[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <87sevwuxlw.fsf@mail.lhotse>
Date: Fri, 26 Jul 2024 22:49:31 +1000
From: Michael Ellerman <mpe@...erman.id.au>
To: Amit Machhiwal <amachhiw@...ux.ibm.com>, Bjorn Helgaas <helgaas@...nel.org>
Cc: Lizhi Hou <lizhi.hou@....com>, Rob Herring <robh@...nel.org>,
linux-pci@...r.kernel.org, linux-kernel@...r.kernel.org,
devicetree@...r.kernel.org, linuxppc-dev@...ts.ozlabs.org,
kvm-ppc@...r.kernel.org, Bjorn Helgaas <bhelgaas@...gle.com>, Saravana
Kannan <saravanak@...gle.com>, Vaibhav Jain <vaibhav@...ux.ibm.com>,
Nicholas Piggin <npiggin@...il.com>, Vaidyanathan Srinivasan
<svaidy@...ux.ibm.com>, Kowshik Jois B S <kowsjois@...ux.ibm.com>, Lukas
Wunner <lukas@...ner.de>
Subject: Re: [PATCH v2] PCI: Fix crash during pci_dev hot-unplug on pseries
KVM guest
Amit Machhiwal <amachhiw@...ux.ibm.com> writes:
> Hi Bjorn,
>
> On 2024/07/25 03:55 PM, Bjorn Helgaas wrote:
>> On Thu, Jul 25, 2024 at 11:15:39PM +0530, Amit Machhiwal wrote:
>> > ...
>> > The crash in question is a critical issue that we would want to have
>> > a fix for soon. And while this is still being figured out, is it
>> > okay to go with the fix I proposed in the V1 of this patch?
>>
>> v6.10 has been released already, and it will be a couple months before
>> the v6.11 release.
>>
>> It looks like the regression is 407d1a51921e, which appeared in v6.6,
>> almost a year ago, so it's fairly old.
>>
>> What target are you thinking about for the V1 patch? I guess if we
>> add it as a v6.11 post-merge window fix, it might get backported to
>> stable kernels before v6.11?
>
> Yes, I think we can go ahead with taking V1 patch for v6.11 post-merge window to
> fix the current bug and ask Ubuntu to pick it while Lizhi's proposed patch goes
> under test and review.
Lizhi's proposed patch (v3?) looks pretty small and straight forward.
It should be possible to get it tested and reviewed and merge it as a
fix during the v6.11-rc series.
Or if the CONFIG option is completely broken as Rob suggests then it
should just be forced off in Kconfig.
cheers
Powered by blists - more mailing lists