Re: system hung up when offlining CPUs

From	Marc Zyngier <marc.zyngier@arm.com>
Newsgroups	linux.kernel
Subject	Re: system hung up when offlining CPUs
Date	2017-08-21 15:40 +0200
Message-ID	<ugV5T-4XE-3@gated-at.bofh.it> (permalink)
References	<ucimu-4kC-21@gated-at.bofh.it> <ucxER-6xm-13@gated-at.bofh.it> <ucEwG-30X-15@gated-at.bofh.it> <ucUi6-55n-25@gated-at.bofh.it> <ugUMy-4R2-25@gated-at.bofh.it>
Organization	ARM Ltd

Show all headers | View raw

On 21/08/17 14:18, Christoph Hellwig wrote:
> Can you try the patch below please?
> 
> ---
> From d5f59cb7a629de8439b318e1384660e6b56e7dd8 Mon Sep 17 00:00:00 2001
> From: Christoph Hellwig <hch@lst.de>
> Date: Mon, 21 Aug 2017 14:24:11 +0200
> Subject: virtio_pci: fix cpu affinity support
> 
> Commit 0b0f9dc5 ("Revert "virtio_pci: use shared interrupts for
> virtqueues"") removed the adjustment of the pre_vectors for the virtio
> MSI-X vector allocation which was added in commit fb5e31d9 ("virtio:
> allow drivers to request IRQ affinity when creating VQs"). This will
> lead to an incorrect assignment of MSI-X vectors, and potential
> deadlocks when offlining cpus.
> 
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> Fixes: 0b0f9dc5 ("Revert "virtio_pci: use shared interrupts for virtqueues")
> Reported-by: YASUAKI ISHIMATSU <yasu.isimatu@gmail.com>

Just gave it a go on an arm64 VM, and the behaviour seems much saner
(the virtio queue affinity now spans the whole system).

Tested-by: Marc Zyngier <marc.zyngier@arm.com>

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...

Back to linux.kernel | Previous | Next — Previous in thread | Find similar | Unroll thread

Thread

Re: system hung up when offlining CPUs Christoph Hellwig <hch@lst.de> - 2017-08-21 15:20 +0200
  Re: system hung up when offlining CPUs Marc Zyngier <marc.zyngier@arm.com> - 2017-08-21 15:40 +0200

csiph-web