Re: [PATCH v5 4/7] iommu/arm-smmu-v3: Pre-allocate a per-master invalidation array

[Nicolin Chen <nicolinc@nvidia.com>]

On Mon, Nov 24, 2025 at 03:31:17PM -0800, Nicolin Chen wrote:
> On Mon, Nov 24, 2025 at 07:08:45PM -0400, Jason Gunthorpe wrote:
> > On Mon, Nov 24, 2025 at 02:43:58PM -0800, Nicolin Chen wrote:
> > > On Mon, Nov 24, 2025 at 09:42:55PM +0000, Will Deacon wrote:
> > > > On Sat, Nov 08, 2025 at 12:08:05AM -0800, Nicolin Chen wrote:
> > > > > +		/* Put the ids into order for sorted to_merge/to_unref arrays */
> > > > > +		sort_nonatomic(fwspec->ids, fwspec->num_ids,
> > > > > +			       sizeof(fwspec->ids[0]), arm_smmu_ids_cmp, NULL);
> > > > > +		/* ATS case adds num_ids of entries, on top of the base case */
> > > > > +		master->build_invs = arm_smmu_invs_alloc(2 + fwspec->num_ids);
> > > > 
> > > > Although I can't point at a specific issue here, I'm nervous about mutating
> > > > the 'fwspec->ids' array from within the driver, The array isn't allocated
> > > > or populated directly by the driver and so I don't think we really have any
> > > > business sorting it. Could we hack iommu_fwspec_add_ids() to keep the array
> > > > ordered instead?
> > > 
> > > Yea, I think it makes sense to do it in the core, once we have the
> > > data structure provided by the core as well.
> > 
> > I would be more worried about sorting it everywhere for every
> > driver. I feel confident SMMUv3 doesn't use it, but something really
> > old and embedded focused like tegra or omap, IDK.
> > 
> > So I wouldn't propose to change iommu_fwspec_add_ids().
> 
> Perhaps a different helper iommu_fwspec_add_ids_sorted()? Drivers
> can choose to use the sorted version, when they want to implement
> the invalidation array. And SMMU can be the first only caller.
> 
> > If you want to be conservative then the thing to do is sort the
> > master->streams that arm_smmu_insert_master() copies the fwspec
> > into. It just has to be sorted prior to feeding it into the rbtree.
> > 
> > Then consistently use master->streams as the sorted list.
> 
> How about kmemdup() an local id array to bridge betwen fwspec->ids
> and rbtree?

Does this look okay?

@@ -3713,6 +3713,7 @@ static int arm_smmu_insert_master(struct arm_smmu_device *smmu,
 {
 	int i;
 	int ret = 0;
+	u32 *ids, *ids_sorted = NULL;
 	struct iommu_fwspec *fwspec = dev_iommu_fwspec_get(master->dev);
 	bool ats_supported = dev_is_pci(master->dev) &&
 			     pci_ats_supported(to_pci_dev(master->dev));
@@ -3723,26 +3724,38 @@ static int arm_smmu_insert_master(struct arm_smmu_device *smmu,
 		return -ENOMEM;
 	master->num_streams = fwspec->num_ids;
 
+	ids = fwspec->ids;
 	if (!ats_supported) {
 		/* Base case has 1 ASID entry or maximum 2 VMID entries */
 		master->build_invs = arm_smmu_invs_alloc(2);
 	} else {
 		/* Put the ids into order for sorted to_merge/to_unref arrays */
-		sort_nonatomic(fwspec->ids, fwspec->num_ids,
-			       sizeof(fwspec->ids[0]), arm_smmu_ids_cmp, NULL);
+		if (fwspec->num_ids > 1) {
+			ids = kmemdup_array(fwspec->ids, fwspec->num_ids,
+					    sizeof(*ids), GFP_KERNEL);
+			if (!ids) {
+				kfree(master->streams);
+				return -ENOMEM;
+			}
+
+			sort_nonatomic(ids, fwspec->num_ids, sizeof(*ids),
+				       arm_smmu_ids_cmp, NULL);
+			ids_sorted = ids;
+		}
 		/* ATS case adds num_ids of entries, on top of the base case */
 		master->build_invs = arm_smmu_invs_alloc(2 + fwspec->num_ids);
 	}
-	if (IS_ERR(master->build_invs)) {
+	if (!master->build_invs) {
 		kfree(master->streams);
-		return PTR_ERR(master->build_invs);
+		ret = -ENOMEM;
+		goto out;
 	}
 
 	mutex_lock(&smmu->streams_mutex);
 	for (i = 0; i < fwspec->num_ids; i++) {
 		struct arm_smmu_stream *new_stream = &master->streams[i];
 		struct rb_node *existing;
-		u32 sid = fwspec->ids[i];
+		u32 sid = ids[i];
 
 		new_stream->id = sid;
 		new_stream->master = master;
@@ -3779,6 +3792,8 @@ static int arm_smmu_insert_master(struct arm_smmu_device *smmu,
 	}
 	mutex_unlock(&smmu->streams_mutex);
 
+out:
+	kfree(ids_sorted);
 	return ret;
 }
 
Thanks
Nicolin