Skip to contentSkip to conversation

iommu/arm-smmu-v3: Cope with duplicated Stream IDs

PATCH New Robin Murphy 4 participants 4 messages Jan 2, 2018 IOMMU

lore
Share & tools
lore

Robin Murphy

author reviewer
#0 lore raw

PATCH 1/1 iommu/arm-smmu-v3: Cope with duplicated Stream IDs

For PCI devices behind an aliasing PCIe-to-PCI/X bridge, the bridge alias to DevFn 0.0 on the subordinate bus may match the original RID of the device, resulting in the same SID being present in the device's fwspec twice. This causes trouble later in arm_smmu_write_strtab_ent() when we wind up visiting the STE a second time and find it already live.

Avoid the issue by giving arm_smmu_install_ste_for_dev() the cleverness to skip over duplicates. It seems mildly counterintuitive compared to preventing the duplicates from existing in the first place, but since the DT and ACPI probe paths build their fwspecs differently, this is actually the cleanest and most self-contained way to deal with it.

Fixes 8f78515425da ("iommu/arm-smmu: Implement of_xlate() for SMMUv3")
---
 drivers/iommu/arm-smmu-v3.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/iommu/arm-smmu-v3.c +8 -1
index 9ce3cde575a8..57c92aa3122e 100644
@@ -1752,7 +1752,7 @@ static __le64 *arm_smmu_get_step_for_sid(struct arm_smmu_device *smmu, u32 sid)
 
 static void arm_smmu_install_ste_for_dev(struct iommu_fwspec *fwspec)
 {
-	int i;
+	int i, j;
 	struct arm_smmu_master_data *master = fwspec->iommu_priv;
 	struct arm_smmu_device *smmu = master->smmu;
 
@@ -1760,6 +1760,13 @@ static void arm_smmu_install_ste_for_dev(struct iommu_fwspec *fwspec)
 		u32 sid = fwspec->ids[i];
 		__le64 *step = arm_smmu_get_step_for_sid(smmu, sid);
 
+		/* Bridged PCI devices may end up with duplicated IDs */
+		for (j = 0; j < i; j++)
+			if (fwspec->ids[j] == sid)
+				break;
+		if (j < i)
+			continue;
+
 		arm_smmu_write_strtab_ent(smmu, sid, step, &master->ste);
 	}
 }
Signature · 1 line
2.13.4.dirty
#1 lore raw

PATCH 1/1 iommu/arm-smmu-v3: Cope with duplicated Stream IDs

Quoted 12 lines · Robin Murphy

For PCI devices behind an aliasing PCIe-to-PCI/X bridge, the bridge alias to DevFn 0.0 on the subordinate bus may match the original RID of the device, resulting in the same SID being present in the device's fwspec twice. This causes trouble later in arm_smmu_write_strtab_ent() when we wind up visiting the STE a second time and find it already live.

Avoid the issue by giving arm_smmu_install_ste_for_dev() the cleverness to skip over duplicates. It seems mildly counterintuitive compared to preventing the duplicates from existing in the first place, but since the DT and ACPI probe paths build their fwspecs differently, this is actually the cleanest and most self-contained way to deal with it.

Quoted 8 lines
Fixes 8f78515425da ("iommu/arm-smmu: Implement of_xlate() for SMMUv3")
---
 drivers/iommu/arm-smmu-v3.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

Cheers, Robin. I'll pick this up and send to Alex later this week along with a couple of other fixes I have kicking around.

Will

#2 lore raw

PATCH 1/1 iommu/arm-smmu-v3: Cope with duplicated Stream IDs

Quoted 22 lines · Will Deacon
Quoted 12 lines · Robin Murphy

For PCI devices behind an aliasing PCIe-to-PCI/X bridge, the bridge alias to DevFn 0.0 on the subordinate bus may match the original RID of the device, resulting in the same SID being present in the device's fwspec twice. This causes trouble later in arm_smmu_write_strtab_ent() when we wind up visiting the STE a second time and find it already live.

Avoid the issue by giving arm_smmu_install_ste_for_dev() the cleverness to skip over duplicates. It seems mildly counterintuitive compared to preventing the duplicates from existing in the first place, but since the DT and ACPI probe paths build their fwspecs differently, this is actually the cleanest and most self-contained way to deal with it.

Quoted 8 lines
Fixes 8f78515425da ("iommu/arm-smmu: Implement of_xlate() for SMMUv3")
---
 drivers/iommu/arm-smmu-v3.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
Quoted 2 lines

Cheers, Robin. I'll pick this up and send to Alex later this week along with a couple of other fixes I have kicking around.

Thanks. Even though the Fixes: tag points to an older commit, we really need the fix in 4.14 where another commit (10631d724def drm/pci:
Deprecate drm_pci_init/exit completely) exposed the issue.

It it is not already planned, can you please cc stable?

JC.

#3 lore raw

PATCH 1/1 iommu/arm-smmu-v3: Cope with duplicated Stream IDs

In reply to Robin Murphy

Hi Robin,

Thank you for fixing this.

Regards,
Tomasz

Quoted 49 lines · Robin Murphy

For PCI devices behind an aliasing PCIe-to-PCI/X bridge, the bridge alias to DevFn 0.0 on the subordinate bus may match the original RID of the device, resulting in the same SID being present in the device's fwspec twice. This causes trouble later in arm_smmu_write_strtab_ent() when we wind up visiting the STE a second time and find it already live.

Avoid the issue by giving arm_smmu_install_ste_for_dev() the cleverness to skip over duplicates. It seems mildly counterintuitive compared to preventing the duplicates from existing in the first place, but since the DT and ACPI probe paths build their fwspecs differently, this is actually the cleanest and most self-contained way to deal with it.

Fixes 8f78515425da ("iommu/arm-smmu: Implement of_xlate() for SMMUv3")
---
  drivers/iommu/arm-smmu-v3.c | 9 ++++++++-
  1 file changed, 8 insertions(+), 1 deletion(-)
drivers/iommu/arm-smmu-v3.c +8 -1
index 9ce3cde575a8..57c92aa3122e 100644
@@ -1752,7 +1752,7 @@ static __le64 *arm_smmu_get_step_for_sid(struct arm_smmu_device *smmu, u32 sid)
 
  static void arm_smmu_install_ste_for_dev(struct iommu_fwspec *fwspec)
  {
-	int i;
+	int i, j;
  	struct arm_smmu_master_data *master = fwspec->iommu_priv;
  	struct arm_smmu_device *smmu = master->smmu;
 
@@ -1760,6 +1760,13 @@ static void arm_smmu_install_ste_for_dev(struct iommu_fwspec *fwspec)
  		u32 sid = fwspec->ids[i];
  		__le64 *step = arm_smmu_get_step_for_sid(smmu, sid);
 
+		/* Bridged PCI devices may end up with duplicated IDs */
+		for (j = 0; j < i; j++)
+			if (fwspec->ids[j] == sid)
+				break;
+		if (j < i)
+			continue;
+
  		arm_smmu_write_strtab_ent(smmu, sid, step, &master->ste);
  	}
  }