Re: [Xen-devel] [PATCH] evtchn: make support for different ABIs tunable

From: Jan Beulich <jbeulich@suse.com>
To: "Woodhouse, David" <dwmw@amazon.co.uk>
Cc: "sstabellini@kernel.org" <sstabellini@kernel.org>,
	"wl@xen.org" <wl@xen.org>,
	"konrad.wilk@oracle.com" <konrad.wilk@oracle.com>,
	"George.Dunlap@eu.citrix.com" <George.Dunlap@eu.citrix.com>,
	"andrew.cooper3@citrix.com" <andrew.cooper3@citrix.com>,
	"ian.jackson@eu.citrix.com" <ian.jackson@eu.citrix.com>,
	Eslam Elnikety <elnikety@amazon.com>, "tim@xen.org" <tim@xen.org>,
	"julien.grall@arm.com" <julien.grall@arm.com>,
	"anthony.perard@citrix.com" <anthony.perard@citrix.com>,
	"xen-devel@lists.xenproject.org" <xen-devel@lists.xenproject.org>
Subject: Re: [Xen-devel] [PATCH] evtchn: make support for different ABIs tunable
Date: Wed, 7 Aug 2019 14:18:00 +0200	[thread overview]
Message-ID: <cfdcb185-07f1-c99a-b231-8ec1e3074ebb@suse.com> (raw)
In-Reply-To: <48bf4b7991aa1bafb96dd30093e07593ce2b8d1f.camel@amazon.co.uk>

On 07.08.2019 14:04,  Woodhouse, David  wrote:
> On Wed, 2019-08-07 at 12:40 +0100, Andrew Cooper wrote:
>> On 07/08/2019 12:20, Eslam Elnikety wrote:
>>> Adding support for FIFO event channel ABI was first introduced in Xen 4.4
>>> (see 88910061ec6). Make this support tunable, since the choice of which
>>> event channel ABI has implications for hibernation. Consider resuming a
>>> pre Xen 4.4 hibernated Linux guest. The guest boot kernel defaults to FIFO
>>> ABI, whereas the resume kernel assumes 2L. This, in turn, results in Xen
>>> and the resumed kernel talking past each other (due to different protocols
>>> FIFO vs 2L).
>>
>> I'm afraid I don't follow.
>>
>> We have a Linux kernel which knows about FIFO, which was first booted on
>> Xen < 4.4, so configured 2L mode.
>>
>> It is then suspended, and resumed on a newer Xen >= 4.4.  The guest now
>> has a choice between 2L mode, and FIFO mode.
>>
>> What is the problem?
>>
>> When resuming, the guest in question should continue to use 2L mode,
>> because that is what it was using previously.
> 
> On resume, the guest first boots a Linux kernel (the 'boot' kernel).
> That kernel then loads the previous state (the 'resumed' kernel) from
> disk and then transfers control to it.
> 
> I believe the problem occurs when the boot kernel sees and enables FIFO
> mode, then transfers control to the resumed kernel which is expecting
> 2L.
> 
> I wonder if treating it more like a kexec and doing SHUTDOWN_soft_reset
> in the handover might be a viable long-term approach (and deal with
> other KASLR-related problems). Not that soft reset currently resets
> this AFAICT at a quick glance, but maybe it should?

Oh, definitely, if it doesn't already. Soft-reset should really do what
its name says and put state back to what it was initially.

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel