diff options
author | Robin Murphy <robin.murphy@arm.com> | 2018-02-15 18:51:42 +0000 |
---|---|---|
committer | Arnd Bergmann <arnd@arndb.de> | 2018-03-06 17:26:17 +0100 |
commit | 3de6be7a3dd8934e59d85fc60a170d4ab2f0a0f2 (patch) | |
tree | 57526a619d1fe827b191906ae2e455eb963afef8 /drivers/bus/arm-cci.c | |
parent | 1888d3ddc3d6a2511be86045cfb2e7ea5fc67c44 (diff) |
drivers/bus: Split Arm CCI driver
The arm-cci driver is really two entirely separate drivers; one for MCPM
port control and the other for the performance monitors. Since they are
already relatively self-contained, let's take the plunge and move the
PMU parts out to drivers/perf where they belong these days. For non-MCPM
systems this leaves a small dependency on the remaining "bus" stub for
initial probing and discovery, but we end up with something that still
fits the general pattern of its fellow system PMU drivers to ease future
maintenance.
Moving code to a new file also offers a perfect excuse to modernise the
license/copyright headers and clean up some funky linewraps on the way.
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Reviewed-by: Suzuki Poulose <suzuki.poulose@arm.com>
Acked-by: Punit Agrawal <punit.agrawal@arm.com>
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Diffstat (limited to 'drivers/bus/arm-cci.c')
-rw-r--r-- | drivers/bus/arm-cci.c | 1745 |
1 files changed, 2 insertions, 1743 deletions
diff --git a/drivers/bus/arm-cci.c b/drivers/bus/arm-cci.c index 5426c04fe24b..503c1789dd02 100644 --- a/drivers/bus/arm-cci.c +++ b/drivers/bus/arm-cci.c @@ -16,20 +16,17 @@ #include <linux/arm-cci.h> #include <linux/io.h> -#include <linux/interrupt.h> #include <linux/module.h> #include <linux/of_address.h> -#include <linux/of_irq.h> #include <linux/of_platform.h> -#include <linux/perf_event.h> #include <linux/platform_device.h> #include <linux/slab.h> -#include <linux/spinlock.h> #include <asm/cacheflush.h> #include <asm/smp_plat.h> -static void __iomem *cci_ctrl_base; +/* Referenced read-only by the PMU driver; see drivers/perf/arm-cci.c */ +void __iomem *cci_ctrl_base; static unsigned long cci_ctrl_phys; #ifdef CONFIG_ARM_CCI400_PORT_CTRL @@ -59,1716 +56,7 @@ static const struct of_device_id arm_cci_matches[] = { {}, }; -#ifdef CONFIG_ARM_CCI_PMU - #define DRIVER_NAME "ARM-CCI" -#define DRIVER_NAME_PMU DRIVER_NAME " PMU" - -#define CCI_PMCR 0x0100 -#define CCI_PID2 0x0fe8 - -#define CCI_PMCR_CEN 0x00000001 -#define CCI_PMCR_NCNT_MASK 0x0000f800 -#define CCI_PMCR_NCNT_SHIFT 11 - -#define CCI_PID2_REV_MASK 0xf0 -#define CCI_PID2_REV_SHIFT 4 - -#define CCI_PMU_EVT_SEL 0x000 -#define CCI_PMU_CNTR 0x004 -#define CCI_PMU_CNTR_CTRL 0x008 -#define CCI_PMU_OVRFLW 0x00c - -#define CCI_PMU_OVRFLW_FLAG 1 - -#define CCI_PMU_CNTR_SIZE(model) ((model)->cntr_size) -#define CCI_PMU_CNTR_BASE(model, idx) ((idx) * CCI_PMU_CNTR_SIZE(model)) -#define CCI_PMU_CNTR_MASK ((1ULL << 32) -1) -#define CCI_PMU_CNTR_LAST(cci_pmu) (cci_pmu->num_cntrs - 1) - -#define CCI_PMU_MAX_HW_CNTRS(model) \ - ((model)->num_hw_cntrs + (model)->fixed_hw_cntrs) - -/* Types of interfaces that can generate events */ -enum { - CCI_IF_SLAVE, - CCI_IF_MASTER, -#ifdef CONFIG_ARM_CCI5xx_PMU - CCI_IF_GLOBAL, -#endif - CCI_IF_MAX, -}; - -struct event_range { - u32 min; - u32 max; -}; - -struct cci_pmu_hw_events { - struct perf_event **events; - unsigned long *used_mask; - raw_spinlock_t pmu_lock; -}; - -struct cci_pmu; -/* - * struct cci_pmu_model: - * @fixed_hw_cntrs - Number of fixed event counters - * @num_hw_cntrs - Maximum number of programmable event counters - * @cntr_size - Size of an event counter mapping - */ -struct cci_pmu_model { - char *name; - u32 fixed_hw_cntrs; - u32 num_hw_cntrs; - u32 cntr_size; - struct attribute **format_attrs; - struct attribute **event_attrs; - struct event_range event_ranges[CCI_IF_MAX]; - int (*validate_hw_event)(struct cci_pmu *, unsigned long); - int (*get_event_idx)(struct cci_pmu *, struct cci_pmu_hw_events *, unsigned long); - void (*write_counters)(struct cci_pmu *, unsigned long *); -}; - -static struct cci_pmu_model cci_pmu_models[]; - -struct cci_pmu { - void __iomem *base; - struct pmu pmu; - int nr_irqs; - int *irqs; - unsigned long active_irqs; - const struct cci_pmu_model *model; - struct cci_pmu_hw_events hw_events; - struct platform_device *plat_device; - int num_cntrs; - atomic_t active_events; - struct mutex reserve_mutex; - struct hlist_node node; - cpumask_t cpus; -}; - -#define to_cci_pmu(c) (container_of(c, struct cci_pmu, pmu)) - -enum cci_models { -#ifdef CONFIG_ARM_CCI400_PMU - CCI400_R0, - CCI400_R1, -#endif -#ifdef CONFIG_ARM_CCI5xx_PMU - CCI500_R0, - CCI550_R0, -#endif - CCI_MODEL_MAX -}; - -static void pmu_write_counters(struct cci_pmu *cci_pmu, - unsigned long *mask); -static ssize_t cci_pmu_format_show(struct device *dev, - struct device_attribute *attr, char *buf); -static ssize_t cci_pmu_event_show(struct device *dev, - struct device_attribute *attr, char *buf); - -#define CCI_EXT_ATTR_ENTRY(_name, _func, _config) \ - &((struct dev_ext_attribute[]) { \ - { __ATTR(_name, S_IRUGO, _func, NULL), (void *)_config } \ - })[0].attr.attr - -#define CCI_FORMAT_EXT_ATTR_ENTRY(_name, _config) \ - CCI_EXT_ATTR_ENTRY(_name, cci_pmu_format_show, (char *)_config) -#define CCI_EVENT_EXT_ATTR_ENTRY(_name, _config) \ - CCI_EXT_ATTR_ENTRY(_name, cci_pmu_event_show, (unsigned long)_config) - -/* CCI400 PMU Specific definitions */ - -#ifdef CONFIG_ARM_CCI400_PMU - -/* Port ids */ -#define CCI400_PORT_S0 0 -#define CCI400_PORT_S1 1 -#define CCI400_PORT_S2 2 -#define CCI400_PORT_S3 3 -#define CCI400_PORT_S4 4 -#define CCI400_PORT_M0 5 -#define CCI400_PORT_M1 6 -#define CCI400_PORT_M2 7 - -#define CCI400_R1_PX 5 - -/* - * Instead of an event id to monitor CCI cycles, a dedicated counter is - * provided. Use 0xff to represent CCI cycles and hope that no future revisions - * make use of this event in hardware. - */ -enum cci400_perf_events { - CCI400_PMU_CYCLES = 0xff -}; - -#define CCI400_PMU_CYCLE_CNTR_IDX 0 -#define CCI400_PMU_CNTR0_IDX 1 - -/* - * CCI PMU event id is an 8-bit value made of two parts - bits 7:5 for one of 8 - * ports and bits 4:0 are event codes. There are different event codes - * associated with each port type. - * - * Additionally, the range of events associated with the port types changed - * between Rev0 and Rev1. - * - * The constants below define the range of valid codes for each port type for - * the different revisions and are used to validate the event to be monitored. - */ - -#define CCI400_PMU_EVENT_MASK 0xffUL -#define CCI400_PMU_EVENT_SOURCE_SHIFT 5 -#define CCI400_PMU_EVENT_SOURCE_MASK 0x7 -#define CCI400_PMU_EVENT_CODE_SHIFT 0 -#define CCI400_PMU_EVENT_CODE_MASK 0x1f -#define CCI400_PMU_EVENT_SOURCE(event) \ - ((event >> CCI400_PMU_EVENT_SOURCE_SHIFT) & \ - CCI400_PMU_EVENT_SOURCE_MASK) -#define CCI400_PMU_EVENT_CODE(event) \ - ((event >> CCI400_PMU_EVENT_CODE_SHIFT) & CCI400_PMU_EVENT_CODE_MASK) - -#define CCI400_R0_SLAVE_PORT_MIN_EV 0x00 -#define CCI400_R0_SLAVE_PORT_MAX_EV 0x13 -#define CCI400_R0_MASTER_PORT_MIN_EV 0x14 -#define CCI400_R0_MASTER_PORT_MAX_EV 0x1a - -#define CCI400_R1_SLAVE_PORT_MIN_EV 0x00 -#define CCI400_R1_SLAVE_PORT_MAX_EV 0x14 -#define CCI400_R1_MASTER_PORT_MIN_EV 0x00 -#define CCI400_R1_MASTER_PORT_MAX_EV 0x11 - -#define CCI400_CYCLE_EVENT_EXT_ATTR_ENTRY(_name, _config) \ - CCI_EXT_ATTR_ENTRY(_name, cci400_pmu_cycle_event_show, \ - (unsigned long)_config) - -static ssize_t cci400_pmu_cycle_event_show(struct device *dev, - struct device_attribute *attr, char *buf); - -static struct attribute *cci400_pmu_format_attrs[] = { - CCI_FORMAT_EXT_ATTR_ENTRY(event, "config:0-4"), - CCI_FORMAT_EXT_ATTR_ENTRY(source, "config:5-7"), - NULL -}; - -static struct attribute *cci400_r0_pmu_event_attrs[] = { - /* Slave events */ - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_any, 0x0), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_device, 0x01), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_normal_or_nonshareable, 0x2), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_inner_or_outershareable, 0x3), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_cache_maintenance, 0x4), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_mem_barrier, 0x5), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_sync_barrier, 0x6), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_dvm_msg, 0x7), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_dvm_msg_sync, 0x8), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_stall_tt_full, 0x9), - CCI_EVENT_EXT_ATTR_ENTRY(si_r_data_last_hs_snoop, 0xA), - CCI_EVENT_EXT_ATTR_ENTRY(si_r_data_stall_rvalids_h_rready_l, 0xB), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_any, 0xC), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_device, 0xD), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_normal_or_nonshareable, 0xE), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_inner_or_outershare_wback_wclean, 0xF), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_write_unique, 0x10), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_write_line_unique, 0x11), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_evict, 0x12), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_stall_tt_full, 0x13), - /* Master events */ - CCI_EVENT_EXT_ATTR_ENTRY(mi_retry_speculative_fetch, 0x14), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall_addr_hazard, 0x15), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall_id_hazard, 0x16), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall_tt_full, 0x17), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall_barrier_hazard, 0x18), - CCI_EVENT_EXT_ATTR_ENTRY(mi_wrq_stall_barrier_hazard, 0x19), - CCI_EVENT_EXT_ATTR_ENTRY(mi_wrq_stall_tt_full, 0x1A), - /* Special event for cycles counter */ - CCI400_CYCLE_EVENT_EXT_ATTR_ENTRY(cycles, 0xff), - NULL -}; - -static struct attribute *cci400_r1_pmu_event_attrs[] = { - /* Slave events */ - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_any, 0x0), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_device, 0x01), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_normal_or_nonshareable, 0x2), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_inner_or_outershareable, 0x3), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_cache_maintenance, 0x4), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_mem_barrier, 0x5), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_sync_barrier, 0x6), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_dvm_msg, 0x7), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_dvm_msg_sync, 0x8), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_stall_tt_full, 0x9), - CCI_EVENT_EXT_ATTR_ENTRY(si_r_data_last_hs_snoop, 0xA), - CCI_EVENT_EXT_ATTR_ENTRY(si_r_data_stall_rvalids_h_rready_l, 0xB), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_any, 0xC), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_device, 0xD), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_normal_or_nonshareable, 0xE), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_inner_or_outershare_wback_wclean, 0xF), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_write_unique, 0x10), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_write_line_unique, 0x11), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_evict, 0x12), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_stall_tt_full, 0x13), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_stall_slave_id_hazard, 0x14), - /* Master events */ - CCI_EVENT_EXT_ATTR_ENTRY(mi_retry_speculative_fetch, 0x0), - CCI_EVENT_EXT_ATTR_ENTRY(mi_stall_cycle_addr_hazard, 0x1), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall_master_id_hazard, 0x2), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall_hi_prio_rtq_full, 0x3), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall_barrier_hazard, 0x4), - CCI_EVENT_EXT_ATTR_ENTRY(mi_wrq_stall_barrier_hazard, 0x5), - CCI_EVENT_EXT_ATTR_ENTRY(mi_wrq_stall_wtq_full, 0x6), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall_low_prio_rtq_full, 0x7), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall_mid_prio_rtq_full, 0x8), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall_qvn_vn0, 0x9), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall_qvn_vn1, 0xA), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall_qvn_vn2, 0xB), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall_qvn_vn3, 0xC), - CCI_EVENT_EXT_ATTR_ENTRY(mi_wrq_stall_qvn_vn0, 0xD), - CCI_EVENT_EXT_ATTR_ENTRY(mi_wrq_stall_qvn_vn1, 0xE), - CCI_EVENT_EXT_ATTR_ENTRY(mi_wrq_stall_qvn_vn2, 0xF), - CCI_EVENT_EXT_ATTR_ENTRY(mi_wrq_stall_qvn_vn3, 0x10), - CCI_EVENT_EXT_ATTR_ENTRY(mi_wrq_unique_or_line_unique_addr_hazard, 0x11), - /* Special event for cycles counter */ - CCI400_CYCLE_EVENT_EXT_ATTR_ENTRY(cycles, 0xff), - NULL -}; - -static ssize_t cci400_pmu_cycle_event_show(struct device *dev, - struct device_attribute *attr, char *buf) -{ - struct dev_ext_attribute *eattr = container_of(attr, - struct dev_ext_attribute, attr); - return snprintf(buf, PAGE_SIZE, "config=0x%lx\n", (unsigned long)eattr->var); -} - -static int cci400_get_event_idx(struct cci_pmu *cci_pmu, - struct cci_pmu_hw_events *hw, - unsigned long cci_event) -{ - int idx; - - /* cycles event idx is fixed */ - if (cci_event == CCI400_PMU_CYCLES) { - if (test_and_set_bit(CCI400_PMU_CYCLE_CNTR_IDX, hw->used_mask)) - return -EAGAIN; - - return CCI400_PMU_CYCLE_CNTR_IDX; - } - - for (idx = CCI400_PMU_CNTR0_IDX; idx <= CCI_PMU_CNTR_LAST(cci_pmu); ++idx) - if (!test_and_set_bit(idx, hw->used_mask)) - return idx; - - /* No counters available */ - return -EAGAIN; -} - -static int cci400_validate_hw_event(struct cci_pmu *cci_pmu, unsigned long hw_event) -{ - u8 ev_source = CCI400_PMU_EVENT_SOURCE(hw_event); - u8 ev_code = CCI400_PMU_EVENT_CODE(hw_event); - int if_type; - - if (hw_event & ~CCI400_PMU_EVENT_MASK) - return -ENOENT; - - if (hw_event == CCI400_PMU_CYCLES) - return hw_event; - - switch (ev_source) { - case CCI400_PORT_S0: - case CCI400_PORT_S1: - case CCI400_PORT_S2: - case CCI400_PORT_S3: - case CCI400_PORT_S4: - /* Slave Interface */ - if_type = CCI_IF_SLAVE; - break; - case CCI400_PORT_M0: - case CCI400_PORT_M1: - case CCI400_PORT_M2: - /* Master Interface */ - if_type = CCI_IF_MASTER; - break; - default: - return -ENOENT; - } - - if (ev_code >= cci_pmu->model->event_ranges[if_type].min && - ev_code <= cci_pmu->model->event_ranges[if_type].max) - return hw_event; - - return -ENOENT; -} - -static int probe_cci400_revision(void) -{ - int rev; - rev = readl_relaxed(cci_ctrl_base + CCI_PID2) & CCI_PID2_REV_MASK; - rev >>= CCI_PID2_REV_SHIFT; - - if (rev < CCI400_R1_PX) - return CCI400_R0; - else - return CCI400_R1; -} - -static const struct cci_pmu_model *probe_cci_model(struct platform_device *pdev) -{ - if (platform_has_secure_cci_access()) - return &cci_pmu_models[probe_cci400_revision()]; - return NULL; -} -#else /* !CONFIG_ARM_CCI400_PMU */ -static inline struct cci_pmu_model *probe_cci_model(struct platform_device *pdev) -{ - return NULL; -} -#endif /* CONFIG_ARM_CCI400_PMU */ - -#ifdef CONFIG_ARM_CCI5xx_PMU - -/* - * CCI5xx PMU event id is an 9-bit value made of two parts. - * bits [8:5] - Source for the event - * bits [4:0] - Event code (specific to type of interface) - * - * - */ - -/* Port ids */ -#define CCI5xx_PORT_S0 0x0 -#define CCI5xx_PORT_S1 0x1 -#define CCI5xx_PORT_S2 0x2 -#define CCI5xx_PORT_S3 0x3 -#define CCI5xx_PORT_S4 0x4 -#define CCI5xx_PORT_S5 0x5 -#define CCI5xx_PORT_S6 0x6 - -#define CCI5xx_PORT_M0 0x8 -#define CCI5xx_PORT_M1 0x9 -#define CCI5xx_PORT_M2 0xa -#define CCI5xx_PORT_M3 0xb -#define CCI5xx_PORT_M4 0xc -#define CCI5xx_PORT_M5 0xd -#define CCI5xx_PORT_M6 0xe - -#define CCI5xx_PORT_GLOBAL 0xf - -#define CCI5xx_PMU_EVENT_MASK 0x1ffUL -#define CCI5xx_PMU_EVENT_SOURCE_SHIFT 0x5 -#define CCI5xx_PMU_EVENT_SOURCE_MASK 0xf -#define CCI5xx_PMU_EVENT_CODE_SHIFT 0x0 -#define CCI5xx_PMU_EVENT_CODE_MASK 0x1f - -#define CCI5xx_PMU_EVENT_SOURCE(event) \ - ((event >> CCI5xx_PMU_EVENT_SOURCE_SHIFT) & CCI5xx_PMU_EVENT_SOURCE_MASK) -#define CCI5xx_PMU_EVENT_CODE(event) \ - ((event >> CCI5xx_PMU_EVENT_CODE_SHIFT) & CCI5xx_PMU_EVENT_CODE_MASK) - -#define CCI5xx_SLAVE_PORT_MIN_EV 0x00 -#define CCI5xx_SLAVE_PORT_MAX_EV 0x1f -#define CCI5xx_MASTER_PORT_MIN_EV 0x00 -#define CCI5xx_MASTER_PORT_MAX_EV 0x06 -#define CCI5xx_GLOBAL_PORT_MIN_EV 0x00 -#define CCI5xx_GLOBAL_PORT_MAX_EV 0x0f - - -#define CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(_name, _config) \ - CCI_EXT_ATTR_ENTRY(_name, cci5xx_pmu_global_event_show, \ - (unsigned long) _config) - -static ssize_t cci5xx_pmu_global_event_show(struct device *dev, - struct device_attribute *attr, char *buf); - -static struct attribute *cci5xx_pmu_format_attrs[] = { - CCI_FORMAT_EXT_ATTR_ENTRY(event, "config:0-4"), - CCI_FORMAT_EXT_ATTR_ENTRY(source, "config:5-8"), - NULL, -}; - -static struct attribute *cci5xx_pmu_event_attrs[] = { - /* Slave events */ - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_arvalid, 0x0), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_dev, 0x1), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_nonshareable, 0x2), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_shareable_non_alloc, 0x3), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_shareable_alloc, 0x4), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_invalidate, 0x5), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_cache_maint, 0x6), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_dvm_msg, 0x7), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_rval, 0x8), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_hs_rlast_snoop, 0x9), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_hs_awalid, 0xA), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_dev, 0xB), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_non_shareable, 0xC), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_share_wb, 0xD), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_share_wlu, 0xE), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_share_wunique, 0xF), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_evict, 0x10), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_wrevict, 0x11), - CCI_EVENT_EXT_ATTR_ENTRY(si_w_data_beat, 0x12), - CCI_EVENT_EXT_ATTR_ENTRY(si_srq_acvalid, 0x13), - CCI_EVENT_EXT_ATTR_ENTRY(si_srq_read, 0x14), - CCI_EVENT_EXT_ATTR_ENTRY(si_srq_clean, 0x15), - CCI_EVENT_EXT_ATTR_ENTRY(si_srq_data_transfer_low, 0x16), - CCI_EVENT_EXT_ATTR_ENTRY(si_rrq_stall_arvalid, 0x17), - CCI_EVENT_EXT_ATTR_ENTRY(si_r_data_stall, 0x18), - CCI_EVENT_EXT_ATTR_ENTRY(si_wrq_stall, 0x19), - CCI_EVENT_EXT_ATTR_ENTRY(si_w_data_stall, 0x1A), - CCI_EVENT_EXT_ATTR_ENTRY(si_w_resp_stall, 0x1B), - CCI_EVENT_EXT_ATTR_ENTRY(si_srq_stall, 0x1C), - CCI_EVENT_EXT_ATTR_ENTRY(si_s_data_stall, 0x1D), - CCI_EVENT_EXT_ATTR_ENTRY(si_rq_stall_ot_limit, 0x1E), - CCI_EVENT_EXT_ATTR_ENTRY(si_r_stall_arbit, 0x1F), - - /* Master events */ - CCI_EVENT_EXT_ATTR_ENTRY(mi_r_data_beat_any, 0x0), - CCI_EVENT_EXT_ATTR_ENTRY(mi_w_data_beat_any, 0x1), - CCI_EVENT_EXT_ATTR_ENTRY(mi_rrq_stall, 0x2), - CCI_EVENT_EXT_ATTR_ENTRY(mi_r_data_stall, 0x3), - CCI_EVENT_EXT_ATTR_ENTRY(mi_wrq_stall, 0x4), - CCI_EVENT_EXT_ATTR_ENTRY(mi_w_data_stall, 0x5), - CCI_EVENT_EXT_ATTR_ENTRY(mi_w_resp_stall, 0x6), - - /* Global events */ - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_access_filter_bank_0_1, 0x0), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_access_filter_bank_2_3, 0x1), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_access_filter_bank_4_5, 0x2), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_access_filter_bank_6_7, 0x3), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_access_miss_filter_bank_0_1, 0x4), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_access_miss_filter_bank_2_3, 0x5), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_access_miss_filter_bank_4_5, 0x6), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_access_miss_filter_bank_6_7, 0x7), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_back_invalidation, 0x8), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_stall_alloc_busy, 0x9), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_stall_tt_full, 0xA), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_wrq, 0xB), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_cd_hs, 0xC), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_rq_stall_addr_hazard, 0xD), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_rq_stall_tt_full, 0xE), - CCI5xx_GLOBAL_EVENT_EXT_ATTR_ENTRY(cci_snoop_rq_tzmp1_prot, 0xF), - NULL -}; - -static ssize_t cci5xx_pmu_global_event_show(struct device *dev, - struct device_attribute *attr, char *buf) -{ - struct dev_ext_attribute *eattr = container_of(attr, - struct dev_ext_attribute, attr); - /* Global events have single fixed source code */ - return snprintf(buf, PAGE_SIZE, "event=0x%lx,source=0x%x\n", - (unsigned long)eattr->var, CCI5xx_PORT_GLOBAL); -} - -/* - * CCI500 provides 8 independent event counters that can count - * any of the events available. - * CCI500 PMU event source ids - * 0x0-0x6 - Slave interfaces - * 0x8-0xD - Master interfaces - * 0xf - Global Events - * 0x7,0xe - Reserved - */ -static int cci500_validate_hw_event(struct cci_pmu *cci_pmu, - unsigned long hw_event) -{ - u32 ev_source = CCI5xx_PMU_EVENT_SOURCE(hw_event); - u32 ev_code = CCI5xx_PMU_EVENT_CODE(hw_event); - int if_type; - - if (hw_event & ~CCI5xx_PMU_EVENT_MASK) - return -ENOENT; - - switch (ev_source) { - case CCI5xx_PORT_S0: - case CCI5xx_PORT_S1: - case CCI5xx_PORT_S2: - case CCI5xx_PORT_S3: - case CCI5xx_PORT_S4: - case CCI5xx_PORT_S5: - case CCI5xx_PORT_S6: - if_type = CCI_IF_SLAVE; - break; - case CCI5xx_PORT_M0: - case CCI5xx_PORT_M1: - case CCI5xx_PORT_M2: - case CCI5xx_PORT_M3: - case CCI5xx_PORT_M4: - case CCI5xx_PORT_M5: - if_type = CCI_IF_MASTER; - break; - case CCI5xx_PORT_GLOBAL: - if_type = CCI_IF_GLOBAL; - break; - default: - return -ENOENT; - } - - if (ev_code >= cci_pmu->model->event_ranges[if_type].min && - ev_code <= cci_pmu->model->event_ranges[if_type].max) - return hw_event; - - return -ENOENT; -} - -/* - * CCI550 provides 8 independent event counters that can count - * any of the events available. - * CCI550 PMU event source ids - * 0x0-0x6 - Slave interfaces - * 0x8-0xe - Master interfaces - * 0xf - Global Events - * 0x7 - Reserved - */ -static int cci550_validate_hw_event(struct cci_pmu *cci_pmu, - unsigned long hw_event) -{ - u32 ev_source = CCI5xx_PMU_EVENT_SOURCE(hw_event); - u32 ev_code = CCI5xx_PMU_EVENT_CODE(hw_event); - int if_type; - - if (hw_event & ~CCI5xx_PMU_EVENT_MASK) - return -ENOENT; - - switch (ev_source) { - case CCI5xx_PORT_S0: - case CCI5xx_PORT_S1: - case CCI5xx_PORT_S2: - case CCI5xx_PORT_S3: - case CCI5xx_PORT_S4: - case CCI5xx_PORT_S5: - case CCI5xx_PORT_S6: - if_type = CCI_IF_SLAVE; - break; - case CCI5xx_PORT_M0: - case CCI5xx_PORT_M1: - case CCI5xx_PORT_M2: - case CCI5xx_PORT_M3: - case CCI5xx_PORT_M4: - case CCI5xx_PORT_M5: - case CCI5xx_PORT_M6: - if_type = CCI_IF_MASTER; - break; - case CCI5xx_PORT_GLOBAL: - if_type = CCI_IF_GLOBAL; - break; - default: - return -ENOENT; - } - - if (ev_code >= cci_pmu->model->event_ranges[if_type].min && - ev_code <= cci_pmu->model->event_ranges[if_type].max) - return hw_event; - - return -ENOENT; -} - -#endif /* CONFIG_ARM_CCI5xx_PMU */ - -/* - * Program the CCI PMU counters which have PERF_HES_ARCH set - * with the event period and mark them ready before we enable - * PMU. - */ -static void cci_pmu_sync_counters(struct cci_pmu *cci_pmu) -{ - int i; - struct cci_pmu_hw_events *cci_hw = &cci_pmu->hw_events; - - DECLARE_BITMAP(mask, cci_pmu->num_cntrs); - - bitmap_zero(mask, cci_pmu->num_cntrs); - for_each_set_bit(i, cci_pmu->hw_events.used_mask, cci_pmu->num_cntrs) { - struct perf_event *event = cci_hw->events[i]; - - if (WARN_ON(!event)) - continue; - - /* Leave the events which are not counting */ - if (event->hw.state & PERF_HES_STOPPED) - continue; - if (event->hw.state & PERF_HES_ARCH) { - set_bit(i, mask); - event->hw.state &= ~PERF_HES_ARCH; - } - } - - pmu_write_counters(cci_pmu, mask); -} - -/* Should be called with cci_pmu->hw_events->pmu_lock held */ -static void __cci_pmu_enable_nosync(struct cci_pmu *cci_pmu) -{ - u32 val; - - /* Enable all the PMU counters. */ - val = readl_relaxed(cci_ctrl_base + CCI_PMCR) | CCI_PMCR_CEN; - writel(val, cci_ctrl_base + CCI_PMCR); -} - -/* Should be called with cci_pmu->hw_events->pmu_lock held */ -static void __cci_pmu_enable_sync(struct cci_pmu *cci_pmu) -{ - cci_pmu_sync_counters(cci_pmu); - __cci_pmu_enable_nosync(cci_pmu); -} - -/* Should be called with cci_pmu->hw_events->pmu_lock held */ -static void __cci_pmu_disable(void) -{ - u32 val; - - /* Disable all the PMU counters. */ - val = readl_relaxed(cci_ctrl_base + CCI_PMCR) & ~CCI_PMCR_CEN; - writel(val, cci_ctrl_base + CCI_PMCR); -} - -static ssize_t cci_pmu_format_show(struct device *dev, - struct device_attribute *attr, char *buf) -{ - struct dev_ext_attribute *eattr = container_of(attr, - struct dev_ext_attribute, attr); - return snprintf(buf, PAGE_SIZE, "%s\n", (char *)eattr->var); -} - -static ssize_t cci_pmu_event_show(struct device *dev, - struct device_attribute *attr, char *buf) -{ - struct dev_ext_attribute *eattr = container_of(attr, - struct dev_ext_attribute, attr); - /* source parameter is mandatory for normal PMU events */ - return snprintf(buf, PAGE_SIZE, "source=?,event=0x%lx\n", - (unsigned long)eattr->var); -} - -static int pmu_is_valid_counter(struct cci_pmu *cci_pmu, int idx) -{ - return 0 <= idx && idx <= CCI_PMU_CNTR_LAST(cci_pmu); -} - -static u32 pmu_read_register(struct cci_pmu *cci_pmu, int idx, unsigned int offset) -{ - return readl_relaxed(cci_pmu->base + - CCI_PMU_CNTR_BASE(cci_pmu->model, idx) + offset); -} - -static void pmu_write_register(struct cci_pmu *cci_pmu, u32 value, - int idx, unsigned int offset) -{ - writel_relaxed(value, cci_pmu->base + - CCI_PMU_CNTR_BASE(cci_pmu->model, idx) + offset); -} - -static void pmu_disable_counter(struct cci_pmu *cci_pmu, int idx) -{ - pmu_write_register(cci_pmu, 0, idx, CCI_PMU_CNTR_CTRL); -} - -static void pmu_enable_counter(struct cci_pmu *cci_pmu, int idx) -{ - pmu_write_register(cci_pmu, 1, idx, CCI_PMU_CNTR_CTRL); -} - -static bool __maybe_unused -pmu_counter_is_enabled(struct cci_pmu *cci_pmu, int idx) -{ - return (pmu_read_register(cci_pmu, idx, CCI_PMU_CNTR_CTRL) & 0x1) != 0; -} - -static void pmu_set_event(struct cci_pmu *cci_pmu, int idx, unsigned long event) -{ - pmu_write_register(cci_pmu, event, idx, CCI_PMU_EVT_SEL); -} - -/* - * For all counters on the CCI-PMU, disable any 'enabled' counters, - * saving the changed counters in the mask, so that we can restore - * it later using pmu_restore_counters. The mask is private to the - * caller. We cannot rely on the used_mask maintained by the CCI_PMU - * as it only tells us if the counter is assigned to perf_event or not. - * The state of the perf_event cannot be locked by the PMU layer, hence - * we check the individual counter status (which can be locked by - * cci_pm->hw_events->pmu_lock). - * - * @mask should be initialised to empty by the caller. - */ -static void __maybe_unused -pmu_save_counters(struct cci_pmu *cci_pmu, unsigned long *mask) -{ - int i; - - for (i = 0; i < cci_pmu->num_cntrs; i++) { - if (pmu_counter_is_enabled(cci_pmu, i)) { - set_bit(i, mask); - pmu_disable_counter(cci_pmu, i); - } - } -} - -/* - * Restore the status of the counters. Reversal of the pmu_save_counters(). - * For each counter set in the mask, enable the counter back. - */ -static void __maybe_unused -pmu_restore_counters(struct cci_pmu *cci_pmu, unsigned long *mask) -{ - int i; - - for_each_set_bit(i, mask, cci_pmu->num_cntrs) - pmu_enable_counter(cci_pmu, i); -} - -/* - * Returns the number of programmable counters actually implemented - * by the cci - */ -static u32 pmu_get_max_counters(void) -{ - return (readl_relaxed(cci_ctrl_base + CCI_PMCR) & - CCI_PMCR_NCNT_MASK) >> CCI_PMCR_NCNT_SHIFT; -} - -static int pmu_get_event_idx(struct cci_pmu_hw_events *hw, struct perf_event *event) -{ - struct cci_pmu *cci_pmu = to_cci_pmu(event->pmu); - unsigned long cci_event = event->hw.config_base; - int idx; - - if (cci_pmu->model->get_event_idx) - return cci_pmu->model->get_event_idx(cci_pmu, hw, cci_event); - - /* Generic code to find an unused idx from the mask */ - for(idx = 0; idx <= CCI_PMU_CNTR_LAST(cci_pmu); idx++) - if (!test_and_set_bit(idx, hw->used_mask)) - return idx; - - /* No counters available */ - return -EAGAIN; -} - -static int pmu_map_event(struct perf_event *event) -{ - struct cci_pmu *cci_pmu = to_cci_pmu(event->pmu); - - if (event->attr.type < PERF_TYPE_MAX || - !cci_pmu->model->validate_hw_event) - return -ENOENT; - - return cci_pmu->model->validate_hw_event(cci_pmu, event->attr.config); -} - -static int pmu_request_irq(struct cci_pmu *cci_pmu, irq_handler_t handler) -{ - int i; - struct platform_device *pmu_device = cci_pmu->plat_device; - - if (unlikely(!pmu_device)) - return -ENODEV; - - if (cci_pmu->nr_irqs < 1) { - dev_err(&pmu_device->dev, "no irqs for CCI PMUs defined\n"); - return -ENODEV; - } - - /* - * Register all available CCI PMU interrupts. In the interrupt handler - * we iterate over the counters checking for interrupt source (the - * overflowing counter) and clear it. - * - * This should allow handling of non-unique interrupt for the counters. - */ - for (i = 0; i < cci_pmu->nr_irqs; i++) { - int err = request_irq(cci_pmu->irqs[i], handler, IRQF_SHARED, - "arm-cci-pmu", cci_pmu); - if (err) { - dev_err(&pmu_device->dev, "unable to request IRQ%d for ARM CCI PMU counters\n", - cci_pmu->irqs[i]); - return err; - } - - set_bit(i, &cci_pmu->active_irqs); - } - - return 0; -} - -static void pmu_free_irq(struct cci_pmu *cci_pmu) -{ - int i; - - for (i = 0; i < cci_pmu->nr_irqs; i++) { - if (!test_and_clear_bit(i, &cci_pmu->active_irqs)) - continue; - - free_irq(cci_pmu->irqs[i], cci_pmu); - } -} - -static u32 pmu_read_counter(struct perf_event *event) -{ - struct cci_pmu *cci_pmu = to_cci_pmu(event->pmu); - struct hw_perf_event *hw_counter = &event->hw; - int idx = hw_counter->idx; - u32 value; - - if (unlikely(!pmu_is_valid_counter(cci_pmu, idx))) { - dev_err(&cci_pmu->plat_device->dev, "Invalid CCI PMU counter %d\n", idx); - return 0; - } - value = pmu_read_register(cci_pmu, idx, CCI_PMU_CNTR); - - return value; -} - -static void pmu_write_counter(struct cci_pmu *cci_pmu, u32 value, int idx) -{ - pmu_write_register(cci_pmu, value, idx, CCI_PMU_CNTR); -} - -static void __pmu_write_counters(struct cci_pmu *cci_pmu, unsigned long *mask) -{ - int i; - struct cci_pmu_hw_events *cci_hw = &cci_pmu->hw_events; - - for_each_set_bit(i, mask, cci_pmu->num_cntrs) { - struct perf_event *event = cci_hw->events[i]; - - if (WARN_ON(!event)) - continue; - pmu_write_counter(cci_pmu, local64_read(&event->hw.prev_count), i); - } -} - -static void pmu_write_counters(struct cci_pmu *cci_pmu, unsigned long *mask) -{ - if (cci_pmu->model->write_counters) - cci_pmu->model->write_counters(cci_pmu, mask); - else - __pmu_write_counters(cci_pmu, mask); -} - -#ifdef CONFIG_ARM_CCI5xx_PMU - -/* - * CCI-500/CCI-550 has advanced power saving policies, which could gate the - * clocks to the PMU counters, which makes the writes to them ineffective. - * The only way to write to those counters is when the global counters - * are enabled and the particular counter is enabled. - * - * So we do the following : - * - * 1) Disable all the PMU counters, saving their current state - * 2) Enable the global PMU profiling, now that all counters are - * disabled. - * - * For each counter to be programmed, repeat steps 3-7: - * - * 3) Write an invalid event code to the event control register for the - counter, so that the counters are not modified. - * 4) Enable the counter control for the counter. - * 5) Set the counter value - * 6) Disable the counter - * 7) Restore the event in the target counter - * - * 8) Disable the global PMU. - * 9) Restore the status of the rest of the counters. - * - * We choose an event which for CCI-5xx is guaranteed not to count. - * We use the highest possible event code (0x1f) for the master interface 0. - */ -#define CCI5xx_INVALID_EVENT ((CCI5xx_PORT_M0 << CCI5xx_PMU_EVENT_SOURCE_SHIFT) | \ - (CCI5xx_PMU_EVENT_CODE_MASK << CCI5xx_PMU_EVENT_CODE_SHIFT)) -static void cci5xx_pmu_write_counters(struct cci_pmu *cci_pmu, unsigned long *mask) -{ - int i; - DECLARE_BITMAP(saved_mask, cci_pmu->num_cntrs); - - bitmap_zero(saved_mask, cci_pmu->num_cntrs); - pmu_save_counters(cci_pmu, saved_mask); - - /* - * Now that all the counters are disabled, we can safely turn the PMU on, - * without syncing the status of the counters - */ - __cci_pmu_enable_nosync(cci_pmu); - - for_each_set_bit(i, mask, cci_pmu->num_cntrs) { - struct perf_event *event = cci_pmu->hw_events.events[i]; - - if (WARN_ON(!event)) - continue; - - pmu_set_event(cci_pmu, i, CCI5xx_INVALID_EVENT); - pmu_enable_counter(cci_pmu, i); - pmu_write_counter(cci_pmu, local64_read(&event->hw.prev_count), i); - pmu_disable_counter(cci_pmu, i); - pmu_set_event(cci_pmu, i, event->hw.config_base); - } - - __cci_pmu_disable(); - - pmu_restore_counters(cci_pmu, saved_mask); -} - -#endif /* CONFIG_ARM_CCI5xx_PMU */ - -static u64 pmu_event_update(struct perf_event *event) -{ - struct hw_perf_event *hwc = &event->hw; - u64 delta, prev_raw_count, new_raw_count; - - do { - prev_raw_count = local64_read(&hwc->prev_count); - new_raw_count = pmu_read_counter(event); - } while (local64_cmpxchg(&hwc->prev_count, prev_raw_count, - new_raw_count) != prev_raw_count); - - delta = (new_raw_count - prev_raw_count) & CCI_PMU_CNTR_MASK; - - local64_add(delta, &event->count); - - return new_raw_count; -} - -static void pmu_read(struct perf_event *event) -{ - pmu_event_update(event); -} - -static void pmu_event_set_period(struct perf_event *event) -{ - struct hw_perf_event *hwc = &event->hw; - /* - * The CCI PMU counters have a period of 2^32. To account for the - * possiblity of extreme interrupt latency we program for a period of - * half that. Hopefully we can handle the interrupt before another 2^31 - * events occur and the counter overtakes its previous value. - */ - u64 val = 1ULL << 31; - local64_set(&hwc->prev_count, val); - - /* - * CCI PMU uses PERF_HES_ARCH to keep track of the counters, whose - * values needs to be sync-ed with the s/w state before the PMU is - * enabled. - * Mark this counter for sync. - */ - hwc->state |= PERF_HES_ARCH; -} - -static irqreturn_t pmu_handle_irq(int irq_num, void *dev) -{ - unsigned long flags; - struct cci_pmu *cci_pmu = dev; - struct cci_pmu_hw_events *events = &cci_pmu->hw_events; - int idx, handled = IRQ_NONE; - - raw_spin_lock_irqsave(&events->pmu_lock, flags); - - /* Disable the PMU while we walk through the counters */ - __cci_pmu_disable(); - /* - * Iterate over counters and update the corresponding perf events. - * This should work regardless of whether we have per-counter overflow - * interrupt or a combined overflow interrupt. - */ - for (idx = 0; idx <= CCI_PMU_CNTR_LAST(cci_pmu); idx++) { - struct perf_event *event = events->events[idx]; - - if (!event) - continue; - - /* Did this counter overflow? */ - if (!(pmu_read_register(cci_pmu, idx, CCI_PMU_OVRFLW) & - CCI_PMU_OVRFLW_FLAG)) - continue; - - pmu_write_register(cci_pmu, CCI_PMU_OVRFLW_FLAG, idx, - CCI_PMU_OVRFLW); - - pmu_event_update(event); - pmu_event_set_period(event); - handled = IRQ_HANDLED; - } - - /* Enable the PMU and sync possibly overflowed counters */ - __cci_pmu_enable_sync(cci_pmu); - raw_spin_unlock_irqrestore(&events->pmu_lock, flags); - - return IRQ_RETVAL(handled); -} - -static int cci_pmu_get_hw(struct cci_pmu *cci_pmu) -{ - int ret = pmu_request_irq(cci_pmu, pmu_handle_irq); - if (ret) { |