qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Cam Macdonell <cam@cs.ualberta.ca>
To: Avi Kivity <avi@redhat.com>
Cc: qemu-devel@nongnu.org, kvm@vger.kernel.org
Subject: [Qemu-devel] Re: [PATCH] Inter-VM shared memory PCI device
Date: Wed, 10 Mar 2010 09:36:55 -0700	[thread overview]
Message-ID: <8286e4ee1003100836g35da5f07ye01fe272dee18dc2@mail.gmail.com> (raw)
In-Reply-To: <4B9764A9.9050003@redhat.com>

On Wed, Mar 10, 2010 at 2:21 AM, Avi Kivity <avi@redhat.com> wrote:
> On 03/09/2010 08:34 PM, Cam Macdonell wrote:
>>
>> On Tue, Mar 9, 2010 at 10:28 AM, Avi Kivity<avi@redhat.com>  wrote:
>>
>>>
>>> On 03/09/2010 05:27 PM, Cam Macdonell wrote:
>>>
>>>>
>>>>
>>>>>
>>>>>
>>>>>>
>>>>>>  Registers are used
>>>>>> for synchronization between guests sharing the same memory object when
>>>>>> interrupts are supported (this requires using the shared memory
>>>>>> server).
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>> How does the driver detect whether interrupts are supported or not?
>>>>>
>>>>>
>>>>
>>>> At the moment, the VM ID is set to -1 if interrupts aren't supported,
>>>> but that may not be the clearest way to do things.  With UIO is there
>>>> a way to detect if the interrupt pin is on?
>>>>
>>>>
>>>
>>> I suggest not designing the device to uio.  Make it a good
>>> guest-independent
>>> device, and if uio doesn't fit it, change it.
>>>
>>> Why not support interrupts unconditionally?  Is the device useful without
>>> interrupts?
>>>
>>
>> Currently my patch works with or without the shared memory server.  If
>> you give the parameter
>>
>> -ivshmem 256,foo
>>
>> then this will create (if necessary) and map /dev/shm/foo as the
>> shared region without interrupt support.  Some users of shared memory
>> are using it this way.
>>
>> Going forward we can require the shared memory server and always have
>> interrupts enabled.
>>
>
> Can you explain how they synchronize?  Polling?  Using the network?  Using
> it as a shared cache?
>
> If it's a reasonable use case it makes sense to keep it.
>

Do you mean how they synchronize without interrupts?  One project I've
been contacted about uses the shared region directly for
synchronization for simulations running in different VMs that share
data in the memory region.  In my tests spinlocks in the shared region
work between guests.

If we want to keep the serverless implementation, do we need to
support shm_open with -chardev somehow? Something like -chardev
shm,name=foo.  Right now my qdev implementation just passes the name
to the -device option and opens it.

> Another thing comes to mind - a shared memory ID, in case a guest has
> multiple cards.

Sure, a number that can be passed on the command-line and stored in a register?

Cam

  reply	other threads:[~2010-03-10 16:37 UTC|newest]

Thread overview: 48+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-03-05 23:52 [Qemu-devel] [PATCH] Support adding a file to qemu's ram allocation Cam Macdonell
2010-03-05 23:52 ` [Qemu-devel] [PATCH] Inter-VM shared memory PCI device Cam Macdonell
2010-03-07 22:53   ` Paul Brook
2010-03-08  1:45     ` Jamie Lokier
2010-03-08  9:48       ` Alexander Graf
2010-03-08  9:54         ` Jamie Lokier
2010-03-08 10:57           ` Alexander Graf
2010-03-09 21:41           ` Anthony Liguori
2010-03-08 13:04       ` Paul Brook
2010-03-09 19:00         ` Jamie Lokier
2010-03-08  9:52     ` Avi Kivity
2010-03-08 13:03       ` Paul Brook
2010-03-08 13:16         ` Avi Kivity
2010-03-09 20:11           ` Jamie Lokier
2010-03-09 21:44           ` Anthony Liguori
2010-03-10  9:25             ` Avi Kivity
2010-03-10 17:13               ` Anthony Liguori
2010-03-10 17:30                 ` Avi Kivity
2010-03-10 17:41                   ` Paul Brook
2010-03-11  6:33                     ` Avi Kivity
2010-03-11 12:21                       ` Paul Brook
2010-03-09 20:12         ` Jamie Lokier
2010-03-10  0:03           ` Paul Brook
2010-03-10  4:38             ` Cam Macdonell
2010-03-10  9:29               ` Avi Kivity
2010-03-10 11:13                 ` Paul Brook
2010-03-11  3:10             ` Jamie Lokier
2010-03-11  4:37               ` Nick Piggin
2010-03-11 14:38                 ` malc
2010-03-08  9:56   ` [Qemu-devel] " Avi Kivity
2010-03-08 17:57     ` Cam Macdonell
2010-03-09 10:29       ` Avi Kivity
     [not found]         ` <8286e4ee1003090724m1ef0b571g8b705a24e36e1753@mail.gmail.com>
2010-03-09 15:27           ` [Qemu-devel] " Cam Macdonell
2010-03-09 17:28             ` [Qemu-devel] " Avi Kivity
2010-03-09 17:34               ` Anthony Liguori
2010-03-09 18:34               ` Cam Macdonell
2010-03-10  9:21                 ` Avi Kivity
2010-03-10 16:36                   ` Cam Macdonell [this message]
2010-03-11  6:49                     ` Avi Kivity
2010-03-09 12:49       ` Arnd Bergmann
2010-03-09 13:03         ` Avi Kivity
2010-03-09 16:44           ` Cam Macdonell
2010-03-10 14:04             ` Arnd Bergmann
2010-03-11  6:50               ` Avi Kivity
2010-03-11 12:57                 ` Arnd Bergmann
2010-03-11 13:07                   ` Avi Kivity
2010-03-11 14:32                     ` Arnd Bergmann
2010-03-08  9:53 ` [Qemu-devel] Re: [PATCH] Support adding a file to qemu's ram allocation Avi Kivity

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=8286e4ee1003100836g35da5f07ye01fe272dee18dc2@mail.gmail.com \
    --to=cam@cs.ualberta.ca \
    --cc=avi@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=qemu-devel@nongnu.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).