[dpdk-dev] [PATCH v4 4/4] virtio: return 1 to tell the upper layer we don't take over this device

Santosh Shukla sshukla at mvista.com
Mon Feb 29 14:15:26 CET 2016


On Fri, Feb 26, 2016 at 7:23 AM, Huawei Xie <huawei.xie at intel.com> wrote:
> v4 changes:
>  Rebase as io port map is moved to eal.
>  Only fall back to PORT IO when there isn't any kernel driver (including

Pl. mention that fallback behaviour applicable to x86 arch only..

However this patch fixes one problem in non-x86 arch issue, Example:
VM has 8 virtio interface and 2 i/f attached out of 8, so in default
case - after 2nd interface, ioport try to program 3..8 ports, result
to failure, lead to exit dpdk application. Patch fixes this problem
for non-x86 arch, test on arm64 platform.

> VFIO/UIO) managing the device. Before v4, we fall back to PORT IO even if
> VFIO/UIO fails.
>  Reword the commit message.
>
> v3 changes:
>  Change log message to tell user that the virtio device is skipped
> due to it is managed by kernel driver, instead of asking user to
> unbind it from kernel driver.
>
> v2 changes:
>  Remove unnecessary assignment of NULL to dev->data->mac_addrs.
>  Ajust one comment's position.
>
> virtio PMD could use IO port to configure the virtio device without
> using UIO/VFIO driver in legacy mode.
>
> There are two issues with the previous implementation:
> 1) virtio PMD will take over the virtio device(s) blindly even if not
> intended for DPDK.
> 2) driver conflict between virtio PMD and virtio-net kernel driver.
>
> This patch checks if there is kernel driver other than UIO/VFIO managing
> the virtio device before using port IO.
>
> If legacy_virtio_resource_init fails and kernel driver other than
> VFIO/UIO is managing the device, return 1 to tell the upper layer we
> don't take over this device.
> For all other IO port mapping errors, return -1.
>
> Note than if VFIO/UIO fails, now we don't fall back to port IO.
>
> Fixes: da978dfdc43b ("virtio: use port IO to get PCI resource")
>
> Signed-off-by: Huawei Xie <huawei.xie at intel.com>
> ---
>  drivers/net/virtio/virtio_ethdev.c |  9 +++++++--
>  drivers/net/virtio/virtio_pci.c    | 15 ++++++++++++++-
>  2 files changed, 21 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/net/virtio/virtio_ethdev.c b/drivers/net/virtio/virtio_ethdev.c
> index caa970c..8601080 100644
> --- a/drivers/net/virtio/virtio_ethdev.c
> +++ b/drivers/net/virtio/virtio_ethdev.c
> @@ -1,4 +1,5 @@
>  /*-
> +
>   *   BSD LICENSE
>   *
>   *   Copyright(c) 2010-2015 Intel Corporation. All rights reserved.
> @@ -1015,6 +1016,7 @@ eth_virtio_dev_init(struct rte_eth_dev *eth_dev)
>         struct virtio_net_config *config;
>         struct virtio_net_config local_config;
>         struct rte_pci_device *pci_dev;
> +       int ret;
>
>         RTE_BUILD_BUG_ON(RTE_PKTMBUF_HEADROOM < sizeof(struct virtio_net_hdr));
>
> @@ -1037,8 +1039,11 @@ eth_virtio_dev_init(struct rte_eth_dev *eth_dev)
>
>         pci_dev = eth_dev->pci_dev;
>
> -       if (vtpci_init(pci_dev, hw) < 0)
> -               return -1;
> +       ret = vtpci_init(pci_dev, hw);
> +       if (ret) {
> +               rte_free(eth_dev->data->mac_addrs);
> +               return ret;
> +       }
>
>         /* Reset the device although not necessary at startup */
>         vtpci_reset(hw);
> diff --git a/drivers/net/virtio/virtio_pci.c b/drivers/net/virtio/virtio_pci.c
> index 85fbe88..f159b2a 100644
> --- a/drivers/net/virtio/virtio_pci.c
> +++ b/drivers/net/virtio/virtio_pci.c
> @@ -622,6 +622,13 @@ next:
>         return 0;
>  }
>
> +/*
> + * Return -1:
> + *   if there is error mapping with VFIO/UIO.
> + *   if port map error when driver type is KDRV_NONE.
> + * Return 1 if kernel driver is managing the device.
> + * Return 0 on success.
> + */
>  int
>  vtpci_init(struct rte_pci_device *dev, struct virtio_hw *hw)
>  {
> @@ -641,8 +648,14 @@ vtpci_init(struct rte_pci_device *dev, struct virtio_hw *hw)
>         }
>
>         PMD_INIT_LOG(INFO, "trying with legacy virtio pci.");
> -       if (legacy_virtio_resource_init(dev, hw) < 0)
> +       if (legacy_virtio_resource_init(dev, hw) < 0) {
> +               if (dev->kdrv == RTE_KDRV_UNKNOWN) {
> +                       PMD_INIT_LOG(INFO,
> +                               "skip kernel managed virtio device.");
> +                       return 1;
> +               }
>                 return -1;
> +       }
>
>         hw->vtpci_ops = &legacy_ops;
>         hw->use_msix = legacy_virtio_has_msix(&dev->addr);

Tested-by: Santosh Shukla <sshukla at mvista.com>
Acked-by: Santosh Shukla <sshukla at mvista.com>

> --
> 1.8.1.4
>


More information about the dev mailing list