From: Avi Kivity <avi@redhat.com>
To: Jan Kiszka <jan.kiszka@siemens.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>,
Marcelo Tosatti <mtosatti@redhat.com>,
liu ping fan <qemulist@gmail.com>,
Anthony Liguori <anthony@codemonkey.ws>,
"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>
Subject: Re: [Qemu-devel] [PATCH V3 10/11] vcpu: introduce lockmap
Date: Tue, 11 Sep 2012 15:30:30 +0300 [thread overview]
Message-ID: <504F2EE6.8060606@redhat.com> (raw)
In-Reply-To: <504F2DD6.8070807@siemens.com>
On 09/11/2012 03:25 PM, Jan Kiszka wrote:
>>>>
>>>> Most DMA today happens without the big qemu lock. We only need to
>>>> convert the paths that actually access memory, these are the block and
>>>> network layers (for the respective devices).
>>>
>>> ...and the devices themselves, of course.
>>
>> Right, for descriptors and stuff. So a guest can set a DMA address to
>> point at its own BAR, and cause qemu to deadlock.
>>
>> The problem is not limited to locking. A guest can also cause a write
>> to a BAR to initiate a synchronous write with the same address and data,
>> triggering infinite recursion.
>>
>> Perhaps one fix is the mythical DMA API, which will provide each DMA
>> initiator with its own view of memory (a private MemoryRegion root
>> node). Even that can be worked around with a pair of devices accessing
>> each other.
>
> Hmm, don't see an alternative to runtime loop detection yet.
See an expensive one on the other branch of this thread.
>
>>
>>>
>>>>
>>>>> The other option is to keep DMA requests issued by devices synchronous
>>>>> but let them fail if we are about to lock up. Still requires changes,
>>>>> but is probably more comprehensible for device model developers.
>>>>
>>>> How do you handle failures?
>>>
>>> By not sending a network frame or dropping an incoming one, e.g., and
>>> signaling this in a device specific way.
>>
>> Doesn't work for block devices.
>
> Because the block layer API cannot report errors to the devices? What
> happens if there is a real I/O error?
We report real I/O errors. But if we report a transient error due to
some lock being taken as an I/O error, the guest will take unwarranted
action.
If the errors are not expected in normal operation (we can avoid them if
all DMA is to real RAM) then this is an acceptable solution. Still it
generates a lot of rarely used code paths and so isn't very good for
security.
--
error compiling committee.c: too many arguments to function
next prev parent reply other threads:[~2012-09-11 12:30 UTC|newest]
Thread overview: 72+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-09-11 7:51 [Qemu-devel] [PATCH V3 0/10] prepare unplug out of protection of global lock Liu Ping Fan
2012-09-11 7:51 ` [Qemu-devel] [PATCH V3 01/11] atomic: introduce atomic operations Liu Ping Fan
2012-09-11 8:04 ` Avi Kivity
2012-09-13 6:54 ` liu ping fan
2012-09-13 8:14 ` Avi Kivity
2012-09-13 8:19 ` Paolo Bonzini
2012-09-13 8:23 ` Avi Kivity
2012-09-13 8:29 ` Paolo Bonzini
2012-09-13 8:45 ` liu ping fan
2012-09-19 13:16 ` Jamie Lokier
2012-09-19 13:32 ` Jamie Lokier
2012-09-19 14:12 ` Peter Maydell
2012-09-19 15:53 ` Jamie Lokier
2012-09-11 8:15 ` Peter Maydell
2012-09-13 6:53 ` liu ping fan
2012-09-11 7:51 ` [Qemu-devel] [PATCH V3 02/11] qom: apply atomic on object's refcount Liu Ping Fan
2012-09-11 7:51 ` [Qemu-devel] [PATCH V3 03/11] hotplug: introduce qdev_unplug_complete() to remove device from views Liu Ping Fan
2012-09-11 7:51 ` [Qemu-devel] [PATCH V3 04/11] pci: remove pci device from mem view when unplug Liu Ping Fan
2012-09-11 7:51 ` [Qemu-devel] [PATCH V3 05/11] memory: introduce ref, unref interface for MemoryRegionOps Liu Ping Fan
2012-09-11 8:08 ` Avi Kivity
2012-09-11 7:51 ` [Qemu-devel] [PATCH V3 06/11] memory: make mmio dispatch able to be out of biglock Liu Ping Fan
2012-09-11 8:25 ` Avi Kivity
2012-09-11 8:47 ` Avi Kivity
2012-09-11 7:51 ` [Qemu-devel] [PATCH V3 07/11] memory: implement e1000's MemoryRegionOps ref/unref Liu Ping Fan
2012-09-11 8:37 ` Avi Kivity
2012-09-11 7:51 ` [Qemu-devel] [PATCH V3 08/11] qom: introduce reclaimer to release obj in async Liu Ping Fan
2012-09-11 8:32 ` Avi Kivity
2012-09-11 9:32 ` liu ping fan
2012-09-11 9:37 ` Avi Kivity
2012-09-13 6:54 ` liu ping fan
2012-09-13 8:45 ` Avi Kivity
2012-09-13 9:59 ` liu ping fan
2012-09-13 10:09 ` Avi Kivity
2012-09-11 7:51 ` [Qemu-devel] [PATCH V3 09/11] vcpu: make QemuThread as tls to store thread-self info Liu Ping Fan
2012-09-11 7:51 ` [Qemu-devel] [PATCH V3 10/11] vcpu: introduce lockmap Liu Ping Fan
2012-09-11 8:35 ` Avi Kivity
2012-09-11 9:44 ` liu ping fan
2012-09-11 9:54 ` Avi Kivity
2012-09-11 10:04 ` Jan Kiszka
2012-09-11 11:03 ` Avi Kivity
2012-09-11 11:08 ` Jan Kiszka
2012-09-11 12:20 ` Avi Kivity
2012-09-11 12:25 ` Jan Kiszka
2012-09-11 12:30 ` Avi Kivity [this message]
2012-09-11 12:35 ` Jan Kiszka
2012-09-11 12:39 ` Avi Kivity
2012-09-19 4:25 ` Peter Crosthwaite
2012-09-19 4:32 ` Edgar E. Iglesias
2012-09-19 4:40 ` Peter Crosthwaite
2012-09-19 7:55 ` Avi Kivity
2012-09-19 11:46 ` Edgar E. Iglesias
2012-09-19 12:12 ` Avi Kivity
2012-09-19 12:17 ` Edgar E. Iglesias
2012-09-19 13:01 ` Igor Mitsyanko
2012-09-19 13:03 ` Avi Kivity
2012-09-19 7:57 ` Jan Kiszka
2012-09-19 13:07 ` Igor Mitsyanko
2012-09-11 9:57 ` Jan Kiszka
2012-09-11 12:24 ` Avi Kivity
2012-09-11 12:41 ` Avi Kivity
2012-09-11 14:54 ` Marcelo Tosatti
2012-09-13 6:55 ` liu ping fan
2012-09-13 6:55 ` liu ping fan
2012-09-13 8:19 ` Avi Kivity
2012-09-17 2:24 ` liu ping fan
2012-09-19 8:01 ` Avi Kivity
2012-09-19 8:36 ` liu ping fan
2012-09-19 9:05 ` Avi Kivity
2012-09-20 7:51 ` liu ping fan
2012-09-20 9:15 ` Avi Kivity
2012-09-21 7:27 ` liu ping fan
2012-09-11 7:51 ` [Qemu-devel] [PATCH V3 11/11] vcpu: push mmio dispatcher out of big lock 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=504F2EE6.8060606@redhat.com \
--to=avi@redhat.com \
--cc=anthony@codemonkey.ws \
--cc=jan.kiszka@siemens.com \
--cc=mtosatti@redhat.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).