On Thu, 21 Jan 2021 13:36:24 +0100 Jean-Philippe Brucker <jean-philippe@xxxxxxxxxx> wrote: > The SMMU provides a Stall model for handling page faults in platform > devices. It is similar to PCIe PRI, but doesn't require devices to have > their own translation cache. Instead, faulting transactions are parked > and the OS is given a chance to fix the page tables and retry the > transaction. > > Enable stall for devices that support it (opt-in by firmware). When an > event corresponds to a translation error, call the IOMMU fault handler. > If the fault is recoverable, it will call us back to terminate or > continue the stall. > > To use stall device drivers need to enable IOMMU_DEV_FEAT_IOPF, which > initializes the fault queue for the device. > > Tested-by: Zhangfei Gao <zhangfei.gao@xxxxxxxxxx> > Signed-off-by: Jean-Philippe Brucker <jean-philippe@xxxxxxxxxx> One thing inline + one comment which was mostly a case of I ran out of time to walk through why probe and release aren't symmetric... > --- > drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h | 43 ++++ > .../iommu/arm/arm-smmu-v3/arm-smmu-v3-sva.c | 59 +++++- > drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 185 +++++++++++++++++- > 3 files changed, 273 insertions(+), 14 deletions(-) > ... > diff --git a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c > index db5d6aa76c3a..af6982aca42e 100644 > --- a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c > +++ b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c > @@ -32,6 +32,7 @@ ... > > master->domain = smmu_domain; > @@ -2484,6 +2624,11 @@ static struct iommu_device *arm_smmu_probe_device(struct device *dev) > master->ssid_bits = min_t(u8, master->ssid_bits, > CTXDESC_LINEAR_CDMAX); > > + if ((smmu->features & ARM_SMMU_FEAT_STALLS && > + device_property_read_bool(dev, "dma-can-stall")) || > + smmu->features & ARM_SMMU_FEAT_STALL_FORCE) > + master->stall_enabled = true; > + > return &smmu->iommu; > > err_free_master: > @@ -2502,6 +2647,7 @@ static void arm_smmu_release_device(struct device *dev) > > master = dev_iommu_priv_get(dev); > WARN_ON(arm_smmu_master_sva_enabled(master)); > + iopf_queue_remove_device(master->smmu->evtq.iopf, dev); > arm_smmu_detach_dev(master); > arm_smmu_disable_pasid(master); > arm_smmu_remove_master(master); The lack of symmetry here bothers me a bit, but it's already true, so I guess this case is fine as well. ... > > @@ -2785,6 +2946,7 @@ static int arm_smmu_cmdq_init(struct arm_smmu_device *smmu) > static int arm_smmu_init_queues(struct arm_smmu_device *smmu) > { > int ret; > + bool sva = smmu->features & ARM_SMMU_FEAT_STALLS; FEAT_SVA? > > /* cmdq */ > ret = arm_smmu_init_one_queue(smmu, &smmu->cmdq.q, ARM_SMMU_CMDQ_PROD, > @@ -2804,6 +2966,12 @@ static int arm_smmu_init_queues(struct arm_smmu_device *smmu) > if (ret) > return ret; > > + if (sva && smmu->features & ARM_SMMU_FEAT_STALLS) { Isn't this checking same thing twice? > + smmu->evtq.iopf = iopf_queue_alloc(dev_name(smmu->dev)); > + if (!smmu->evtq.iopf) > + return -ENOMEM; > + } > + > /* priq */ > if (!(smmu->features & ARM_SMMU_FEAT_PRI)) > return 0; > @@ -3718,6 +3886,7 @@ static int arm_smmu_device_remove(struct platform_device *pdev) > iommu_device_unregister(&smmu->iommu); > iommu_device_sysfs_remove(&smmu->iommu); > arm_smmu_device_disable(smmu); > + iopf_queue_free(smmu->evtq.iopf); > > return 0; > }