From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.133.124]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 0C2E5819 for ; Tue, 18 Feb 2025 00:12:41 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=170.10.133.124 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1739837564; cv=none; b=JWzspElYSrtxP6k8PRDHUDz53Ug7uJL/GQxaNTdnN3Arqgodc2WtObEfurmhQ++cET+xWXWiG0ox/LuhGN39PH4sre1uGFKFNDCOT+fnysz4+on7UFseiP+jrDvJcKR/Rrl2NhCVx2hSD9zjZz9jMl42vp5R4DpPua/DFNkBlOM= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1739837564; c=relaxed/simple; bh=uC8cKZgH5jtMxAMxHIrCMiEUZs5Vp6M4wFq3dcuatR4=; h=Date:From:To:Cc:Subject:Message-ID:References:MIME-Version: In-Reply-To:Content-Type:Content-Disposition; b=BfoSkA4zZkyxZ67+EpVm1hNkrX+/HJ1IGXR+rpBMUFW5LZpKY/+ABTTC998E2KQczWVLUCUD/H5LLCS9m2CuAwYJxxKZF7emdQBp3sQYc6AHBRLTmQP4fgdpdDKc7eo8z85SvCOqeSn3Si8Q8vZy/jNcXvoNELUR90EijqYeAt8= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com; spf=pass smtp.mailfrom=redhat.com; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b=TZ2nxg7M; arc=none smtp.client-ip=170.10.133.124 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=redhat.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="TZ2nxg7M" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1739837560; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=B02gyojDUiuYtKrTl8LJtiDNK23qLyYF7c8avrsDEN4=; b=TZ2nxg7MC1XHdHarv2RLtVzKPor0oMukXir+T47sCDhJBPlCJhOJPNw1L8g3sMgTYcrnnu oQdMAXeHH4D2uSB0BujxJdbqa3LkjtL0/C4/qyD7Tq4fZEMn5wLOfP5pbKytZNTh1RODgh 4og43UOMybXCX3gEa28M2cf9ZVdHQjQ= Received: from mail-wm1-f70.google.com (mail-wm1-f70.google.com [209.85.128.70]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-686-aU1Sb-agOpabASKGlLYo0Q-1; Mon, 17 Feb 2025 19:12:39 -0500 X-MC-Unique: aU1Sb-agOpabASKGlLYo0Q-1 X-Mimecast-MFC-AGG-ID: aU1Sb-agOpabASKGlLYo0Q_1739837558 Received: by mail-wm1-f70.google.com with SMTP id 5b1f17b1804b1-43961fb0bafso24918125e9.2 for ; Mon, 17 Feb 2025 16:12:39 -0800 (PST) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1739837558; x=1740442358; h=in-reply-to:content-transfer-encoding:content-disposition :mime-version:references:message-id:subject:cc:to:from:date :x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=B02gyojDUiuYtKrTl8LJtiDNK23qLyYF7c8avrsDEN4=; b=Gc6NfcujuBr7Ti6xTaXwubKqbWkkweGN+BBFdQbynYvrI8WRRsXRjuaimgDo0Vz3Fa 6qSHJ2ATKHi6jpd2FzTd6jbsEmE1jxHWJH+CMbe/WlmWlyFJWAe5hDRiEobXZpF420iX t/pWiVCfmp74wklq+nZZlgaUsen/UnFgnZuGmx4VS0iAwG6e7voB/RsaKNXvDC+VG2Ki AcqN9RuybGWKt+4DfWzlsE28v4BFUa2Dm5L0hS+QHzIi36hsvgxbQ9GFwHks+JiIanEY n+x+MALz5sQfSRua92S+LBpHvVjqUzjQ4awaou7wuSVLrDGe/Ij2jRD/ZrLN2gO6qfiF KLOQ== X-Forwarded-Encrypted: i=1; AJvYcCW5zdFWVuJLRuTQ0CChcrSPCfEPpP2Ymln/18VE402QUv6hI8xhEJvaXklj8lTf/8AT57XvcqZoRnleSOUpIw==@lists.linux.dev X-Gm-Message-State: AOJu0Yx3Wmyt2ajjuh1s0CC9LWoip93cb/SW/1T3F0lFFteQkyQskuWO k6WIK+vqPi4OflgIMTJ954G8ALKItkJDCrzef7Dd0i7Sbt/XaVWZDZXH3PW1XI4FPX4QX8TGtPj CDdZNM0KtQkCU2tl/7hriE1a/1IFLufglrHpGsQaep/0+hmWpmv6ebb7EGi9TMqkh X-Gm-Gg: ASbGncs9H4RwhU7FW4lz9R8HJOllLU7yxEeXUNd6gABz0glHMfCC3EXTeBYgLvAR2OP Hz96VZnd9QKiCM5vDrJfdqJ9uhUANMCVdyL5QbqAi2BphM4e9qXZs1MMtf2guBRP8HMIrJspZ8M Zf/68pBwCyf7S/ZZUu2rMY2H1+0BVs3MnOVKogm5VskX7i03QlJkMEansg8JoXA8sVEoXcBIDSP Emb0+pIlJwypmlomgmRLHGy7L2DT7gaWZgO6cEGuFVrEMbFmiYv7XSJS7zt1fLWaH+qoMk= X-Received: by 2002:a05:600c:a014:b0:439:8340:637 with SMTP id 5b1f17b1804b1-439861fdba3mr42542975e9.30.1739837558453; Mon, 17 Feb 2025 16:12:38 -0800 (PST) X-Google-Smtp-Source: AGHT+IG5kjO9h+qY7R0mzmILuEK/JSqwr6CGQLnkfN78PWHYwp+tnpQVFFvr841liMidGooiybCi0g== X-Received: by 2002:a05:600c:a014:b0:439:8340:637 with SMTP id 5b1f17b1804b1-439861fdba3mr42542725e9.30.1739837557989; Mon, 17 Feb 2025 16:12:37 -0800 (PST) Received: from redhat.com ([2a02:14f:1f5:7ef9:8f0d:a8:10c1:4db]) by smtp.gmail.com with ESMTPSA id 5b1f17b1804b1-439858ec5fasm39352515e9.29.2025.02.17.16.12.36 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Mon, 17 Feb 2025 16:12:37 -0800 (PST) Date: Mon, 17 Feb 2025 19:12:34 -0500 From: "Michael S. Tsirkin" To: "Ning, Hongyu" Cc: Eric Auger , "Kirill A. Shutemov" , Jason Wang , Xuan Zhuo , Eugenio =?iso-8859-1?Q?P=E9rez?= , Zhenzhong Duan , virtualization@lists.linux.dev, linux-kernel@vger.kernel.org Subject: Re: [PATCH] virtio: Remove virtio devices on device_shutdown() Message-ID: <20250217191103-mutt-send-email-mst@kernel.org> References: <20240808075141.3433253-1-kirill.shutemov@linux.intel.com> <20250203094700-mutt-send-email-mst@kernel.org> <7cee3c9e-515e-41de-a15c-04c7591e83eb@redhat.com> <6bce0f4c-636f-456b-ab21-4a25d3dc8803@redhat.com> <90a09ffa-e316-41f0-916b-25635b1d4bc6@linux.intel.com> <83b43e73-8599-44ff-8657-6d5f2f9b2de5@redhat.com> <20250214070904-mutt-send-email-mst@kernel.org> Precedence: bulk X-Mailing-List: virtualization@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 In-Reply-To: X-Mimecast-Spam-Score: 0 X-Mimecast-MFC-PROC-ID: wdrTrksKvL63efM7Kh9Sc-E7RvAZXhvNIpKB6c12Me4_1739837558 X-Mimecast-Originator: redhat.com Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit On Tue, Feb 18, 2025 at 07:57:23AM +0800, Ning, Hongyu wrote: > > > On 2025/2/17 17:25, Eric Auger wrote: > > Hi Michael, Hongyu, > > > > On 2/14/25 1:16 PM, Michael S. Tsirkin wrote: > > > On Fri, Feb 14, 2025 at 08:56:56AM +0100, Eric Auger wrote: > > > > Hi, > > > > > > > > On 2/14/25 8:21 AM, Ning, Hongyu wrote: > > > > > > > > > > > > > > > On 2025/2/6 16:59, Eric Auger wrote: > > > > > > Hi, > > > > > > > > > > > > On 2/4/25 12:46 PM, Eric Auger wrote: > > > > > > > Hi, > > > > > > > > > > > > > > On 2/3/25 3:48 PM, Michael S. Tsirkin wrote: > > > > > > > > On Fri, Jan 31, 2025 at 10:53:15AM +0100, Eric Auger wrote: > > > > > > > > > Hi Kirill, Michael > > > > > > > > > > > > > > > > > > On 8/8/24 9:51 AM, Kirill A. Shutemov wrote: > > > > > > > > > > Hongyu reported a hang on kexec in a VM. QEMU reported invalid memory > > > > > > > > > > accesses during the hang. > > > > > > > > > > > > > > > > > > > >     Invalid read at addr 0x102877002, size 2, region '(null)', > > > > > > > > > > reason: rejected > > > > > > > > > >     Invalid write at addr 0x102877A44, size 2, region '(null)', > > > > > > > > > > reason: rejected > > > > > > > > > >     ... > > > > > > > > > > > > > > > > > > > > It was traced down to virtio-console. Kexec works fine if virtio- > > > > > > > > > > console > > > > > > > > > > is not in use. > > > > > > > > > > > > > > > > > > > > Looks like virtio-console continues to write to the MMIO even after > > > > > > > > > > underlying virtio-pci device is removed. > > > > > > > > > > > > > > > > > > > > The problem can be mitigated by removing all virtio devices on virtio > > > > > > > > > > bus shutdown. > > > > > > > > > > > > > > > > > > > > Signed-off-by: Kirill A. Shutemov > > > > > > > > > > Reported-by: Hongyu Ning > > > > > > > > > > > > > > > > > > Gentle ping on that patch that seems to have fallen though the cracks. > > > > > > > > > > > > > > > > > > I think this fix is really needed. I have another test case with a > > > > > > > > > rebooting guest exposed with virtio-net (backed by vhost-net) and > > > > > > > > > viommu. Since there is currently no shutdown for the virtio-net, on > > > > > > > > > reboot, the IOMMU is disabled through the native_machine_shutdown()/ > > > > > > > > > x86_platform.iommu_shutdown() while the virtio-net is still alive. > > > > > > > > > > > > > > > > > > Normally device_shutdown() should call virtio-net shutdown before the > > > > > > > > > IOMMU tear down and we wouldn't see any spurious transactions after > > > > > > > > > iommu shutdown. > > > > > > > > > > > > > > > > > > With that fix, the above test case is fixed and I do not see spurious > > > > > > > > > vhost IOTLB miss spurious requests. > > > > > > > > > > > > > > > > > > For more details, see qemu thread ([PATCH] hw/virtio/vhost: Disable > > > > > > > > > IOTLB callbacks when IOMMU gets disabled, > > > > > > > > > https://lore.kernel.org/all/20250120173339.865681-1- > > > > > > > > > eric.auger@redhat.com/) > > > > > > > > > > > > > > > > > > > > > > > > > > > Reviewed-by: Eric Auger > > > > > > > > > Tested-by: Eric Auger > > > > > > > > > > > > > > > > > > Thanks > > > > > > > > > > > > > > > > > > Eric > > > > > > > > > > > > > > > > > > > --- > > > > > > > > > >   drivers/virtio/virtio.c | 10 ++++++++++ > > > > > > > > > >   1 file changed, 10 insertions(+) > > > > > > > > > > > > > > > > > > > > diff --git a/drivers/virtio/virtio.c b/drivers/virtio/virtio.c > > > > > > > > > > index a9b93e99c23a..6c2f908eb22c 100644 > > > > > > > > > > --- a/drivers/virtio/virtio.c > > > > > > > > > > +++ b/drivers/virtio/virtio.c > > > > > > > > > > @@ -356,6 +356,15 @@ static void virtio_dev_remove(struct device *_d) > > > > > > > > > >       of_node_put(dev->dev.of_node); > > > > > > > > > >   } > > > > > > > > > >   +static void virtio_dev_shutdown(struct device *_d) > > > > > > > > > > +{ > > > > > > > > > > +    struct virtio_device *dev = dev_to_virtio(_d); > > > > > > > > > > +    struct virtio_driver *drv = drv_to_virtio(dev->dev.driver); > > > > > > > > > > + > > > > > > > > > > +    if (drv && drv->remove) > > > > > > > > > > +        drv->remove(dev); > > > > > > > > > > > > > > > > > > > > > > > > I am concerned that full remove is a heavyweight operation. > > > > > > > > Do not want to slow down reboots even more. > > > > > > > > How about just doing a reset, instead? > > > > > > > > > > > > > > I tested with > > > > > > > > > > > > > > static void virtio_dev_shutdown(struct device *_d) > > > > > > > { > > > > > > >          struct virtio_device *dev = dev_to_virtio(_d); > > > > > > > > > > > > > >          virtio_reset_device(dev); > > > > > > > } > > > > > > > > > > > > > > > > > > > > > and it fixes my issue. > > > > > > > > > > > > > > Kirill, would that fix you issue too? > > > > > > > > > > Hi, > > > > > > > > > > sorry for my late response, I synced with Kirill offline and did a retest. > > > > > > > > > > The issue is still reproduced on my side, kexec will be stuck in case of > > > > > "console=hvc0" append in kernel cmdline and even with such patch applied. > > > > > > > > Thanks for testing! > > > > > > > > Michael, it looks like the initial patch from Kyrill is the one that > > > > fixes both issues. virtio_reset_device() usage does not work for the > > > > initial bug report while it works for me. Other ideas? > > > > > > > > Thanks > > > > > > > > Eric > > > > > > Ah, wait a second. > > > > > > Looks like virtio-console continues to write to the MMIO even after > > > underlying virtio-pci device is removed. > > > > > > Hmm. I am not sure why that is a problem, but I assume some hypervisors just > > > hang the system if you try to kick them after reset. > > > Unfortunate that spec did not disallow it. > > > > > > If we want to prevent that, we want to do something like this: > > > > > > > > > /* > > > * Some devices get wedged if you kick them after they are > > > * reset. Mark all vqs as broken to make sure we don't. > > > */ > > > virtio_break_device(dev); > > > /* > > > * The below virtio_synchronize_cbs() guarantees that any > > > * interrupt for this line arriving after > > > * virtio_synchronize_vqs() has completed is guaranteed to see > > > * vq->broken as true. > > > */ > > > virtio_synchronize_cbs(dev); > > > dev->config->reset(dev); > > > > > > > > > I assume this still works for you, yes? > > Would that still been done in the virtio_dev_shutdown()? > > > > Is that what you tested Hongyu? > > Hi Eric, > > my patch applied based on Michael's comments: > > --- > drivers/virtio/virtio.c | 20 ++++++++++++++++++++ > 1 file changed, 20 insertions(+) > > diff --git a/drivers/virtio/virtio.c b/drivers/virtio/virtio.c > index ba37665188b5..458dc28be060 100644 > --- a/drivers/virtio/virtio.c > +++ b/drivers/virtio/virtio.c > @@ -395,6 +395,25 @@ static const struct cpumask > *virtio_irq_get_affinity(struct device *_d, > return dev->config->get_vq_affinity(dev, irq_vec); > } > > +static void virtio_dev_shutdown(struct device *_d) > +{ > + struct virtio_device *dev = dev_to_virtio(_d); > + /* > + * Some devices get wedged if you kick them after they are > + * reset. Mark all vqs as broken to make sure we don't. > + */ > + virtio_break_device(dev); > + /* > + * The below virtio_synchronize_cbs() guarantees that any > + * interrupt for this line arriving after > + * virtio_synchronize_vqs() has completed is guaranteed to see > + * vq->broken as true. > + */ > + virtio_synchronize_cbs(dev); > + dev->config->reset(dev); > + //virtio_reset_device(dev); > +} > + > static const struct bus_type virtio_bus = { > .name = "virtio", > .match = virtio_dev_match, > @@ -403,6 +422,7 @@ static const struct bus_type virtio_bus = { > .probe = virtio_dev_probe, > .remove = virtio_dev_remove, > .irq_get_affinity = virtio_irq_get_affinity, > + .shutdown = virtio_dev_shutdown, > }; > > int __register_virtio_driver(struct virtio_driver *driver, struct module > *owner) > -- > 2.43.0 yes, that's the patch. Does it work for you? Addresses the problem? If yes I'll repost. > > > > > Eric > > > > > > > > > > > > > > > > > > > my kernel code base is 6.14.0-rc2. > > > > > > > > > > let me know if any more experiments needed. > > > > > > > > > > --- > > > > >  drivers/virtio/virtio.c | 8 ++++++++ > > > > >  1 file changed, 8 insertions(+) > > > > > > > > > > diff --git a/drivers/virtio/virtio.c b/drivers/virtio/virtio.c > > > > > index ba37665188b5..f9f885d04763 100644 > > > > > --- a/drivers/virtio/virtio.c > > > > > +++ b/drivers/virtio/virtio.c > > > > > @@ -395,6 +395,13 @@ static const struct cpumask > > > > > *virtio_irq_get_affinity(struct device *_d, > > > > >         return dev->config->get_vq_affinity(dev, irq_vec); > > > > >  } > > > > > > > > > > +static void virtio_dev_shutdown(struct device *_d) > > > > > +{ > > > > > +        struct virtio_device *dev = dev_to_virtio(_d); > > > > > + > > > > > +        virtio_reset_device(dev); > > > > > +} > > > > > + > > > > >  static const struct bus_type virtio_bus = { > > > > >         .name  = "virtio", > > > > >         .match = virtio_dev_match, > > > > > @@ -403,6 +410,7 @@ static const struct bus_type virtio_bus = { > > > > >         .probe = virtio_dev_probe, > > > > >         .remove = virtio_dev_remove, > > > > >         .irq_get_affinity = virtio_irq_get_affinity, > > > > > +       .shutdown = virtio_dev_shutdown, > > > > >  }; > > > > > > > > > >  int __register_virtio_driver(struct virtio_driver *driver, struct > > > > > module *owner) > > > > > -- > > > > > 2.43.0 > > > > > > > > > > > > > > > > gentle ping. > > > > > > > > > > > > this also fixes another issue with qemu vSMMU + virtio-scsi-pci. With > > > > > > the above addition I get rid of spurious warning in qemu on guest reboot. > > > > > > > > > > > > qemu-system-aarch64: virtio: zero sized buffers are not allowed > > > > > > qemu-system-aarch64: vhost vring error in virtqueue 0: Invalid > > > > > > argument (22) > > > > > > > > > > > > Would you mind if I respin? > > > > > > > > > > > > Thanks > > > > > > > > > > > > Eric > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Thanks > > > > > > > > > > > > > > Eric > > > > > > > > > > > > > > > > > > +} > > > > > > > > > > + > > > > > > > > > >   static const struct bus_type virtio_bus = { > > > > > > > > > >       .name  = "virtio", > > > > > > > > > >       .match = virtio_dev_match, > > > > > > > > > > @@ -363,6 +372,7 @@ static const struct bus_type virtio_bus = { > > > > > > > > > >       .uevent = virtio_uevent, > > > > > > > > > >       .probe = virtio_dev_probe, > > > > > > > > > >       .remove = virtio_dev_remove, > > > > > > > > > > +    .shutdown = virtio_dev_shutdown, > > > > > > > > > >   }; > > > > > > > > > >     int __register_virtio_driver(struct virtio_driver *driver, > > > > > > > > > > struct module *owner) > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > >