Re: [PATCH v9 02/12] iommu/arm-smmu-v3: Skip EVTQ/PRIQ setup in kdump kernel
From: Jason Gunthorpe <jgg@nvidia.com>
Date: 2026-08-21 23:15:54
Also in:
linux-iommu, lkml
quoted hunk ↗ jump to hunk
@@ -4821,21 +4821,36 @@ static int arm_smmu_device_reset(struct arm_smmu_device *smmu) arm_smmu_cmdq_issue_cmd_with_sync( smmu, arm_smmu_make_cmd_op(CMDQ_OP_TLBI_NSNH_ALL)); - /* Event queue */ - writeq_relaxed(smmu->evtq.q.q_base, smmu->base + ARM_SMMU_EVTQ_BASE); - writel_relaxed(smmu->evtq.q.llq.prod, smmu->page1 + ARM_SMMU_EVTQ_PROD); - writel_relaxed(smmu->evtq.q.llq.cons, smmu->page1 + ARM_SMMU_EVTQ_CONS); + /* + * Event queue + * + * Do not enable in a kdump case, as the crashed kernel's CDs and page + * tables may be corrupted, triggering event spamming. A disabled queue + * simply discards new events, without raising any global error. + */ + if (!is_kdump_kernel()) {
same remark about a FEAT_EVTQ
- /* PRI queue */
- if (smmu->features & ARM_SMMU_FEAT_PRI) {
+ /*
+ * PRI queue
+ *
+ * Do not enable in a kdump case, as we cannot serve page requests.
+ */
+ if (!is_kdump_kernel() && (smmu->features & ARM_SMMU_FEAT_PRI)) {
writeq_relaxed(smmu->priq.q.q_base,I'm reading this again and wondering why not drop FEAT_PRI much earlier so we don't need this is_kdump_kernel() here? The logic still looks OK to me -- Jason