| Message ID | f707b4248e1d33b6d2c7f1d7c94febb802cf9890.1649161199.git.robin.murphy@arm.com (mailing list archive) |
|---|---|
| State | Accepted |
| Delegated to: | Jason Gunthorpe |
| Headers | show |
| Series | RDMA/usnic: Stop using iommu_present() | expand |
On Tue, Apr 05, 2022 at 01:19:59PM +0100, Robin Murphy wrote: > Even if an IOMMU might be present for some PCI segment in the system, > that doesn't necessarily mean it provides translation for the device(s) > we care about. Replace iommu_present() with a more appropriate check at > probe time, and garbage-collect the resulting empty init function. > > Signed-off-by: Robin Murphy <robin.murphy@arm.com> > --- > drivers/infiniband/hw/usnic/usnic_ib_main.c | 11 +++++------ > drivers/infiniband/hw/usnic/usnic_uiom.c | 10 ---------- > drivers/infiniband/hw/usnic/usnic_uiom.h | 1 - > 3 files changed, 5 insertions(+), 17 deletions(-) Applied to for-next, thanks Jason
diff --git a/drivers/infiniband/hw/usnic/usnic_ib_main.c b/drivers/infiniband/hw/usnic/usnic_ib_main.c index d346dd48e731..46653ad56f5a 100644 --- a/drivers/infiniband/hw/usnic/usnic_ib_main.c +++ b/drivers/infiniband/hw/usnic/usnic_ib_main.c @@ -534,6 +534,11 @@ static int usnic_ib_pci_probe(struct pci_dev *pdev, struct usnic_ib_vf *vf; enum usnic_vnic_res_type res_type; + if (!device_iommu_mapped(&pdev->dev)) { + usnic_err("IOMMU required but not present or enabled. USNIC QPs will not function w/o enabling IOMMU\n"); + return -EPERM; + } + vf = kzalloc(sizeof(*vf), GFP_KERNEL); if (!vf) return -ENOMEM; @@ -642,12 +647,6 @@ static int __init usnic_ib_init(void) printk_once(KERN_INFO "%s", usnic_version); - err = usnic_uiom_init(DRV_NAME); - if (err) { - usnic_err("Unable to initialize umem with err %d\n", err); - return err; - } - err = pci_register_driver(&usnic_ib_pci_driver); if (err) { usnic_err("Unable to register with PCI\n"); diff --git a/drivers/infiniband/hw/usnic/usnic_uiom.c b/drivers/infiniband/hw/usnic/usnic_uiom.c index 760b254ba42d..8c48027614a1 100644 --- a/drivers/infiniband/hw/usnic/usnic_uiom.c +++ b/drivers/infiniband/hw/usnic/usnic_uiom.c @@ -556,13 +556,3 @@ void usnic_uiom_free_dev_list(struct device **devs) { kfree(devs); } - -int usnic_uiom_init(char *drv_name) -{ - if (!iommu_present(&pci_bus_type)) { - usnic_err("IOMMU required but not present or enabled. USNIC QPs will not function w/o enabling IOMMU\n"); - return -EPERM; - } - - return 0; -} diff --git a/drivers/infiniband/hw/usnic/usnic_uiom.h b/drivers/infiniband/hw/usnic/usnic_uiom.h index 7ec8991ace67..9407522179e9 100644 --- a/drivers/infiniband/hw/usnic/usnic_uiom.h +++ b/drivers/infiniband/hw/usnic/usnic_uiom.h @@ -91,5 +91,4 @@ struct usnic_uiom_reg *usnic_uiom_reg_get(struct usnic_uiom_pd *pd, unsigned long addr, size_t size, int access, int dmasync); void usnic_uiom_reg_release(struct usnic_uiom_reg *uiomr); -int usnic_uiom_init(char *drv_name); #endif /* USNIC_UIOM_H_ */
Even if an IOMMU might be present for some PCI segment in the system, that doesn't necessarily mean it provides translation for the device(s) we care about. Replace iommu_present() with a more appropriate check at probe time, and garbage-collect the resulting empty init function. Signed-off-by: Robin Murphy <robin.murphy@arm.com> --- drivers/infiniband/hw/usnic/usnic_ib_main.c | 11 +++++------ drivers/infiniband/hw/usnic/usnic_uiom.c | 10 ---------- drivers/infiniband/hw/usnic/usnic_uiom.h | 1 - 3 files changed, 5 insertions(+), 17 deletions(-)