From: Jan Kiszka <jan.kiszka@siemens.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: qemu-devel <qemu-devel@nongnu.org>, Liu Ping Fan <qemulist@gmail.com>
Subject: Re: [Qemu-devel] [PATCH v7 0/7] push mmio dispatch out of big lock
Date: Mon, 06 May 2013 14:06:50 +0200 [thread overview]
Message-ID: <51879CDA.3000800@siemens.com> (raw)
In-Reply-To: <51879844.1030101@redhat.com>
On 2013-05-06 13:47, Paolo Bonzini wrote:
> Il 06/05/2013 13:39, Jan Kiszka ha scritto:
>> On 2013-05-06 13:28, Paolo Bonzini wrote:
>>> Il 06/05/2013 13:11, Jan Kiszka ha scritto:
>>>> On 2013-05-06 12:58, Paolo Bonzini wrote:
>>>>> Il 06/05/2013 12:56, Jan Kiszka ha scritto:
>>>>>>> The problem is that even if I/O for a region is supposed to happen
>>>>>>> within the BQL, lookup can happen outside the BQL. Lookup will use the
>>>>>>> region even if it is just to discard it:
>>>>>>>
>>>>>>> VCPU thread (under BQL) device thread
>>>>>>> --------------------------------------------------------------------------------------
>>>>>>> flatview_ref
>>>>>>> memory_region_find returns d->mr
>>>>>>> memory_region_ref(d->mr) /* nop */
>>>>>>> qdev_free(d)
>>>>>>> object_unparent(d)
>>>>>>> unrealize(d)
>>>>>>> memory_region_del_subregion(d->mr)
>>>>>>> FlatView updated, d->mr not in the new view
>>>>>>>
>>>>>>> flatview_unref
>>>>>>> memory_region_unref(d->mr)
>>>>>>> object_unref(d)
>>>>>>> free(d)
>>>>>>> if (!d->mr->is_ram) { /* BAD! */
>>>>>>> memory_region_unref(d->mr) /* nop */
>>>>>>> return error
>>>>>>> }
>>>>>>>
>>>>>>>
>>>>>>> Here, the memory region is dereferenced *before* we know that it is BQL-free
>>>>>>> (in fact, exactly to ascertain whether it is BQL-free).
>>>>>>
>>>>>> Both flatview update and lookup *plus* locking type evaluation (i.e.
>>>>>> memory region dereferencing) always happen under the address space lock.
>>>>>> See Pingfan's patch.
>>>>>
>>>>> That's true of address_space_rw/map, but I don't think it holds for
>>>>> memory_region_find.
>>>>
>>>> It has to, or it would be broken: Either it is called on a region that
>>>> supports reference counting
>>>
>>> You cannot know that in advance, can you? The address is decided by the
>>> guest.
>>
>> Need to help me again to get the context: In which case is this a
>> hot-path that we want to keep BQL-free? Current users of
>> memory_region_find appear to be all relatively slow paths, thus are fine
>> with staying under BQL.
>
> virtio-blk-dataplane is basically redoing memory_region_find with a
> separate data structure, exactly so that it can run outside the BQL
> before we get BQL-free MMIO dispatch.
>
> I can try to post patches later today that actually use
> memory_region_find instead.
We could define its semantics as follows: return a reference to the
corresponding memory region, provide this is safe. A reference is safe when
- the region supports BQL-free operation (thus provides an owner to
apply reference counting on)
- the caller holds the BQL (check via qemu_mutex_iothread_is_locked()
- to be implemented)
The latter implies that the BQL is not dropped before returning the
reference, but that's nothing memory_region_find can enforce.
Jan
--
Siemens AG, Corporate Technology, CT RTC ITP SDP-DE
Corporate Competence Center Embedded Linux
next prev parent reply other threads:[~2013-05-06 12:06 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-11-25 2:02 [Qemu-devel] [PATCH v7 0/7] push mmio dispatch out of big lock Liu Ping Fan
2012-11-25 2:02 ` [Qemu-devel] [PATCH v7 1/7] qom: apply atomic on object's refcount Liu Ping Fan
2012-11-28 17:16 ` Richard Henderson
2012-11-29 8:35 ` liu ping fan
2012-11-25 2:02 ` [Qemu-devel] [PATCH v7 2/7] hotplug: introduce qdev_unplug_complete() to remove device from views Liu Ping Fan
2012-11-25 2:03 ` [Qemu-devel] [PATCH v7 3/7] pci: remove pci device from mem view when unplug Liu Ping Fan
2012-11-25 2:03 ` [Qemu-devel] [PATCH v7 4/7] memory: introduce local lock for address space Liu Ping Fan
2012-11-25 2:03 ` [Qemu-devel] [PATCH v7 5/7] memory: make mmio dispatch able to be out of biglock Liu Ping Fan
2013-05-06 11:21 ` Paolo Bonzini
2013-05-06 11:25 ` Jan Kiszka
2013-05-06 11:30 ` Paolo Bonzini
2012-11-25 2:03 ` [Qemu-devel] [PATCH v7 6/7] memory: introduce tls context to trace nested mmio request issue Liu Ping Fan
2012-11-25 2:03 ` [Qemu-devel] [PATCH v7 7/7] vcpu: push mmio dispatcher out of big lock Liu Ping Fan
2012-12-06 7:28 ` [Qemu-devel] [PATCH v7 0/7] push mmio dispatch " liu ping fan
2013-05-02 16:58 ` Jan Kiszka
2013-05-02 17:14 ` Jan Kiszka
2013-05-03 7:37 ` liu ping fan
2013-05-03 8:04 ` Jan Kiszka
2013-05-04 9:47 ` Paolo Bonzini
2013-05-04 10:42 ` Jan Kiszka
2013-05-06 8:07 ` Paolo Bonzini
2013-05-06 8:40 ` Jan Kiszka
2013-05-06 10:27 ` Paolo Bonzini
2013-05-06 10:56 ` Jan Kiszka
2013-05-06 10:58 ` Paolo Bonzini
2013-05-06 11:11 ` Jan Kiszka
2013-05-06 11:28 ` Paolo Bonzini
2013-05-06 11:39 ` Jan Kiszka
2013-05-06 11:47 ` Paolo Bonzini
2013-05-06 12:06 ` Jan Kiszka [this message]
2013-05-06 13:09 ` Paolo Bonzini
2013-05-06 14:05 ` Jan Kiszka
2013-05-06 14:28 ` Paolo Bonzini
2013-05-06 1:46 ` liu ping fan
2013-05-06 1:57 ` liu ping fan
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=51879CDA.3000800@siemens.com \
--to=jan.kiszka@siemens.com \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=qemulist@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 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).