Re: [PATCH v3 4/7] xen/9pfs: connect to the backend

From: Stefano Stabellini
Date: Thu Mar 16 2017 - 14:03:52 EST


On Thu, 16 Mar 2017, Juergen Gross wrote:
> On 15/03/17 19:44, Stefano Stabellini wrote:
> > On Wed, 15 Mar 2017, Juergen Gross wrote:
> >> On 14/03/17 22:22, Stefano Stabellini wrote:
> >>> Hi Juergen,
> >>>
> >>> thank you for the review!
> >>>
> >>> On Tue, 14 Mar 2017, Juergen Gross wrote:
> >>>> On 14/03/17 00:50, Stefano Stabellini wrote:
> >>>>> Implement functions to handle the xenbus handshake. Upon connection,
> >>>>> allocate the rings according to the protocol specification.
> >>>>>
> >>>>> Initialize a work_struct and a wait_queue. The work_struct will be used
> >>>>> to schedule work upon receiving an event channel notification from the
> >>>>> backend. The wait_queue will be used to wait when the ring is full and
> >>>>> we need to send a new request.
> >>>>>
> >>>>> Signed-off-by: Stefano Stabellini <stefano@xxxxxxxxxxx>
> >>>>> CC: boris.ostrovsky@xxxxxxxxxx
> >>>>> CC: jgross@xxxxxxxx
> >>>>> CC: Eric Van Hensbergen <ericvh@xxxxxxxxx>
> >>>>> CC: Ron Minnich <rminnich@xxxxxxxxxx>
> >>>>> CC: Latchesar Ionkov <lucho@xxxxxxxxxx>
> >>>>> CC: v9fs-developer@xxxxxxxxxxxxxxxxxxxxx
> >>>>> ---
> >>
> >>>> Did you think about using request_threaded_irq() instead of a workqueue?
> >>>> For an example see e.g. drivers/scsi/xen-scsifront.c
> >>>
> >>> I like workqueues :-) It might come down to personal preferences, but I
> >>> think workqueues are more flexible and a better fit for this use case.
> >>> Not only it is easy to schedule work in a workqueue from the interrupt
> >>> handler, but also they can be used for sleeping in the request function
> >>> if there is not enough room on the ring. Besides, they can easily be
> >>> configured to share a single thread or to have multiple independent
> >>> threads.
> >>
> >> I'm fine with the workqueues as long as you have decided to use them
> >> considering the alternatives. :-)
> >>
> >>>> Can't you use xenbus_read_unsigned() instead of xenbus_read()?
> >>>
> >>> I can use xenbus_read_unsigned in the other cases below, but not here,
> >>> because versions is in the form: "1,3,4"
> >>
> >> Is this documented somewhere?
> >>
> >> Hmm, are any of the Xenstore entries documented? Shouldn't this be done
> >> in xen_9pfs.h ?
> >
> > They are documented in docs/misc/9pfs.markdown, under "Xenstore". Given
> > that it's all written there, especially the semantics, I didn't repeat
> > it in xen_9pfs.h
>
> Looking at it from the Linux kernel perspective this documentation is
> not really highly visible. For me it is okay, but there have been
> multiple examples in the past where documentation in the Xen repository
> wasn't regarded as being sufficient.
>
> I recommend moving the documentation regarding the interface into the
> header file like for the other pv interfaces.

What about adding a link such as:

http://xenbits.xen.org/gitweb/?p=xen.git;a=blob_plain;f=docs/misc/9pfs.markdown;hb=HEAD

that should be easily accessible, right? For other specifications, such
as 9p, only links are provided (see Documentation/filesystems/9p.txt).
I am suggesting a link, because then we are sure the specs don't go out
of sync. I realize that older PV protocols were described in header
files, but that was before Xen Project had a formal process for getting
new specifications accepted, and a formal place where to publish them.