All of lore.kernel.org
 help / color / mirror / Atom feed
From: liu ping fan <qemulist@gmail.com>
To: Avi Kivity <avi@redhat.com>
Cc: Peter Maydell <peter.maydell@linaro.org>,
	Stefan Hajnoczi <stefanha@gmail.com>,
	Marcelo Tosatti <mtosatti@redhat.com>,
	"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>,
	Anthony Liguori <anthony@codemonkey.ws>,
	Jan Kiszka <jan.kiszka@siemens.com>,
	Paolo Bonzini <pbonzini@redhat.com>
Subject: Re: [Qemu-devel] [patch v4 13/16] e1000: add busy flag to anti broken device state
Date: Fri, 26 Oct 2012 11:05:06 +0800	[thread overview]
Message-ID: <CAJnKYQntJZBJjwCMj7T0rxKgGP51A+ehmyNc=O3RBCicmHD-kg@mail.gmail.com> (raw)
In-Reply-To: <50890087.1080706@redhat.com>

On Thu, Oct 25, 2012 at 5:04 PM, Avi Kivity <avi@redhat.com> wrote:
> On 10/25/2012 11:00 AM, Peter Maydell wrote:
>> On 23 October 2012 10:37, Avi Kivity <avi@redhat.com> wrote:
>>> On 10/23/2012 11:32 AM, liu ping fan wrote:
>>>> On Tue, Oct 23, 2012 at 5:07 PM, Jan Kiszka <jan.kiszka@siemens.com> wrote:
>>>>> On 2012-10-23 07:52, liu ping fan wrote:
>>>>>> On Mon, Oct 22, 2012 at 6:40 PM, Avi Kivity <avi@redhat.com> wrote:
>>>>>>> On 10/22/2012 11:23 AM, Liu Ping Fan wrote:
>>>>>> It will only record and fix the issue on one thread. But guest can
>>>>>> touch the emulated device on muti-threads.
>>>>>
>>>>> Sorry, what does that mean? A second VCPU accessing the device will
>>>>> simply be ignored when it races with another VCPU? Specifically
>>>>>
>>>> Yes, just ignored.  For device which support many logic in parallel,
>>>> it should use independent busy flag for each logic
>>>
>>> We don't actually know that e1000 doesn't.  Why won't writing into
>>> different registers in parallel work?
>>
>> Unless the device we're emulating supports multiple in
>> parallel accesses (and I bet 99.9% of the devices we model
>> don't) then the memory framework needs to serialise the
>> loads/stores. Otherwise it's just going to be excessively
>> hard to write a reliable device model.
>
> That's why we have a per-device lock.  The busy flag breaks that model
> by discarding accesses that occur in parallel.
>
I think by adopting the model, we can avoid this.

struct device_logic {
  bool busy;
  qemu_mutex lock;
  QemuCond wait;
};

LOCK(logic->lock)
while (logic->busy) {
qemu_cond_wait(&logic->wait, &logic->lock);
}
....
do hardware emulation
...
logic->busy = false;
qemu_cond_signal(&logic->wait);

This is identical to the biglock's behavior for parallel access to
device for nowadays. And then, the problem left is what level for
parallel we want. If we expect more parallel, we need to degrade the
device into more logic unit.


Regards,
pingfan
>
> --
> error compiling committee.c: too many arguments to function

  reply	other threads:[~2012-10-26  3:05 UTC|newest]

Thread overview: 102+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-10-22  9:23 [Qemu-devel] [patch v4 00/16] push mmio dispatch out of big lock Liu Ping Fan
2012-10-22  9:23 ` [Qemu-devel] [patch v4 01/16] atomic: introduce atomic operations Liu Ping Fan
2012-10-22  9:23 ` [Qemu-devel] [patch v4 02/16] qom: apply atomic on object's refcount Liu Ping Fan
2012-10-22  9:23 ` [Qemu-devel] [patch v4 03/16] hotplug: introduce qdev_unplug_complete() to remove device from views Liu Ping Fan
2012-10-22  9:23 ` [Qemu-devel] [patch v4 04/16] pci: remove pci device from mem view when unplug Liu Ping Fan
2012-10-22  9:23 ` [Qemu-devel] [patch v4 05/16] memory: introduce ref, unref interface for MemoryRegionOps Liu Ping Fan
2012-10-22  9:38   ` Avi Kivity
2012-10-23 11:51     ` Paolo Bonzini
2012-10-23 11:55       ` Avi Kivity
2012-10-23 11:57         ` Paolo Bonzini
2012-10-23 12:02           ` Avi Kivity
2012-10-23 12:06             ` Paolo Bonzini
2012-10-23 12:15               ` Avi Kivity
2012-10-23 12:32                 ` Paolo Bonzini
2012-10-23 14:49                   ` Avi Kivity
2012-10-23 15:26                     ` Paolo Bonzini
2012-10-23 16:09                       ` Avi Kivity
2012-10-24  7:29                         ` Paolo Bonzini
2012-10-25 16:28                           ` Avi Kivity
2012-10-26 15:05                             ` Paolo Bonzini
2012-10-23 12:04         ` Jan Kiszka
2012-10-23 12:12           ` Paolo Bonzini
2012-10-23 12:16             ` Jan Kiszka
2012-10-23 12:28               ` Avi Kivity
2012-10-23 12:40                 ` Jan Kiszka
2012-10-23 14:37                   ` Avi Kivity
2012-10-22  9:23 ` [Qemu-devel] [patch v4 06/16] memory: document ref, unref interface Liu Ping Fan
2012-10-22  9:23 ` [Qemu-devel] [patch v4 07/16] memory: make mmio dispatch able to be out of biglock Liu Ping Fan
2012-10-23 12:12   ` Jan Kiszka
2012-10-23 12:36     ` Avi Kivity
2012-10-24  6:31       ` liu ping fan
2012-10-24  6:56         ` liu ping fan
2012-10-25  8:57           ` Avi Kivity
2012-10-22  9:23 ` [Qemu-devel] [patch v4 08/16] QemuThread: make QemuThread as tls to store extra info Liu Ping Fan
2012-10-22  9:30   ` Jan Kiszka
2012-10-22 17:13     ` Peter Maydell
2012-10-23  5:58       ` liu ping fan
2012-10-23 11:48       ` Paolo Bonzini
2012-10-23 11:50         ` Peter Maydell
2012-10-23 11:51           ` Jan Kiszka
2012-10-23 12:00           ` Paolo Bonzini
2012-10-23 12:27             ` Peter Maydell
2012-11-18 10:02             ` Brad Smith
2012-11-18 16:14               ` Paolo Bonzini
2012-11-18 16:15                 ` Paolo Bonzini
2012-10-22  9:23 ` [Qemu-devel] [patch v4 09/16] memory: introduce mmio request pending to anti nested DMA Liu Ping Fan
2012-10-22 10:28   ` Avi Kivity
2012-10-23 12:38   ` Gleb Natapov
2012-10-24  6:31     ` liu ping fan
2012-10-22  9:23 ` [Qemu-devel] [patch v4 10/16] memory: introduce lock ops for MemoryRegionOps Liu Ping Fan
2012-10-22 10:30   ` Avi Kivity
2012-10-23  5:53     ` liu ping fan
2012-10-23  8:53       ` Jan Kiszka
2012-10-22  9:23 ` [Qemu-devel] [patch v4 11/16] vcpu: push mmio dispatcher out of big lock Liu Ping Fan
2012-10-22 10:31   ` Avi Kivity
2012-10-22 10:36     ` Jan Kiszka
2012-10-22  9:23 ` [Qemu-devel] [patch v4 12/16] e1000: apply fine lock on e1000 Liu Ping Fan
2012-10-22 10:37   ` Avi Kivity
2012-10-23  9:04   ` Jan Kiszka
2012-10-24  6:31     ` liu ping fan
2012-10-24  7:17       ` Jan Kiszka
2012-10-25  9:01         ` Avi Kivity
2012-10-25  9:31           ` Jan Kiszka
2012-10-25 16:21             ` Avi Kivity
2012-10-25 16:39               ` Jan Kiszka
2012-10-25 17:02                 ` Avi Kivity
2012-10-25 18:48                   ` Jan Kiszka
2012-10-29  5:24                     ` liu ping fan
2012-10-24  7:29     ` liu ping fan
2012-10-25 13:34       ` Jan Kiszka
2012-10-25 16:23         ` Avi Kivity
2012-10-25 16:41           ` Jan Kiszka
2012-10-25 17:03             ` Avi Kivity
2012-10-29  5:24         ` liu ping fan
2012-10-31  7:03           ` Jan Kiszka
2012-10-22  9:23 ` [Qemu-devel] [patch v4 13/16] e1000: add busy flag to anti broken device state Liu Ping Fan
2012-10-22 10:40   ` Avi Kivity
2012-10-23  5:52     ` liu ping fan
2012-10-23  9:06       ` Avi Kivity
2012-10-23  9:07       ` Jan Kiszka
2012-10-23  9:32         ` liu ping fan
2012-10-23  9:37           ` Avi Kivity
2012-10-24  6:36             ` liu ping fan
2012-10-25  8:55               ` Avi Kivity
2012-10-25  9:00             ` Peter Maydell
2012-10-25  9:04               ` Avi Kivity
2012-10-26  3:05                 ` liu ping fan [this message]
2012-10-26  3:08                   ` liu ping fan
2012-10-26 10:25                     ` Jan Kiszka
2012-10-29  5:24                       ` liu ping fan
2012-10-29  7:50                         ` Peter Maydell
2012-10-22  9:23 ` [Qemu-devel] [patch v4 14/16] qdev: introduce stopping state Liu Ping Fan
2012-10-22  9:23 ` [Qemu-devel] [patch v4 15/16] e1000: introduce unmap() to fix unplug issue Liu Ping Fan
2012-10-22  9:23 ` [Qemu-devel] [patch v4 16/16] e1000: implement MemoryRegionOps's ref&lock interface Liu Ping Fan
2012-10-25 14:04 ` [Qemu-devel] [patch v4 00/16] push mmio dispatch out of big lock Peter Maydell
2012-10-25 16:44   ` Jan Kiszka
2012-10-25 17:07   ` Avi Kivity
2012-10-25 17:13     ` Peter Maydell
2012-10-25 18:13       ` Marcelo Tosatti
2012-10-25 19:00         ` Jan Kiszka
2012-10-25 19:06           ` Peter Maydell
2012-10-29 15:24       ` 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='CAJnKYQntJZBJjwCMj7T0rxKgGP51A+ehmyNc=O3RBCicmHD-kg@mail.gmail.com' \
    --to=qemulist@gmail.com \
    --cc=anthony@codemonkey.ws \
    --cc=avi@redhat.com \
    --cc=jan.kiszka@siemens.com \
    --cc=mtosatti@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-devel@nongnu.org \
    --cc=stefanha@gmail.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.