Re: [PATCH 18/23] iommu: Express DMA strictness via the domain type

From: Lu Baolu
Date: Mon Jul 26 2021 - 08:30:05 EST

On 2021/7/26 16:27, Robin Murphy wrote:
On 2021-07-24 06:29, Lu Baolu wrote:
Hi Robin,

On 2021/7/22 2:20, Robin Murphy wrote:
Eliminate the iommu_get_dma_strict() indirection and pipe the
information through the domain type from the beginning. Besides
the flow simplification this also has several nice side-effects:

  - Automatically implies strict mode for untrusted devices by
    virtue of their IOMMU_DOMAIN_DMA override.
  - Ensures that we only ends up using flush queues for drivers
    which are aware of them and can actually benefit.

Is this expressed by vendor iommu driver has ops->flush_iotlb_all?

No, it's literally whether ->domain_alloc accepts the DMA_DOMAIN_FQ type or not.

  - Allows us to handle flush queue init failure by falling back
    to strict mode instead of leaving it to possibly blow up later.

Signed-off-by: Robin Murphy <robin.murphy@xxxxxxx>
---
  drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c |  2 +-
  drivers/iommu/arm/arm-smmu/arm-smmu.c       |  2 +-
  drivers/iommu/dma-iommu.c                   | 10 ++++++----
  drivers/iommu/iommu.c                       | 14 ++++----------
  include/linux/iommu.h                       |  1 -
  5 files changed, 12 insertions(+), 17 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 fa41026d272e..260b560d0075 100644
--- a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c
+++ b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c
@@ -2175,7 +2175,7 @@ static int arm_smmu_domain_finalise(struct iommu_domain *domain,
          .iommu_dev    = smmu->dev,
      };
-    if (!iommu_get_dma_strict(domain))
+    if (domain->type == IOMMU_DOMAIN_DMA_FQ)
          pgtbl_cfg.quirks |= IO_PGTABLE_QUIRK_NON_STRICT;
      pgtbl_ops = alloc_io_pgtable_ops(fmt, &pgtbl_cfg, smmu_domain);
diff --git a/drivers/iommu/arm/arm-smmu/arm-smmu.c b/drivers/iommu/arm/arm-smmu/arm-smmu.c
index dbc14c265b15..2c717f3be056 100644
--- a/drivers/iommu/arm/arm-smmu/arm-smmu.c
+++ b/drivers/iommu/arm/arm-smmu/arm-smmu.c
@@ -765,7 +765,7 @@ static int arm_smmu_init_domain_context(struct iommu_domain *domain,
          .iommu_dev    = smmu->dev,
      };
-    if (!iommu_get_dma_strict(domain))
+    if (domain->type == IOMMU_DOMAIN_DMA_FQ)
          pgtbl_cfg.quirks |= IO_PGTABLE_QUIRK_NON_STRICT;
      if (smmu->impl && smmu->impl->init_context) {
diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c
index b1af1ff324c5..a114a7ad88ec 100644
--- a/drivers/iommu/dma-iommu.c
+++ b/drivers/iommu/dma-iommu.c
@@ -363,13 +363,15 @@ static int iommu_dma_init_domain(struct iommu_domain *domain, dma_addr_t base,
      init_iova_domain(iovad, 1UL << order, base_pfn);
-    if (!cookie->fq_domain && !dev_is_untrusted(dev) &&
-        domain->ops->flush_iotlb_all && !iommu_get_dma_strict(domain)) {
+    if (domain->type == IOMMU_DOMAIN_DMA_FQ && !cookie->fq_domain &&
+        domain->ops->flush_iotlb_all) {

Perhaps we can remove the ops->flush_iotlb_all check with the
assumption that any vendor iommu driver with DMA_FQ domain support
should always provides this callback?

Best regards,
baolu

          if (init_iova_flush_queue(iovad, iommu_dma_flush_iotlb_all,
-                      iommu_dma_entry_dtor))
+                      iommu_dma_entry_dtor)) {
              pr_warn("iova flush queue initialization failed\n");
-        else
+            domain->type = IOMMU_DOMAIN_DMA;
+        } else {
              cookie->fq_domain = domain;
+        }
      }
      return iova_reserve_iommu_regions(dev, domain);
diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
index 8333c334891e..d7eaacae0944 100644
--- a/drivers/iommu/iommu.c
+++ b/drivers/iommu/iommu.c
@@ -135,6 +135,9 @@ static int __init iommu_subsys_init(void)
          }
      }
+    if (!iommu_default_passthrough() && !iommu_dma_strict)
+        iommu_def_domain_type = IOMMU_DOMAIN_DMA_FQ;
+
      pr_info("Default domain type: %s %s\n",
          iommu_domain_type_str(iommu_def_domain_type),
          (iommu_cmd_line & IOMMU_CMD_LINE_DMA_API) ?
@@ -352,15 +355,6 @@ void iommu_set_dma_strict(bool strict)
          iommu_dma_strict = strict;
  }
-bool iommu_get_dma_strict(struct iommu_domain *domain)
-{
-    /* only allow lazy flushing for DMA domains */
-    if (domain->type == IOMMU_DOMAIN_DMA)
-        return iommu_dma_strict;
-    return true;
-}
-EXPORT_SYMBOL_GPL(iommu_get_dma_strict);
-
  static ssize_t iommu_group_attr_show(struct kobject *kobj,
                       struct attribute *__attr, char *buf)
  {
@@ -764,7 +758,7 @@ static int iommu_create_device_direct_mappings(struct iommu_group *group,
      unsigned long pg_size;
      int ret = 0;
-    if (!domain || domain->type != IOMMU_DOMAIN_DMA)
+    if (!domain || !(domain->type & __IOMMU_DOMAIN_DMA_API))

Nit: probably move above change to patch 14?

Indeed I'm not sure why this one ended up here, good catch!

Thanks,
Robin.

          return 0;
      BUG_ON(!domain->pgsize_bitmap);
diff --git a/include/linux/iommu.h b/include/linux/iommu.h
index 56519110d43f..557c4c12e2cf 100644
--- a/include/linux/iommu.h
+++ b/include/linux/iommu.h
@@ -484,7 +484,6 @@ int iommu_set_pgtable_quirks(struct iommu_domain *domain,
          unsigned long quirks);
  void iommu_set_dma_strict(bool val);
-bool iommu_get_dma_strict(struct iommu_domain *domain);
  extern int report_iommu_fault(struct iommu_domain *domain, struct device *dev,
                    unsigned long iova, int flags);


Best regards,
baolu