[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [RFC PATCH 04/13] ppc/spapr: Disable FWNMI capability if KVM does no
From: |
Nicholas Piggin |
Subject: |
Re: [RFC PATCH 04/13] ppc/spapr: Disable FWNMI capability if KVM does not support it |
Date: |
Thu, 05 Mar 2020 15:23:57 +1000 |
User-agent: |
astroid/0.15.0 (https://github.com/astroidmail/astroid) |
David Gibson's on March 5, 2020 2:37 pm:
> On Thu, Mar 05, 2020 at 01:51:40PM +1000, Nicholas Piggin wrote:
>> David Gibson's on March 4, 2020 3:57 pm:
>> > On Wed, Mar 04, 2020 at 03:28:41PM +1000, Nicholas Piggin wrote:
>> >> This is a fatal error if KVM does not support this capability, so try to
>> >> disable it if that's the case.
>> >
>> > Nack.
>> >
>> > Changing the guest visible environment based on host capabilities
>> > totally breaks the migration model. If the user has requested an
>> > environment the host can't provide, we have to outright fail instead.
>>
>> Okay. How are these capabilities requested, do you tend to default on
>> and pass the cap-flag=off parameter?
>
> I don't entirely understand the question. The user will use machine
> options of some sort to request the environment they want - plus we
> try to pick a good default for the base machine types.
>
> The cap-* options are one subset of that, with some inbuilt handling
> (which comes with some restrictions on how they operate). Above I was
> using "capability" in a broader sense, though.
Yeah okay, I think I answered my own question or fixed my
misunderstanding after going over your comments and reading the
code some more.
Thakns,
Nick
- [RFC PATCH 02/13] ppc/spapr: Add FWNMI System Reset state, (continued)
[RFC PATCH 05/13] ppc/spapr: Fix FWNMI machine check interrupt delivery, Nicholas Piggin, 2020/03/04
[RFC PATCH 06/13] ppc/spapr: Fix FWNMI machine check failure handling, Nicholas Piggin, 2020/03/04
[RFC PATCH 09/13] ppc/spapr: Don't kill the guest if a recovered FWNMI machine check delivery fails, Nicholas Piggin, 2020/03/04
[RFC PATCH 07/13] ppc/spapr: Improve FWNMI machine check delivery corner case comments, Nicholas Piggin, 2020/03/04
[RFC PATCH 08/13] ppc/spapr: Add FWNMI machine check delivery warnings, Nicholas Piggin, 2020/03/04
[RFC PATCH 10/13] target/ppc: allow ppc_cpu_do_system_reset to take an alternate vector, Nicholas Piggin, 2020/03/04
[RFC PATCH 12/13] ppc/spapr: Allow FWNMI on TCG, Nicholas Piggin, 2020/03/04