Skip to contentSkip to conversation

iommu/arm-smmu-v3: Factor out CMDQ batch force-sync conditions

PATCH Superseded Ashish Mhetre 2 participants 10 messages Jul 9 – Jul 10

Subsystems (5)

v5 of 9 Superseded Latest is v9 →

lore
Share & tools
lore
#1 lore raw

PATCH v5 2/3 iommu/arm-smmu-v3: Introduce CFGI/TLBI-repeat workaround infrastructure

In reply to Ashish Mhetre

Tegra264 SMMU instances need every CFGI/TLBI command sequence issued twice, with the second issue executing only after the first issue's CMD_SYNC has completed:

TLBI/CFGI ... CMD_SYNC TLBI/CFGI ... CMD_SYNC

ATC_INV is not affected and must never be doubled.

Add arm_smmu_erratum_repeat_tlbi_cfgi_key static key and an arm_smmu_erratum_cmd_needs_repeating() helper in arm-smmu-v3.h that gates on the static key first and then range-checks the opcode (CFGI_STE .. ATC_INV), so subsequent changes wiring the workaround into the CMDQ submission and iommufd batching paths can share a single predicate.

Rename the existing arm_smmu_cmdq_issue_cmdlist() to __arm_smmu_cmdq_issue_cmdlist() and add a thin wrapper that re-issues the same cmdlist a second time when the predicate fires. Register the new condition with arm_smmu_cmdq_batch_force_sync() and add arm_vsmmu_can_batch_cmd() so iommufd batches split at every "needs repeating" transition.

No callers enable the static key yet, so this patch introduces no functional change. A subsequent change hooks the DT probe to enable the key on affected instances.

---
 .../arm/arm-smmu-v3/arm-smmu-v3-iommufd.c     | 15 +++++++-
 drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c   | 36 ++++++++++++++++---
 drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h   | 15 ++++++++
 3 files changed, 61 insertions(+), 5 deletions(-)
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3-iommufd.c +14 -1
index 1e9f7d2de344..11d22acae613 100644
@@ -350,6 +350,18 @@ static int arm_vsmmu_convert_user_cmd(struct arm_vsmmu *vsmmu,
 	return 0;
 }
 
+static bool arm_vsmmu_can_batch_cmd(struct arm_smmu_device *smmu,
+				    struct arm_vsmmu_invalidation_cmd *last,
+				    struct arm_vsmmu_invalidation_cmd *next)
+{
+	struct arm_smmu_cmd next_cmd = {
+		.data[0] = le64_to_cpu(next->ucmd.cmd[0]),
+	};
+
+	return arm_smmu_erratum_cmd_needs_repeating(smmu, &last->cmd) ==
+	       arm_smmu_erratum_cmd_needs_repeating(smmu, &next_cmd);
+}
+
 int arm_vsmmu_cache_invalidate(struct iommufd_viommu *viommu,
 			       struct iommu_user_data_array *array)
 {
@@ -382,7 +394,8 @@ int arm_vsmmu_cache_invalidate(struct iommufd_viommu *viommu,
 
 		/* FIXME work in blocks of CMDQ_BATCH_ENTRIES and copy each block? */
 		cur++;
-		if (cur != end && (cur - last) != CMDQ_BATCH_ENTRIES - 1)
+		if (cur != end && (cur - last) != CMDQ_BATCH_ENTRIES - 1 &&
+		    arm_vsmmu_can_batch_cmd(smmu, last, cur))
 			continue;
 
 		/* FIXME always uses the main cmdq rather than trying to group by type */
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +32 -4
index 76efe479e80f..15b9d0170520 100644
@@ -41,6 +41,7 @@ MODULE_PARM_DESC(disable_msipolling,
 
 static const struct iommu_ops arm_smmu_ops;
 static struct iommu_dirty_ops arm_smmu_dirty_ops;
+DEFINE_STATIC_KEY_FALSE(arm_smmu_erratum_repeat_tlbi_cfgi_key);
 
 enum arm_smmu_msi_index {
 	EVTQ_MSI_INDEX,
@@ -698,10 +699,10 @@ static void arm_smmu_cmdq_write_entries(struct arm_smmu_cmdq *cmdq,
  *   insert their own list of commands then all of the commands from one
  *   CPU will appear before any of the commands from the other CPU.
  */
-int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu,
-				struct arm_smmu_cmdq *cmdq,
-				struct arm_smmu_cmd *cmds, int n,
-				bool sync)
+static int __arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu,
+					 struct arm_smmu_cmdq *cmdq,
+					 struct arm_smmu_cmd *cmds, int n,
+					 bool sync)
 {
 	struct arm_smmu_cmd cmd_sync;
 	u32 prod;
@@ -820,6 +821,28 @@ int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu,
 	return ret;
 }
 
+int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu,
+				struct arm_smmu_cmdq *cmdq,
+				struct arm_smmu_cmd *cmds, int n,
+				bool sync)
+{
+	int ret = __arm_smmu_cmdq_issue_cmdlist(smmu, cmdq, cmds, n, sync);
+
+	/*
+	 * arm_smmu_cmdq_batch_add_cmd_p() can flush its current batch
+	 * with sync=true and n=0 (bare SYNC) when the next command is
+	 * not supported by the batch's pre-selected cmdq, so the
+	 * repeat path must not inspect cmds[0].
+	 */
+	if (!n || ret || !sync)
+		return ret;
+
+	if (arm_smmu_erratum_cmd_needs_repeating(smmu, &cmds[0]))
+		ret = __arm_smmu_cmdq_issue_cmdlist(smmu, cmdq, cmds, n, sync);
+
+	return ret;
+}
+
 static int arm_smmu_cmdq_issue_cmd_p(struct arm_smmu_device *smmu,
 				     struct arm_smmu_cmd *cmd, bool sync)
 {
@@ -860,6 +883,11 @@ static bool arm_smmu_cmdq_batch_force_sync(struct arm_smmu_device *smmu,
 	    (smmu->options & ARM_SMMU_OPT_CMDQ_FORCE_SYNC))
 		return true;
 
+	/* See arm_smmu_erratum_cmd_needs_repeating() */
+	if (cmds->num == CMDQ_BATCH_ENTRIES &&
+	    arm_smmu_erratum_cmd_needs_repeating(smmu, &cmds->cmds[0]))
+		return true;
+
 	return false;
 }
 
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h +15 -0
index c909c9a88538..31121d7c7841 100644
@@ -11,6 +11,7 @@
 #include <linux/bitfield.h>
 #include <linux/iommu.h>
 #include <linux/iommufd.h>
+#include <linux/jump_label.h>
 #include <linux/kernel.h>
 #include <linux/mmzone.h>
 #include <linux/sizes.h>
@@ -1211,6 +1212,20 @@ int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu,
 				struct arm_smmu_cmdq *cmdq,
 				struct arm_smmu_cmd *cmds, int n,
 				bool sync);
+DECLARE_STATIC_KEY_FALSE(arm_smmu_erratum_repeat_tlbi_cfgi_key);
+
+static inline bool
+arm_smmu_erratum_cmd_needs_repeating(struct arm_smmu_device *smmu,
+				     struct arm_smmu_cmd *cmd)
+{
+	u8 opcode;
+
+	if (!static_branch_unlikely(&arm_smmu_erratum_repeat_tlbi_cfgi_key))
+		return false;
+
+	opcode = FIELD_GET(CMDQ_0_OP, cmd->data[0]);
+	return opcode >= CMDQ_OP_CFGI_STE && opcode < CMDQ_OP_ATC_INV;
+}
 
 #ifdef CONFIG_ARM_SMMU_V3_SVA
 bool arm_smmu_sva_supported(struct arm_smmu_device *smmu);
Signature · 1 line
2.50.1
#0 lore raw

PATCH v5 1/3 iommu/arm-smmu-v3: Factor out CMDQ batch force-sync conditions

From: Nicolin Chen <[email protected]>

arm_smmu_cmdq_batch_add_cmd_p() carries two distinct reasons for flushing the current batch with a CMD_SYNC before appending the new command:

- The batch's pre-assigned cmdq does not support the new command.
- The Arm erratum 2812531 workaround (ARM_SMMU_OPT_CMDQ_FORCE_SYNC) forces a SYNC at one entry before the batch is full.

Lift those checks into a new arm_smmu_cmdq_batch_force_sync() helper so that adding another force-sync condition becomes a one-line addition. No functional change.

---
 drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 23 +++++++++++++++------
 1 file changed, 17 insertions(+), 6 deletions(-)
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +17 -6
index a10affb483a4..76efe479e80f 100644
@@ -847,16 +847,27 @@ static void arm_smmu_cmdq_batch_init_cmd(struct arm_smmu_device *smmu,
 	cmds->cmdq = arm_smmu_get_cmdq(smmu, cmd);
 }
 
+static bool arm_smmu_cmdq_batch_force_sync(struct arm_smmu_device *smmu,
+					   struct arm_smmu_cmdq_batch *cmds,
+					   struct arm_smmu_cmd *cmd)
+{
+	/* The batch's pre-assigned cmdq doesn't support the new command */
+	if (!arm_smmu_cmdq_supports_cmd(cmds->cmdq, cmd))
+		return true;
+
+	/* Arm erratum 2812531 */
+	if (cmds->num == CMDQ_BATCH_ENTRIES - 1 &&
+	    (smmu->options & ARM_SMMU_OPT_CMDQ_FORCE_SYNC))
+		return true;
+
+	return false;
+}
+
 static void arm_smmu_cmdq_batch_add_cmd_p(struct arm_smmu_device *smmu,
 					  struct arm_smmu_cmdq_batch *cmds,
 					  struct arm_smmu_cmd *cmd)
 {
-	bool force_sync = (cmds->num == CMDQ_BATCH_ENTRIES - 1) &&
-			  (smmu->options & ARM_SMMU_OPT_CMDQ_FORCE_SYNC);
-	bool unsupported_cmd;
-
-	unsupported_cmd = !arm_smmu_cmdq_supports_cmd(cmds->cmdq, cmd);
-	if (force_sync || unsupported_cmd) {
+	if (arm_smmu_cmdq_batch_force_sync(smmu, cmds, cmd)) {
 		arm_smmu_cmdq_issue_cmdlist(smmu, cmds->cmdq, cmds->cmds,
 					    cmds->num, true);
 		arm_smmu_cmdq_batch_init_cmd(smmu, cmds, cmd);
Signature · 1 line
2.50.1
#4 lore raw

PATCH v5 3/3 iommu/arm-smmu-v3: Enable CFGI/TLBI-repeat workaround on Tegra264

Nvidia Tegra264 SMMU is affected by an erratum where a TLB entry can survive an invalidation that races with concurrent traffic targeting the same entry. The hardware-recommended software workaround is to issue every CFGI/TLBI command (each followed by CMD_SYNC) twice, and that infrastructure is already in place behind arm_smmu_erratum_repeat_tlbi_cfgi_key.

Neither IDR nor IIDR flags this Tegra264-specific bug, so hardware detection is not possible. Tegra264 is device-tree-only (no ACPI/IORT support) and already has a dedicated "nvidia,tegra264-smmu" compatible, so DT-probe is the only viable detection path.

Enable the workaround on instances matching the existing "nvidia,tegra264-smmu" compatible by calling static_branch_enable() on arm_smmu_erratum_repeat_tlbi_cfgi_key. Document the erratum in Documentation/arch/arm64/silicon-errata.rst.

---
 Documentation/arch/arm64/silicon-errata.rst | 2 ++
 drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 4 +++-
 2 files changed, 5 insertions(+), 1 deletion(-)
Documentation/arch/arm64/silicon-errata.rst +2 -0
index 014aa1c215a1..076b3947d259 100644
@@ -312,6 +312,8 @@ stable kernels.
 |                |                 | T241-MPAM-4,    |                             |
 |                |                 | T241-MPAM-6     |                             |
 +----------------+-----------------+-----------------+-----------------------------+
+| NVIDIA         | T264 SMMU       | T264-SMMU-3     | N/A                         |
++----------------+-----------------+-----------------+-----------------------------+
 +----------------+-----------------+-----------------+-----------------------------+
 | Freescale/NXP  | LS2080A/LS1043A | A-008585        | FSL_ERRATUM_A008585         |
 +----------------+-----------------+-----------------+-----------------------------+
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +3 -1
index 15b9d0170520..edb7a5d38cf9 100644
@@ -5331,8 +5331,10 @@ static int arm_smmu_device_dt_probe(struct platform_device *pdev,
 	if (of_dma_is_coherent(dev->of_node))
 		smmu->features |= ARM_SMMU_FEAT_COHERENCY;
 
-	if (of_device_is_compatible(dev->of_node, "nvidia,tegra264-smmu"))
+	if (of_device_is_compatible(dev->of_node, "nvidia,tegra264-smmu")) {
 		tegra_cmdqv_dt_probe(dev->of_node, smmu);
+		static_branch_enable(&arm_smmu_erratum_repeat_tlbi_cfgi_key);
+	}
 
 	return ret;
 }
Signature · 1 line
2.50.1
#7 lore raw

v5 cover iommu/arm-smmu-v3: Tegra264 invalidation workaround

In reply to Ashish Mhetre

Nvidia Tegra264 SMMUs are affected by an erratum where a TLB entry can survive an invalidation that races with concurrent traffic targeting the same entry. The hardware-recommended software workaround is to issue every CFGI/TLBI command (each followed by CMD_SYNC) twice. The second issue must execute only after the first issue's CMD_SYNC has completed, giving the sequence:

TLBI/CFGI ... CMD_SYNC TLBI/CFGI ... CMD_SYNC

ATC_INV is not affected and must not be doubled.

The erratum is not flagged by any SMMUv3 IDR/IIDR register, so it cannot be detected from hardware ID. Tegra264 is device-tree-only (no ACPI/IORT support), so detection is purely by compatible string.

This series is structured as a small refactor + infrastructure + enable sequence so that each step is reviewable in isolation:

1/3 Pure refactor (no functional change): lift the existing
force-sync conditions out of arm_smmu_cmdq_batch_add_cmd_p() into a new arm_smmu_cmdq_batch_force_sync() helper, so that adding another condition (in patch 2) is a one-line addition. Authored by Nicolin Chen.

2/3 Add the workaround infrastructure without enabling it. Defines
the arm_smmu_erratum_repeat_tlbi_cfgi_key static key, the shared arm_smmu_erratum_cmd_needs_repeating() predicate, the arm_smmu_cmdq_issue_cmdlist() wrapper that can re-issue matching cmdlists, the batch-helper force-sync condition, and the iommufd batching split for mixed command classes.

3/3 Enable the workaround for the existing "nvidia,tegra264-smmu"
compatible and document the erratum in silicon-errata.rst.

The series applies cleanly on linux-next/master (base-commit below).

Changes since v4:
- Drop ARM_SMMU_OPT_REPEAT_TLBI_CFGI entirely: the option bit was set and read on the exact same "nvidia,tegra264-smmu" compatible as the static key, so it added no per-instance signal that the static key did not already carry. The predicate now gates purely on arm_smmu_erratum_repeat_tlbi_cfgi_key. (Nicolin: "Are they redundant? / Drop the deadcode then.")
- Reorder the series so the compatible-string detection lands last, once all the infrastructure exists (Nicolin):

     1/3  factor out force_sync helper (unchanged)
     2/3  add static key + WAR functions (no functional change)
     3/3  enable the key on nvidia,tegra264-smmu + silicon-errata

Split the old v4 "Detect" and "Issue twice" patches accordingly.
- Update the /* See ARM_SMMU_OPT_REPEAT_TLBI_CFGI */ comment inside arm_smmu_cmdq_batch_force_sync() to reference arm_smmu_erratum_cmd_needs_repeating() instead.

Changes since v3:
- Drop the cmds->num == 0 early-return so the refactor is truly "no functional change".
- Rename ARM_SMMU_OPT_TLBI_TWICE -> ARM_SMMU_OPT_REPEAT_TLBI_CFGI and rephrase its kdoc to be hardware-agnostic.
- Rename arm_smmu_cmd_needs_tlbi_twice() -> arm_smmu_erratum_cmd_needs_repeating() and drop the kdoc above it.
- Replace the explicit opcode switch with a single range check opcode >= CMDQ_OP_CFGI_STE && opcode < CMDQ_OP_ATC_INV.
- Introduce arm_smmu_erratum_repeat_tlbi_cfgi_key static key:
the predicate gates on it first so unaffected kernels pay only a single static_branch_unlikely() check.
- Drop the verbose Tegra264-specific comments above arm_vsmmu_can_batch_cmd() and inside the batch helper.
- Document the erratum in Documentation/arch/arm64/silicon-errata.rst.
- Guard the repeat path in arm_smmu_cmdq_issue_cmdlist() with an n > 0 check so we never inspect cmds[0] on the bare-SYNC flush emitted by arm_smmu_cmdq_batch_add_cmd_p() when the next command is unsupported by the batch's pre-selected cmdq.
- Drop the carried Reviewed-by tags now that the patch shape has changed; re-review appreciated.

Changes since v2:
- Split into a 3-patch series (refactor / detect / apply) to keep each step small and bisectable.
- Move the classifier to arm-smmu-v3.h as static inline so the iommufd file can share it.
- Add arm_vsmmu_can_batch_cmd() to split iommufd batches at "needs repeating" transitions so the per-batch decision based on the first command stays correct under mixed user input.
- Spell out in the commit message why detection is via DT and not via IIDR/ACPI.

Changes since v1:
- Detect the erratum from the existing "nvidia,tegra264-smmu" compatible instead of adding a new property.
- Centralise the doubling at the CMDQ submission layer and only apply it to CFGI/TLBI (not ATC_INV).
- Drop the binding/dtsi patches accordingly.

Ashish Mhetre (2):
iommu/arm-smmu-v3: Introduce CFGI/TLBI-repeat workaround
infrastructure
iommu/arm-smmu-v3: Enable CFGI/TLBI-repeat workaround on Tegra264

Nicolin Chen (1):
iommu/arm-smmu-v3: Factor out CMDQ batch force-sync conditions

 Documentation/arch/arm64/silicon-errata.rst   |  2 +
 .../arm/arm-smmu-v3/arm-smmu-v3-iommufd.c     | 15 ++++-
 drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c   | 63 +++++++++++++++----
 drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h   | 15 +++++
 4 files changed, 83 insertions(+), 12 deletions(-)

base-commit: b9810cd75b9fb56a3425d391cba3f608502bd474

Signature · 1 line
2.50.1
#2 lore raw

Re: [PATCH v5 2/3] iommu/arm-smmu-v3: Introduce CFGI/TLBI-repeat workaround infrastructure

In reply to Ashish Mhetre
Quoted 3 lines · Ashish Mhetre

No callers enable the static key yet, so this patch introduces no functional change. A subsequent change hooks the DT probe to enable the key on affected instances.

Nit: once a patch is merged, it's no longer a "patch" but "commit".

Instead,

No callers enable the static key yet, so there is no functional change. A subsequent change will enable the key on affected instances.

drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c MODULE_PARM_DESC(disable_msipolling, · line 44

Quoted 12 lines
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +1 -0
index 76efe479e80f..15b9d0170520 100644
@@ -41,6 +41,7 @@ MODULE_PARM_DESC(disable_msipolling,
 
 static const struct iommu_ops arm_smmu_ops;
 static struct iommu_dirty_ops arm_smmu_dirty_ops;
+DEFINE_STATIC_KEY_FALSE(arm_smmu_erratum_repeat_tlbi_cfgi_key);

It's an erratum; it should have an inline description, like the one you had at ARM_SMMU_OPT_TLBI_TWICE previously.

drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu, · line 834

Quoted 11 lines
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +11 -0
+int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu,
+				struct arm_smmu_cmdq *cmdq,
+				struct arm_smmu_cmd *cmds, int n,
+				bool sync)
+{
+	int ret = __arm_smmu_cmdq_issue_cmdlist(smmu, cmdq, cmds, n, sync);
+
+	/*
+	 * arm_smmu_cmdq_batch_add_cmd_p() can flush its current batch
+	 * with sync=true and n=0 (bare SYNC) when the next command is
+	 * not supported by the batch's pre-selected cmdq, so the

After a second thought, this "supported" case is not reachable.

When a batch is init-ed, it's given an opcode (same as cmd[0]'s).

If an opcode is not supported, it's init-ed with smmu->cmdq that must support all commands.

So, adding the !n is more like a defensive play. But the comments here would be misleading.

Quoted 15 lines
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +11 -0
+	 * repeat path must not inspect cmds[0].
+	 */
+	if (!n || ret || !sync)
+		return ret;
+
+	if (arm_smmu_erratum_cmd_needs_repeating(smmu, &cmds[0]))
+		ret = __arm_smmu_cmdq_issue_cmdlist(smmu, cmdq, cmds, n, sync);
+
+	return ret;
+}
+
 static int arm_smmu_cmdq_issue_cmd_p(struct arm_smmu_device *smmu,
 				     struct arm_smmu_cmd *cmd, bool sync)
 {

drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c static bool arm_smmu_cmdq_batch_force_sync(struct arm_smmu_device *smmu, · line 888

Quoted 6 lines
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +3 -0
@@ -860,6 +883,11 @@ static bool arm_smmu_cmdq_batch_force_sync(struct arm_smmu_device *smmu,
 	    (smmu->options & ARM_SMMU_OPT_CMDQ_FORCE_SYNC))
 		return true;
 
+	/* See arm_smmu_erratum_cmd_needs_repeating() */
+	if (cmds->num == CMDQ_BATCH_ENTRIES &&
+	    arm_smmu_erratum_cmd_needs_repeating(smmu, &cmds->cmds[0]))

The comment above is useless; arm_smmu_erratum_cmd_needs_repeating doesn't give any useful explanation either.

Like I suggested, add the missing piece of war description to the arm_smmu_erratum_repeat_tlbi_cfgi_key at the top. Then,

	/* See the description at arm_smmu_erratum_repeat_tlbi_cfgi_key */

drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu, · line 1219

Quoted 10 lines
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h +5 -0
@@ -1211,6 +1212,20 @@ int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu,
 				struct arm_smmu_cmdq *cmdq,
 				struct arm_smmu_cmd *cmds, int n,
 				bool sync);
+DECLARE_STATIC_KEY_FALSE(arm_smmu_erratum_repeat_tlbi_cfgi_key);
+
+static inline bool
+arm_smmu_erratum_cmd_needs_repeating(struct arm_smmu_device *smmu,
+				     struct arm_smmu_cmd *cmd)

smmu is unused.

drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu, · line 1228

Quoted 9 lines
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h +9 -0
+{
+	u8 opcode;
+
+	if (!static_branch_unlikely(&arm_smmu_erratum_repeat_tlbi_cfgi_key))
+		return false;
+
+	opcode = FIELD_GET(CMDQ_0_OP, cmd->data[0]);
+	return opcode >= CMDQ_OP_CFGI_STE && opcode < CMDQ_OP_ATC_INV;
+}

Maybe move this to the first caller in arm-smmu-v3.c? Instead, add its stub here, so we can drop that DECLARE_STATIC_KEY_FALSE. Then, DEFINE_STATIC_KEY_FALSE will need a "static".

Nicolin

Reviewed-by on v5 3/3
#5 lore raw

Re: [PATCH v5 3/3] iommu/arm-smmu-v3: Enable CFGI/TLBI-repeat workaround on Tegra264

In reply to Ashish Mhetre
Quoted 12 lines · Ashish Mhetre

Nvidia Tegra264 SMMU is affected by an erratum where a TLB entry can survive an invalidation that races with concurrent traffic targeting the same entry. The hardware-recommended software workaround is to issue every CFGI/TLBI command (each followed by CMD_SYNC) twice, and that infrastructure is already in place behind arm_smmu_erratum_repeat_tlbi_cfgi_key.

Neither IDR nor IIDR flags this Tegra264-specific bug, so hardware detection is not possible. Tegra264 is device-tree-only (no ACPI/IORT support) and already has a dedicated "nvidia,tegra264-smmu" compatible, so DT-probe is the only viable detection path.

Quoted 6 lines

Enable the workaround on instances matching the existing "nvidia,tegra264-smmu" compatible by calling static_branch_enable() on arm_smmu_erratum_repeat_tlbi_cfgi_key. Document the erratum in Documentation/arch/arm64/silicon-errata.rst.

Sashiko pointed out a concern at PATCH-3 regarding the static key:
https://sashiko.dev/#/patchset/20260709095613.831769-1-amhetre%40nvidia.com

It's a false positive. But perhaps we could fold in an inline note; it'd belong to the missing description that I commented in PATCH-2.

Nicolin

#8 lore raw
In reply to Ashish Mhetre
Quoted 6 lines · Ashish Mhetre

From: Nicolin Chen <[email protected]>

arm_smmu_cmdq_batch_add_cmd_p() carries two distinct reasons for flushing the current batch with a CMD_SYNC before appending the new command:

Quoted 7 lines

- The batch's pre-assigned cmdq does not support the new command.
- The Arm erratum 2812531 workaround (ARM_SMMU_OPT_CMDQ_FORCE_SYNC) forces a SYNC at one entry before the batch is full.

Lift those checks into a new arm_smmu_cmdq_batch_force_sync() helper so that adding another force-sync condition becomes a one-line addition. No functional change.

Patch is fine, but you're missing the cover-letter and changelogs.

Nicolin

#3 lore raw

Re: [PATCH v5 2/3] iommu/arm-smmu-v3: Introduce CFGI/TLBI-repeat workaround infrastructure

In reply to Nicolin Chen
Quoted 10 lines · Nicolin Chen
Quoted 3 lines · Ashish Mhetre

No callers enable the static key yet, so this patch introduces no functional change. A subsequent change hooks the DT probe to enable the key on affected instances.

Nit: once a patch is merged, it's no longer a "patch" but "commit".

Instead,

No callers enable the static key yet, so there is no functional change. A subsequent change will enable the key on affected instances.

Ack.

Quoted 8 lines
Quoted 8 lines
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +0 -0
index 76efe479e80f..15b9d0170520 100644
@@ -41,6 +41,7 @@ MODULE_PARM_DESC(disable_msipolling,
 
Quoted 5 lines

drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c MODULE_PARM_DESC(disable_msipolling, · line 44

Quoted 3 lines
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +1 -0
  static const struct iommu_ops arm_smmu_ops;
  static struct iommu_dirty_ops arm_smmu_dirty_ops;
+DEFINE_STATIC_KEY_FALSE(arm_smmu_erratum_repeat_tlbi_cfgi_key);

It's an erratum; it should have an inline description, like the one you had at ARM_SMMU_OPT_TLBI_TWICE previously.

Okay, I'll restore the description I had for option bit here.

Quoted 13 lines

drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu, · line 834

Quoted 11 lines
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +11 -0
+int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu,
+				struct arm_smmu_cmdq *cmdq,
+				struct arm_smmu_cmd *cmds, int n,
+				bool sync)
+{
+	int ret = __arm_smmu_cmdq_issue_cmdlist(smmu, cmdq, cmds, n, sync);
+
+	/*
+	 * arm_smmu_cmdq_batch_add_cmd_p() can flush its current batch
+	 * with sync=true and n=0 (bare SYNC) when the next command is
+	 * not supported by the batch's pre-selected cmdq, so the

After a second thought, this "supported" case is not reachable.

Quoted 7 lines

When a batch is init-ed, it's given an opcode (same as cmd[0]'s).

If an opcode is not supported, it's init-ed with smmu->cmdq that must support all commands.

So, adding the !n is more like a defensive play. But the comments here would be misleading.

You're right, the unsupported-cmdq force-sync path cannot run with n == 0. I'll fix the misleading comment. The !n guard stays as a safety check for empty batch_submit(), where we must not inspect cmds[0].

Quoted 24 lines
Quoted 15 lines
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +11 -0
+	 * repeat path must not inspect cmds[0].
+	 */
+	if (!n || ret || !sync)
+		return ret;
+
+	if (arm_smmu_erratum_cmd_needs_repeating(smmu, &cmds[0]))
+		ret = __arm_smmu_cmdq_issue_cmdlist(smmu, cmdq, cmds, n, sync);
+
+	return ret;
+}
+
  static int arm_smmu_cmdq_issue_cmd_p(struct arm_smmu_device *smmu,
  				     struct arm_smmu_cmd *cmd, bool sync)
  {

drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c static bool arm_smmu_cmdq_batch_force_sync(struct arm_smmu_device *smmu, · line 888

Quoted 6 lines
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +3 -0
@@ -860,6 +883,11 @@ static bool arm_smmu_cmdq_batch_force_sync(struct arm_smmu_device *smmu,
  	    (smmu->options & ARM_SMMU_OPT_CMDQ_FORCE_SYNC))
  		return true;
 
+	/* See arm_smmu_erratum_cmd_needs_repeating() */
+	if (cmds->num == CMDQ_BATCH_ENTRIES &&
+	    arm_smmu_erratum_cmd_needs_repeating(smmu, &cmds->cmds[0]))

The comment above is useless; arm_smmu_erratum_cmd_needs_repeating doesn't give any useful explanation either.

Quoted 4 lines

Like I suggested, add the missing piece of war description to the arm_smmu_erratum_repeat_tlbi_cfgi_key at the top. Then,

	/* See the description at arm_smmu_erratum_repeat_tlbi_cfgi_key */

Ack.

Quoted 11 lines

drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu, · line 1219

Quoted 10 lines
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h +5 -0
@@ -1211,6 +1212,20 @@ int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu,
  				struct arm_smmu_cmdq *cmdq,
  				struct arm_smmu_cmd *cmds, int n,
  				bool sync);
+DECLARE_STATIC_KEY_FALSE(arm_smmu_erratum_repeat_tlbi_cfgi_key);
+
+static inline bool
+arm_smmu_erratum_cmd_needs_repeating(struct arm_smmu_device *smmu,
+				     struct arm_smmu_cmd *cmd)

smmu is unused.

Thanks for pointing. With static key, smmu pointer is not used anymore. I'll remove it.

Quoted 13 lines

drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h int arm_smmu_cmdq_issue_cmdlist(struct arm_smmu_device *smmu, · line 1228

Quoted 9 lines
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h +9 -0
+{
+	u8 opcode;
+
+	if (!static_branch_unlikely(&arm_smmu_erratum_repeat_tlbi_cfgi_key))
+		return false;
+
+	opcode = FIELD_GET(CMDQ_0_OP, cmd->data[0]);
+	return opcode >= CMDQ_OP_CFGI_STE && opcode < CMDQ_OP_ATC_INV;
+}

Maybe move this to the first caller in arm-smmu-v3.c? Instead, add its stub here, so we can drop that DECLARE_STATIC_KEY_FALSE. Then, DEFINE_STATIC_KEY_FALSE will need a "static".

Quoted 1 line

Nicolin

Makes sense as iommufd only needs the classifier, not the key itself. I'll move arm_smmu_erratum_cmd_needs_repeating() into arm-smmu-v3.c, leave a declaration in the header.

Thanks,
Ashish Mhetre

#6 lore raw

Re: [PATCH v5 3/3] iommu/arm-smmu-v3: Enable CFGI/TLBI-repeat workaround on Tegra264

In reply to Nicolin Chen
Quoted 21 lines · Nicolin Chen
Quoted 12 lines · Ashish Mhetre

Nvidia Tegra264 SMMU is affected by an erratum where a TLB entry can survive an invalidation that races with concurrent traffic targeting the same entry. The hardware-recommended software workaround is to issue every CFGI/TLBI command (each followed by CMD_SYNC) twice, and that infrastructure is already in place behind arm_smmu_erratum_repeat_tlbi_cfgi_key.

Neither IDR nor IIDR flags this Tegra264-specific bug, so hardware detection is not possible. Tegra264 is device-tree-only (no ACPI/IORT support) and already has a dedicated "nvidia,tegra264-smmu" compatible, so DT-probe is the only viable detection path.

Quoted 6 lines

Enable the workaround on instances matching the existing "nvidia,tegra264-smmu" compatible by calling static_branch_enable() on arm_smmu_erratum_repeat_tlbi_cfgi_key. Document the erratum in Documentation/arch/arm64/silicon-errata.rst.

Quoted 7 lines

Sashiko pointed out a concern at PATCH-3 regarding the static key:
https://sashiko.dev/#/patchset/20260709095613.831769-1-amhetre%40nvidia.com

It's a false positive. But perhaps we could fold in an inline note; it'd belong to the missing description that I commented in PATCH-2.

Nicolin

Ack, I will address this in v6.

Thanks,
Ashish Mhetre

#9 lore raw
In reply to Nicolin Chen
Quoted 11 lines · Nicolin Chen
Quoted 10 lines · Ashish Mhetre

From: Nicolin Chen <[email protected]>

arm_smmu_cmdq_batch_add_cmd_p() carries two distinct reasons for flushing the current batch with a CMD_SYNC before appending the new command:

- The batch's pre-assigned cmdq does not support the new command.
- The Arm erratum 2812531 workaround (ARM_SMMU_OPT_CMDQ_FORCE_SYNC) forces a SYNC at one entry before the batch is full.

Quoted 6 lines
Quoted 3 lines

Lift those checks into a new arm_smmu_cmdq_batch_force_sync() helper so that adding another force-sync condition becomes a one-line addition. No functional change.

Patch is fine, but you're missing the cover-letter and changelogs.

Nicolin

I had sent out the cover-letter along with change logs. The sequence got messed up while sending. Here's the cover letter.
https://lore.kernel.org/linux-tegra/[email protected]/T/#u

I will fix this in next version.

Thanks,
Ashish Mhetre