passing intel_iommu=sm_on to the kernel. If not sure, please use
the default value.
+config INTEL_IOMMU_PERF_EVENTS
+ def_bool y
+ bool "Intel IOMMU performance events"
+ depends on INTEL_IOMMU && PERF_EVENTS
+ help
+ Selecting this option will enable the performance monitoring
+ infrastructure in the Intel IOMMU. It collects information about
+ key events occurring during operation of the remapping hardware,
+ to aid performance tuning and debug. These are available on modern
+ processors which support Intel VT-d 4.0 and later.
+
endif # INTEL_IOMMU
obj-$(CONFIG_INTEL_IOMMU_DEBUGFS) += debugfs.o
obj-$(CONFIG_INTEL_IOMMU_SVM) += svm.o
obj-$(CONFIG_IRQ_REMAP) += irq_remapping.o
+obj-$(CONFIG_INTEL_IOMMU_PERF_EVENTS) += perfmon.o
#include "../irq_remapping.h"
#include "perf.h"
#include "trace.h"
+#include "perfmon.h"
typedef int (*dmar_res_handler_t)(struct acpi_dmar_header *, void *);
struct dmar_res_callback {
if (sts & DMA_GSTS_QIES)
iommu->gcmd |= DMA_GCMD_QIE;
+ if (alloc_iommu_pmu(iommu))
+ pr_debug("Cannot alloc PMU for iommu (seq_id = %d)\n", iommu->seq_id);
+
raw_spin_lock_init(&iommu->register_lock);
/*
err_sysfs:
iommu_device_sysfs_remove(&iommu->iommu);
err_unmap:
+ free_iommu_pmu(iommu);
unmap_iommu(iommu);
error_free_seq_id:
ida_free(&dmar_seq_ids, iommu->seq_id);
iommu_device_sysfs_remove(&iommu->iommu);
}
+ free_iommu_pmu(iommu);
+
if (iommu->irq) {
if (iommu->pr_irq) {
free_irq(iommu->pr_irq, iommu);
#define DMAR_MTRR_PHYSMASK8_REG 0x208
#define DMAR_MTRR_PHYSBASE9_REG 0x210
#define DMAR_MTRR_PHYSMASK9_REG 0x218
+#define DMAR_PERFCAP_REG 0x300
+#define DMAR_PERFCFGOFF_REG 0x310
+#define DMAR_PERFOVFOFF_REG 0x318
+#define DMAR_PERFCNTROFF_REG 0x31c
+#define DMAR_PERFEVNTCAP_REG 0x380
#define DMAR_VCCAP_REG 0xe30 /* Virtual command capability register */
#define DMAR_VCMD_REG 0xe00 /* Virtual command register */
#define DMAR_VCRSP_REG 0xe10 /* Virtual command response register */
*/
#define cap_esrtps(c) (((c) >> 63) & 1)
#define cap_esirtps(c) (((c) >> 62) & 1)
+#define cap_ecmds(c) (((c) >> 61) & 1)
#define cap_fl5lp_support(c) (((c) >> 60) & 1)
#define cap_pi_support(c) (((c) >> 59) & 1)
#define cap_fl1gp_support(c) (((c) >> 56) & 1)
* Extended Capability Register
*/
-#define ecap_rps(e) (((e) >> 49) & 0x1)
+#define ecap_pms(e) (((e) >> 51) & 0x1)
+#define ecap_rps(e) (((e) >> 49) & 0x1)
#define ecap_smpwc(e) (((e) >> 48) & 0x1)
#define ecap_flts(e) (((e) >> 47) & 0x1)
#define ecap_slts(e) (((e) >> 46) & 0x1)
#define ecap_max_handle_mask(e) (((e) >> 20) & 0xf)
#define ecap_sc_support(e) (((e) >> 7) & 0x1) /* Snooping Control */
+/*
+ * Decoding Perf Capability Register
+ */
+#define pcap_num_cntr(p) ((p) & 0xffff)
+#define pcap_cntr_width(p) (((p) >> 16) & 0x7f)
+#define pcap_num_event_group(p) (((p) >> 24) & 0x1f)
+#define pcap_filters_mask(p) (((p) >> 32) & 0x1f)
+#define pcap_interrupt(p) (((p) >> 50) & 0x1)
+/* The counter stride is calculated as 2 ^ (x+10) bytes */
+#define pcap_cntr_stride(p) (1ULL << ((((p) >> 52) & 0x7) + 10))
+
+/*
+ * Decoding Perf Event Capability Register
+ */
+#define pecap_es(p) ((p) & 0xfffffff)
+
/* Virtual command interface capability */
#define vccap_pasid(v) (((v) & DMA_VCS_PAS)) /* PASID allocation */
iommu core */
};
+struct iommu_pmu {
+ struct intel_iommu *iommu;
+ u32 num_cntr; /* Number of counters */
+ u32 num_eg; /* Number of event group */
+ u32 cntr_width; /* Counter width */
+ u32 cntr_stride; /* Counter Stride */
+ u32 filter; /* Bitmask of filter support */
+ void __iomem *base; /* the PerfMon base address */
+ void __iomem *cfg_reg; /* counter configuration base address */
+ void __iomem *cntr_reg; /* counter 0 address*/
+ void __iomem *overflow; /* overflow status register */
+
+ u64 *evcap; /* Indicates all supported events */
+ u32 **cntr_evcap; /* Supported events of each counter. */
+};
+
struct intel_iommu {
void __iomem *reg; /* Pointer to hardware regs, virtual addr */
u64 reg_phys; /* physical address of hw register set */
struct dmar_drhd_unit *drhd;
void *perf_statistic;
+
+ struct iommu_pmu *pmu;
};
/* PCI domain-device relationship */
--- /dev/null
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Support Intel IOMMU PerfMon
+ * Copyright(c) 2023 Intel Corporation.
+ */
+#define pr_fmt(fmt) "DMAR: " fmt
+#define dev_fmt(fmt) pr_fmt(fmt)
+
+#include <linux/dmar.h>
+#include "iommu.h"
+#include "perfmon.h"
+
+static inline void __iomem *
+get_perf_reg_address(struct intel_iommu *iommu, u32 offset)
+{
+ u32 off = dmar_readl(iommu->reg + offset);
+
+ return iommu->reg + off;
+}
+
+int alloc_iommu_pmu(struct intel_iommu *iommu)
+{
+ struct iommu_pmu *iommu_pmu;
+ int i, j, ret;
+ u64 perfcap;
+ u32 cap;
+
+ if (!ecap_pms(iommu->ecap))
+ return 0;
+
+ /* The IOMMU PMU requires the ECMD support as well */
+ if (!cap_ecmds(iommu->cap))
+ return -ENODEV;
+
+ perfcap = dmar_readq(iommu->reg + DMAR_PERFCAP_REG);
+ /* The performance monitoring is not supported. */
+ if (!perfcap)
+ return -ENODEV;
+
+ /* Sanity check for the number of the counters and event groups */
+ if (!pcap_num_cntr(perfcap) || !pcap_num_event_group(perfcap))
+ return -ENODEV;
+
+ /* The interrupt on overflow is required */
+ if (!pcap_interrupt(perfcap))
+ return -ENODEV;
+
+ iommu_pmu = kzalloc(sizeof(*iommu_pmu), GFP_KERNEL);
+ if (!iommu_pmu)
+ return -ENOMEM;
+
+ iommu_pmu->num_cntr = pcap_num_cntr(perfcap);
+ iommu_pmu->cntr_width = pcap_cntr_width(perfcap);
+ iommu_pmu->filter = pcap_filters_mask(perfcap);
+ iommu_pmu->cntr_stride = pcap_cntr_stride(perfcap);
+ iommu_pmu->num_eg = pcap_num_event_group(perfcap);
+
+ iommu_pmu->evcap = kcalloc(iommu_pmu->num_eg, sizeof(u64), GFP_KERNEL);
+ if (!iommu_pmu->evcap) {
+ ret = -ENOMEM;
+ goto free_pmu;
+ }
+
+ /* Parse event group capabilities */
+ for (i = 0; i < iommu_pmu->num_eg; i++) {
+ u64 pcap;
+
+ pcap = dmar_readq(iommu->reg + DMAR_PERFEVNTCAP_REG +
+ i * IOMMU_PMU_CAP_REGS_STEP);
+ iommu_pmu->evcap[i] = pecap_es(pcap);
+ }
+
+ iommu_pmu->cntr_evcap = kcalloc(iommu_pmu->num_cntr, sizeof(u32 *), GFP_KERNEL);
+ if (!iommu_pmu->cntr_evcap) {
+ ret = -ENOMEM;
+ goto free_pmu_evcap;
+ }
+ for (i = 0; i < iommu_pmu->num_cntr; i++) {
+ iommu_pmu->cntr_evcap[i] = kcalloc(iommu_pmu->num_eg, sizeof(u32), GFP_KERNEL);
+ if (!iommu_pmu->cntr_evcap[i]) {
+ ret = -ENOMEM;
+ goto free_pmu_cntr_evcap;
+ }
+ /*
+ * Set to the global capabilities, will adjust according
+ * to per-counter capabilities later.
+ */
+ for (j = 0; j < iommu_pmu->num_eg; j++)
+ iommu_pmu->cntr_evcap[i][j] = (u32)iommu_pmu->evcap[j];
+ }
+
+ iommu_pmu->cfg_reg = get_perf_reg_address(iommu, DMAR_PERFCFGOFF_REG);
+ iommu_pmu->cntr_reg = get_perf_reg_address(iommu, DMAR_PERFCNTROFF_REG);
+ iommu_pmu->overflow = get_perf_reg_address(iommu, DMAR_PERFOVFOFF_REG);
+
+ /*
+ * Check per-counter capabilities. All counters should have the
+ * same capabilities on Interrupt on Overflow Support and Counter
+ * Width.
+ */
+ for (i = 0; i < iommu_pmu->num_cntr; i++) {
+ cap = dmar_readl(iommu_pmu->cfg_reg +
+ i * IOMMU_PMU_CFG_OFFSET +
+ IOMMU_PMU_CFG_CNTRCAP_OFFSET);
+ if (!iommu_cntrcap_pcc(cap))
+ continue;
+
+ /*
+ * It's possible that some counters have a different
+ * capability because of e.g., HW bug. Check the corner
+ * case here and simply drop those counters.
+ */
+ if ((iommu_cntrcap_cw(cap) != iommu_pmu->cntr_width) ||
+ !iommu_cntrcap_ios(cap)) {
+ iommu_pmu->num_cntr = i;
+ pr_warn("PMU counter capability inconsistent, counter number reduced to %d\n",
+ iommu_pmu->num_cntr);
+ }
+
+ /* Clear the pre-defined events group */
+ for (j = 0; j < iommu_pmu->num_eg; j++)
+ iommu_pmu->cntr_evcap[i][j] = 0;
+
+ /* Override with per-counter event capabilities */
+ for (j = 0; j < iommu_cntrcap_egcnt(cap); j++) {
+ cap = dmar_readl(iommu_pmu->cfg_reg + i * IOMMU_PMU_CFG_OFFSET +
+ IOMMU_PMU_CFG_CNTREVCAP_OFFSET +
+ (j * IOMMU_PMU_OFF_REGS_STEP));
+ iommu_pmu->cntr_evcap[i][iommu_event_group(cap)] = iommu_event_select(cap);
+ /*
+ * Some events may only be supported by a specific counter.
+ * Track them in the evcap as well.
+ */
+ iommu_pmu->evcap[iommu_event_group(cap)] |= iommu_event_select(cap);
+ }
+ }
+
+ iommu_pmu->iommu = iommu;
+ iommu->pmu = iommu_pmu;
+
+ return 0;
+
+free_pmu_cntr_evcap:
+ for (i = 0; i < iommu_pmu->num_cntr; i++)
+ kfree(iommu_pmu->cntr_evcap[i]);
+ kfree(iommu_pmu->cntr_evcap);
+free_pmu_evcap:
+ kfree(iommu_pmu->evcap);
+free_pmu:
+ kfree(iommu_pmu);
+
+ return ret;
+}
+
+void free_iommu_pmu(struct intel_iommu *iommu)
+{
+ struct iommu_pmu *iommu_pmu = iommu->pmu;
+
+ if (!iommu_pmu)
+ return;
+
+ if (iommu_pmu->evcap) {
+ int i;
+
+ for (i = 0; i < iommu_pmu->num_cntr; i++)
+ kfree(iommu_pmu->cntr_evcap[i]);
+ kfree(iommu_pmu->cntr_evcap);
+ }
+ kfree(iommu_pmu->evcap);
+ kfree(iommu_pmu);
+ iommu->pmu = NULL;
+}
--- /dev/null
+/* SPDX-License-Identifier: GPL-2.0 */
+
+/*
+ * PERFCFGOFF_REG, PERFFRZOFF_REG
+ * PERFOVFOFF_REG, PERFCNTROFF_REG
+ */
+#define IOMMU_PMU_NUM_OFF_REGS 4
+#define IOMMU_PMU_OFF_REGS_STEP 4
+
+#define IOMMU_PMU_CFG_OFFSET 0x100
+#define IOMMU_PMU_CFG_CNTRCAP_OFFSET 0x80
+#define IOMMU_PMU_CFG_CNTREVCAP_OFFSET 0x84
+#define IOMMU_PMU_CFG_SIZE 0x8
+#define IOMMU_PMU_CFG_FILTERS_OFFSET 0x4
+
+#define IOMMU_PMU_CAP_REGS_STEP 8
+
+#define iommu_cntrcap_pcc(p) ((p) & 0x1)
+#define iommu_cntrcap_cw(p) (((p) >> 8) & 0xff)
+#define iommu_cntrcap_ios(p) (((p) >> 16) & 0x1)
+#define iommu_cntrcap_egcnt(p) (((p) >> 28) & 0xf)
+
+#define iommu_event_select(p) ((p) & 0xfffffff)
+#define iommu_event_group(p) (((p) >> 28) & 0xf)
+
+#ifdef CONFIG_INTEL_IOMMU_PERF_EVENTS
+int alloc_iommu_pmu(struct intel_iommu *iommu);
+void free_iommu_pmu(struct intel_iommu *iommu);
+#else
+static inline int
+alloc_iommu_pmu(struct intel_iommu *iommu)
+{
+ return 0;
+}
+
+static inline void
+free_iommu_pmu(struct intel_iommu *iommu)
+{
+}
+#endif /* CONFIG_INTEL_IOMMU_PERF_EVENTS */