zhang
2012-10-30 1:04 ` [PATCH 08/12] VMCI: resource object implementation George Zhang
2012-10-30 2:29 ` Greg KH
2012-10-30 5:20 ` [Pv-drivers] " Dmitry Torokhov
2012-10-30 15:51 ` Greg KH
2012-10-30 16:11 ` Dmitry Torokhov
2012-10-30 2:29 ` Greg KH
2012-10-30 5:21 ` [Pv-drivers] " Dmitry Torokhov
2012-10-30 1:05 ` [PATCH 09/12] VMCI: routing implementation George Zhang
2012-10-30 1:05 ` [PATCH 10/12] VMCI: guest side driver implementation George Zhang
2012-10-30 1:05 ` [PATCH 11/12] VMCI: host " George Zhang
2012-10-30 1:05 ` [PATCH 12/12] VMCI: Some header and config files George Zhang
2012-10-30 2:32 ` Greg KH
2012-10-30 5:22 ` [Pv-drivers] " Dmitry Torokhov
2012-10-30 2:38 ` Greg KH
2012-10-30 2:19 ` [PATCH 00/12] VMCI for Linux upstreaming Greg KH
2012-10-30 4:07 ` [Pv-drivers] " Dmitry Torokhov
2012-10-30 15:48 ` Greg KH
2012-10-30 16:18 ` Dmitry Torokhov
2012-10-30 16:27 ` Greg KH
2012-10-30 19:43 ` Dmitry Torokhov
2012-10-30 19:59 ` Greg KH
2012-11-21 20:31 George Zhang
2012-11-21 20:31 ` [PATCH 01/12] VMCI: context implementation George Zhang
2012-11-21 21:04 ` Joe Perches
2012-11-21 21:10 ` [Pv-drivers] " Andy King
2012-11-21 21:29 ` Dmitry Torokhov
2012-11-21 21:35 ` Joe Perches
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=6866362.eNDmFOn8qX@dtor-d630.eng.vmware.com \
--to=dtor@vmware.com \
--cc=georgezhang@vmware.com \
--cc=gregkh@linuxfoundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=pv-drivers@vmware.com \
--cc=virtualization@lists.linux-foundation.org \
/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).