linux-kernel - Re: vhost changes (batched) in linux-next after 12/13 trigger random crashes in KVM guests after reboot

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <2a63b15f-8cf5-5868-550c-42e2cfd92c60@de.ibm.com>
Date:   Wed, 22 Jan 2020 20:32:21 +0100
From:   Christian Borntraeger <borntraeger@...ibm.com>
To:     "Michael S. Tsirkin" <mst@...hat.com>
Cc:     "virtualization@...ts.linux-foundation.org" 
        <virtualization@...ts.linux-foundation.org>,
        Stephen Rothwell <sfr@...b.auug.org.au>,
        Linux Next Mailing List <linux-next@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        kvm list <kvm@...r.kernel.org>,
        Halil Pasic <pasic@...ux.ibm.com>
Subject: Re: vhost changes (batched) in linux-next after 12/13 trigger random
 crashes in KVM guests after reboot



On 20.01.20 07:27, Michael S. Tsirkin wrote:
> On Tue, Jan 07, 2020 at 01:16:50PM +0100, Christian Borntraeger wrote:
>> On 07.01.20 12:55, Michael S. Tsirkin wrote:
>>
>>>
>>> I pushed batched-v3 - same head but bisect should work now.
>>>
>>
>> With 
>> commit 38ced0208491103b50f1056f0d1c8f28e2e13d08 (HEAD)
>> Author:     Michael S. Tsirkin <mst@...hat.com>
>> AuthorDate: Wed Dec 11 12:19:26 2019 -0500
>> Commit:     Michael S. Tsirkin <mst@...hat.com>
>> CommitDate: Tue Jan 7 06:52:42 2020 -0500
>>
>>     vhost: use batched version by default
>>
>>
>> I have exactly one successful ping and then the network inside the guest is broken (no packet
>> anymore).
> 
> Does anything appear in host's dmesg when this happens?

I think there was nothing, but I am not sure. I would need to redo the test if this is important to know.

> 
> 
>> So you could consider this commit broken (but in a different way and also without any
>> guest reboot necessary).
>>
>>
>> bisect log:
>> git bisect start
>> # bad: [d2f6175f52062ee51ee69754a6925608213475d2] vhost: use vhost_desc instead of vhost_log
>> git bisect bad d2f6175f52062ee51ee69754a6925608213475d2
>> # good: [d1281e3a562ec6a08f944a876481dd043ba739b9] virtio-blk: remove VIRTIO_BLK_F_SCSI support
>> git bisect good d1281e3a562ec6a08f944a876481dd043ba739b9
>> # good: [fac7c0f46996e32d996f5c46121df24a6b95ec3b] vhost: option to fetch descriptors through an independent struct
>> git bisect good fac7c0f46996e32d996f5c46121df24a6b95ec3b
>> # bad: [539eb9d738f048cd7be61f404e8f9c7d9d2ff3cc] vhost: batching fetches
>> git bisect bad 539eb9d738f048cd7be61f404e8f9c7d9d2ff3cc
>