iommu: Don't warn prematurely about dodgy probes

The warning for suspect probe conditions inadvertently got moved too
early in a prior respin - it happened to work out OK for fwspecs, but in
general still needs to be after the ops->probe_device call so drivers
which filter devices for themselves have a chance to do that.

Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Fixes: bcb81ac ("iommu: Get DT/ACPI parsing into the proper probe path")
Link: https://lore.kernel.org/r/72a4853e7ef36e7c1c4ca171ed4ed8e1a463a61a.1741791691.git.robin.murphy@arm.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>

drivers/iommu/iommu.c

@@ -441,12 +441,6 @@ static int iommu_init_device(struct device *dev)
 		ret = -ENODEV;
 		goto err_free;
 	}
-	/*
-	 * And if we do now see any replay calls, they would indicate someone
-	 * misusing the dma_configure path outside bus code.
-	 */
-	if (dev->driver)
-		dev_WARN(dev, "late IOMMU probe at driver bind, something fishy here!\n");
 
 	if (!try_module_get(ops->owner)) {
 		ret = -EINVAL;
@@ -569,6 +563,12 @@ static int __iommu_probe_device(struct device *dev, struct list_head *group_list
 	ret = iommu_init_device(dev);
 	if (ret)
 		return ret;
+	/*
+	 * And if we do now see any replay calls, they would indicate someone
+	 * misusing the dma_configure path outside bus code.
+	 */
+	if (dev->driver)
+		dev_WARN(dev, "late IOMMU probe at driver bind, something fishy here!\n");
 
 	group = dev->iommu_group;
 	gdev = iommu_group_alloc_device(group, dev);