From: Joerg Roedel <jroedel@suse.de>
To: Bjorn Helgaas <bhelgaas@google.com>
Cc: linux-pci@vger.kernel.org, David Woodhouse <dwmw2@infradead.org>,
iommu@lists.linux-foundation.org,
Gregor Dick <gdick@solarflare.com>
Subject: Re: [PATCH v2 02/11] PCI: Allocate ATS struct during enumeration
Date: Mon, 27 Jul 2015 14:40:24 +0200 [thread overview]
Message-ID: <20150727124024.GB27614@suse.de> (raw)
In-Reply-To: <20150721001405.28145.43004.stgit@bhelgaas-glaptop2.roam.corp.google.com>
On Mon, Jul 20, 2015 at 07:14:05PM -0500, Bjorn Helgaas wrote:
> Previously, we allocated pci_ats structures when an IOMMU driver called
> pci_enable_ats(). An SR-IOV VF shares the STU setting with its PF, so when
> enabling ATS on the VF, we allocated a pci_ats struct for the PF if it
> didn't already have one. We held the sriov->lock to serialize threads
> concurrently enabling ATS on several VFS so only one would allocate the PF
> pci_ats.
>
> Gregor reported a deadlock here:
>
> pci_enable_sriov
> sriov_enable
> virtfn_add
> mutex_lock(dev->sriov->lock) # acquire sriov->lock
> pci_device_add
> device_add
> BUS_NOTIFY_ADD_DEVICE notifier chain
> iommu_bus_notifier
> amd_iommu_add_device # iommu_ops.add_device
> init_iommu_group
> iommu_group_get_for_dev
> iommu_group_add_device
> __iommu_attach_device
> amd_iommu_attach_device # iommu_ops.attach_device
> attach_device
> pci_enable_ats
> mutex_lock(dev->sriov->lock) # deadlock
>
> There's no reason to delay allocating the pci_ats struct, and if we
> allocate it for each device at enumeration-time, there's no need for
> locking in pci_enable_ats().
>
> Allocate pci_ats struct during enumeration, when we initialize other
> capabilities.
>
> Note that this implementation requires ATS to be enabled on the PF first,
> before on any of the VFs because the PF controls the STU for all the VFs.
>
> Link: http://permalink.gmane.org/gmane.linux.kernel.iommu/9433
> Reported-by: Gregor Dick <gdick@solarflare.com>
> Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
> ---
> drivers/pci/ats.c | 98 ++++++++++++++++++++---------------------------
> drivers/pci/probe.c | 3 +
> drivers/pci/remove.c | 1
> include/linux/pci-ats.h | 2 -
> include/linux/pci.h | 9 ++++
> 5 files changed, 56 insertions(+), 57 deletions(-)
Looks good to me now.
Reviewed-by: Joerg Roedel <jroedel@suse.de>
next prev parent reply other threads:[~2015-07-27 12:40 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-07-21 0:13 [PATCH v2 00/11] PCI: Fix ATS deadlock Bjorn Helgaas
2015-07-21 0:13 ` [PATCH v2 01/11] iommu/vt-d: Cache PCI ATS state and Invalidate Queue Depth Bjorn Helgaas
2015-07-27 13:08 ` Joerg Roedel
2015-07-27 22:54 ` Bjorn Helgaas
2015-07-28 7:14 ` Joerg Roedel
2015-07-21 0:14 ` [PATCH v2 02/11] PCI: Allocate ATS struct during enumeration Bjorn Helgaas
2015-07-27 12:40 ` Joerg Roedel [this message]
2015-07-21 0:14 ` [PATCH v2 03/11] PCI: Embed ATS info directly into struct pci_dev Bjorn Helgaas
2015-07-27 12:45 ` Joerg Roedel
2015-07-21 0:14 ` [PATCH v2 04/11] PCI: Reduce size of ATS structure elements Bjorn Helgaas
2015-07-21 0:14 ` [PATCH v2 05/11] PCI: Rationalize pci_ats_queue_depth() error checking Bjorn Helgaas
2015-07-21 0:14 ` [PATCH v2 06/11] PCI: Inline the ATS setup code into pci_ats_init() Bjorn Helgaas
2015-07-21 0:14 ` [PATCH v2 07/11] PCI: Use pci_physfn() rather than looking up physfn by hand Bjorn Helgaas
2015-07-21 0:14 ` [PATCH v2 08/11] PCI: Clean up ATS error handling Bjorn Helgaas
2015-07-27 12:56 ` Joerg Roedel
2015-07-21 0:15 ` [PATCH v2 09/11] PCI: Move ATS declarations to linux/pci.h so they're all together Bjorn Helgaas
2015-07-21 0:15 ` [PATCH v2 10/11] PCI: Stop caching ATS Invalidate Queue Depth Bjorn Helgaas
2015-07-27 12:57 ` Joerg Roedel
2015-07-27 14:00 ` Don Dutile
2015-07-27 22:27 ` Bjorn Helgaas
2015-07-27 23:13 ` Don Dutile
2015-07-21 0:15 ` [PATCH v2 11/11] PCI: Remove pci_ats_enabled() Bjorn Helgaas
2015-07-27 12:58 ` Joerg Roedel
2015-07-28 15:16 ` [PATCH v2 00/11] PCI: Fix ATS deadlock Joerg Roedel
2015-07-29 16:07 ` Bjorn Helgaas
2015-08-06 16:03 ` Yinghai Lu
2015-08-07 1:06 ` Yinghai Lu
2015-08-10 17:33 ` Bjorn Helgaas
2015-08-10 22:54 ` Yinghai Lu
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20150727124024.GB27614@suse.de \
--to=jroedel@suse.de \
--cc=bhelgaas@google.com \
--cc=dwmw2@infradead.org \
--cc=gdick@solarflare.com \
--cc=iommu@lists.linux-foundation.org \
--cc=linux-pci@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).