[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Thu, 5 May 2022 22:00:20 +0000
From: "Michael Kelley (LINUX)" <mikelley@...rosoft.com>
To: "Andrea Parri (Microsoft)" <parri.andrea@...il.com>,
KY Srinivasan <kys@...rosoft.com>,
Haiyang Zhang <haiyangz@...rosoft.com>,
Stephen Hemminger <sthemmin@...rosoft.com>,
Wei Liu <wei.liu@...nel.org>, Dexuan Cui <decui@...rosoft.com>,
Lorenzo Pieralisi <lorenzo.pieralisi@....com>,
Rob Herring <robh@...nel.org>,
Krzysztof Wilczynski <kw@...ux.com>,
Bjorn Helgaas <bhelgaas@...gle.com>
CC: "linux-hyperv@...r.kernel.org" <linux-hyperv@...r.kernel.org>,
"linux-pci@...r.kernel.org" <linux-pci@...r.kernel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: RE: [PATCH 2/2] PCI: hv: Fix synchronization between channel callback
and hv_pci_bus_exit()
From: Andrea Parri (Microsoft) <parri.andrea@...il.com> Sent: Wednesday, May 4, 2022 5:51 AM
>
> [ Similarly to commit a765ed47e4516 ("PCI: hv: Fix synchronization
> between channel callback and hv_compose_msi_msg()"): ]
>
> The (on-stack) teardown packet becomes invalid once the completion
> timeout in hv_pci_bus_exit() has expired and hv_pci_bus_exit() has
> returned. Prevent the channel callback from accessing the invalid
> packet by removing the ID associated to such packet from the VMbus
> requestor in hv_pci_bus_exit().
>
> Signed-off-by: Andrea Parri (Microsoft) <parri.andrea@...il.com>
> ---
> drivers/pci/controller/pci-hyperv.c | 26 +++++++++++++++++++-------
> 1 file changed, 19 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/pci/controller/pci-hyperv.c b/drivers/pci/controller/pci-hyperv.c
> index 9a3e17b682eb7..db4b3f86726b2 100644
> --- a/drivers/pci/controller/pci-hyperv.c
> +++ b/drivers/pci/controller/pci-hyperv.c
> @@ -3620,6 +3620,7 @@ static int hv_pci_probe(struct hv_device *hdev,
> static int hv_pci_bus_exit(struct hv_device *hdev, bool keep_devs)
> {
> struct hv_pcibus_device *hbus = hv_get_drvdata(hdev);
> + struct vmbus_channel *chan = hdev->channel;
> struct {
> struct pci_packet teardown_packet;
> u8 buffer[sizeof(struct pci_message)];
> @@ -3627,13 +3628,14 @@ static int hv_pci_bus_exit(struct hv_device *hdev, bool
> keep_devs)
> struct hv_pci_compl comp_pkt;
> struct hv_pci_dev *hpdev, *tmp;
> unsigned long flags;
> + u64 trans_id;
> int ret;
>
> /*
> * After the host sends the RESCIND_CHANNEL message, it doesn't
> * access the per-channel ringbuffer any longer.
> */
> - if (hdev->channel->rescind)
> + if (chan->rescind)
> return 0;
>
> if (!keep_devs) {
> @@ -3670,16 +3672,26 @@ static int hv_pci_bus_exit(struct hv_device *hdev, bool
> keep_devs)
> pkt.teardown_packet.compl_ctxt = &comp_pkt;
> pkt.teardown_packet.message[0].type = PCI_BUS_D0EXIT;
>
> - ret = vmbus_sendpacket(hdev->channel, &pkt.teardown_packet.message,
> - sizeof(struct pci_message),
> - (unsigned long)&pkt.teardown_packet,
> - VM_PKT_DATA_INBAND,
> - VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
> + ret = vmbus_sendpacket_getid(chan, &pkt.teardown_packet.message,
> + sizeof(struct pci_message),
> + (unsigned long)&pkt.teardown_packet,
> + &trans_id, VM_PKT_DATA_INBAND,
> +
> VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
> if (ret)
> return ret;
>
> - if (wait_for_completion_timeout(&comp_pkt.host_event, 10 * HZ) == 0)
> + if (wait_for_completion_timeout(&comp_pkt.host_event, 10 * HZ) == 0) {
> + /*
> + * The completion packet on the stack becomes invalid after
> + * 'return'; remove the ID from the VMbus requestor if the
> + * identifier is still mapped to/associated with the packet.
> + *
> + * Cf. hv_pci_onchannelcallback().
> + */
> + vmbus_request_addr_match(chan, trans_id,
> + (unsigned long)&pkt.teardown_packet);
> return -ETIMEDOUT;
> + }
>
> return 0;
> }
> --
> 2.25.1
Reviewed-by: Michael Kelley <mikelley@...rosoft.com>
Powered by blists - more mailing lists