On 25/08/2023 10:02, Vikram Garhwal wrote:
> Rename iommu_dt_device_is_assigned() to iommu_dt_device_is_assigned_locked().
>
> Moving spin_lock to caller was done to prevent the concurrent access to
> iommu_dt_device_is_assigned while doing add/remove/assign/deassign. Follow-up
> patches in this series introduces node add/remove feature.
>
> Signed-off-by: Vikram Garhwal <vikram.garh...@amd.com>
>
> ---
> Changes from v9:
> Make iommu_dt_device_is_assigned_locked() static and delete header.
> Move dtdevs_lock before iommu_dt_device_is_assigned_locked().
> Changes from v7:
> Update commit message.
> Add ASSERT().
> ---
> ---
> xen/drivers/passthrough/device_tree.c | 16 ++++++++++++----
> 1 file changed, 12 insertions(+), 4 deletions(-)
>
> diff --git a/xen/drivers/passthrough/device_tree.c
> b/xen/drivers/passthrough/device_tree.c
> index 1c32d7b50c..5d84c07b50 100644
> --- a/xen/drivers/passthrough/device_tree.c
> +++ b/xen/drivers/passthrough/device_tree.c
> @@ -83,16 +83,17 @@ fail:
> return rc;
> }
>
> -static bool_t iommu_dt_device_is_assigned(const struct dt_device_node *dev)
> +static bool_t
> +iommu_dt_device_is_assigned_locked(const struct dt_device_node *dev)
This does not apply cleanly due to recent change from bool_t to bool. Please
rebase for v11 (the function
should then fit in a single line I think).
> {
> bool_t assigned = 0;
>
> + ASSERT(spin_is_locked(&dtdevs_lock));
> +
> if ( !dt_device_is_protected(dev) )
> return 0;
>
> - spin_lock(&dtdevs_lock);
> assigned = !list_empty(&dev->domain_list);
> - spin_unlock(&dtdevs_lock);
>
> return assigned;
> }
> @@ -223,17 +224,24 @@ int iommu_do_dt_domctl(struct xen_domctl *domctl,
> struct domain *d,
> if ( ret )
> break;
>
> + spin_lock(&dtdevs_lock);
Why is this lock placed here instead of ...
> +
> if ( domctl->cmd == XEN_DOMCTL_test_assign_device )
> {
> - if ( iommu_dt_device_is_assigned(dev) )
> +
... here, right before iommu_dt_device_is_assigned_locked()?
> + if ( iommu_dt_device_is_assigned_locked(dev) )
> {
> printk(XENLOG_G_ERR "%s already assigned.\n",
> dt_node_full_name(dev));
> ret = -EINVAL;
> }
> +
> + spin_unlock(&dtdevs_lock);
> break;
> }
>
> + spin_unlock(&dtdevs_lock);
You could then remove this one.
With the remarks addressed:
Reviewed-by: Michal Orzel <michal.or...@amd.com>
~Michal
