On 21/06/2021 12:59, Robin Murphy wrote:

+ Nadav

quoted

On a personal level I would be happy with that approach, but I think
it's better to not start changing things right away in a follow-up series.

So how about we add this patch (which replaces 6/6 "iommu: Remove mode
argument from iommu_set_dma_strict()")?

Robin, any opinion?

For me it boils down to whether there are any realistic workloads where
non-strict mode*would*  still perform better under virtualisation. The
only reason for the user to explicitly pass "iommu.strict=0" is because
they expect it to increase unmap performance; if it's only ever going to
lead to an unexpected performance loss, I don't see any value in
overriding the kernel's decision purely for the sake of subservience.

If there*are*  certain valid cases for allowing it for people who really
know what they're doing, then we should arguably also log a counterpart
message to say "we're honouring your override but beware it may have the
opposite effect to what you expect" for the benefit of other users who
assume it's a generic go-faster knob. At that point it starts getting
non-trivial enough that I'd want to know for sure it's worthwhile.

The other reason this might be better to revisit later is that an AMD
equivalent is still in flight[1], and there might be more that can
eventually be factored out. I think both series are pretty much good to
merge for 5.14, but time's already tight to sort out the conflicts which
exist as-is, without making them any worse.

ok, fine. Can revisit.

As for getting these merged, I'll dry-run merging both of those series 
to see the conflicts. It doesn't look too problematic from a glance.

Cheers,
John

Robin.

[1]
https://lore.kernel.org/linux-iommu/20210616100500.174507-3-namit@vmware.com/ (local)

quoted

------->8---------

[PATCH] iommu/vt-d: Make "iommu.strict" override batching due to
   virtualization

As a change in policy, make iommu.strict cmdline argument override
whether we disable batching due to virtualization.

The API of iommu_set_dma_strict() is changed to accept a "force"
argument, which means that we always set iommu_dma_strict true,
regardless of whether we already set via cmdline. Also return a boolean,
to tell whether iommu_dma_strict was set or not.

Note that in all pre-existing callsites of iommu_set_dma_strict(),
argument strict was true, so this argument is dropped.

Signed-off-by: John Garry<redacted>

diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c
index 06666f9d8116..e8d65239b359 100644
--- a/drivers/iommu/intel/iommu.c
+++ b/drivers/iommu/intel/iommu.c

@@ -4380,10 +4380,8 @@ int __init intel_iommu_init(void)
            * is likely to be much lower than the overhead of synchronizing
            * the virtual and physical IOMMU page-tables.
            */
-        if (cap_caching_mode(iommu->cap)) {
+        if (cap_caching_mode(iommu->cap) && iommu_set_dma_strict(false))
               pr_info_once("IOMMU batching disallowed due to

virtualization\n");
-            iommu_set_dma_strict(true);
-        }
           iommu_device_sysfs_add(&iommu->iommu, NULL,
                          intel_iommu_groups,
                          "%s", iommu->name);

diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
index 60b1ec42e73b..1434bee64af3 100644
--- a/drivers/iommu/iommu.c
+++ b/drivers/iommu/iommu.c

@@ -349,10 +349,14 @@ static int __init iommu_dma_setup(char *str)
   }
   early_param("iommu.strict", iommu_dma_setup);

-void iommu_set_dma_strict(bool strict)
+/* Return true if we set iommu_dma_strict */
+bool iommu_set_dma_strict(bool force)
   {
-    if (strict || !(iommu_cmd_line & IOMMU_CMD_LINE_STRICT))
-        iommu_dma_strict = strict;
+    if (force || !(iommu_cmd_line & IOMMU_CMD_LINE_STRICT)) {
+        iommu_dma_strict = true;
+        return true;
+    }
+    return false;
   }

   bool iommu_get_dma_strict(struct iommu_domain *domain)

diff --git a/include/linux/iommu.h b/include/linux/iommu.h
index 32d448050bf7..f17b20234296 100644
--- a/include/linux/iommu.h
+++ b/include/linux/iommu.h

@@ -476,7 +476,7 @@ int iommu_enable_nesting(struct iommu_domain *domain);
   int iommu_set_pgtable_quirks(struct iommu_domain *domain,
           unsigned long quirks);

-void iommu_set_dma_strict(bool val);
+bool iommu_set_dma_strict(bool force);
   bool iommu_get_dma_strict(struct iommu_domain *domain);

   extern int report_iommu_fault(struct iommu_domain *domain, struct

device *dev,