MIPS: perf: Add perf functionality for BMIPS5000

Add hardware performance counter support to kernel "perf" code for
BMIPS5000. The BMIPS5000 performance counters are similar to MIPS
MTI cores, so the changes were mostly made in perf_event_mipsxx.c
which is typically for MTI cores.

Signed-off-by: Al Cooper <[e-mail]>
Cc: [e-mail]
Cc: [e-mail]
Patchwork: https://patchwork.linux-mips.org/patch/4109/
Signed-off-by: Ralf Baechle <[e-mail]>

arch/mips/kernel/perf_event_mipsxx.c

···
 #include <asm/time.h> /* For perf_irq */
 
 #define MIPS_MAX_HWEVENTS 4
+#define MIPS_TCS_PER_COUNTER 2
+#define MIPS_CPUID_TO_COUNTER_MASK (MIPS_TCS_PER_COUNTER - 1)
 
 struct cpu_hw_events {
 	/* Array of events on this cpu. */
···
 #define M_PERFCTL_INTERRUPT_ENABLE	(1      <<  4)
 #define M_PERFCTL_EVENT(event)		(((event) & 0x3ff)  << 5)
 #define M_PERFCTL_VPEID(vpe)		((vpe)    << 16)
+
+#ifdef CONFIG_CPU_BMIPS5000
+#define M_PERFCTL_MT_EN(filter)		0
+#else /* !CONFIG_CPU_BMIPS5000 */
 #define M_PERFCTL_MT_EN(filter)		((filter) << 20)
+#endif /* CONFIG_CPU_BMIPS5000 */
+
 #define    M_TC_EN_ALL			M_PERFCTL_MT_EN(0)
 #define    M_TC_EN_VPE			M_PERFCTL_MT_EN(1)
 #define    M_TC_EN_TC			M_PERFCTL_MT_EN(2)
 #define M_PERFCTL_TCID(tcid)		((tcid)   << 22)
 #define M_PERFCTL_WIDE			(1      << 30)
 #define M_PERFCTL_MORE			(1      << 31)
+#define M_PERFCTL_TC			(1      << 30)
 
 #define M_PERFCTL_COUNT_EVENT_WHENEVER	(M_PERFCTL_EXL |		\
 					M_PERFCTL_KERNEL |		\
···
 
 static DEFINE_RWLOCK(pmuint_rwlock);
 
+#if defined(CONFIG_CPU_BMIPS5000)
+#define vpe_id()	(cpu_has_mipsmt_pertccounters ? \
+			 0 : (smp_processor_id() & MIPS_CPUID_TO_COUNTER_MASK))
+#else
 /*
  * FIXME: For VSMP, vpe_id() is redefined for Perf-events, because
  * cpu_data[cpuid].vpe_id reports 0 for _both_ CPUs.
  */
 #define vpe_id()	(cpu_has_mipsmt_pertccounters ? \
-			0 : smp_processor_id())
+			 0 : smp_processor_id())
+#endif
 
 /* Copied from op_model_mipsxx.c */
 static unsigned int vpe_shift(void)
···
 		(evt->config_base & M_PERFCTL_CONFIG_MASK) |
 		/* Make sure interrupt enabled. */
 		M_PERFCTL_INTERRUPT_ENABLE;
+	if (IS_ENABLED(CONFIG_CPU_BMIPS5000))
+		/* enable the counter for the calling thread */
+		cpuc->saved_ctrl[idx] |=
+			(1 << (12 + vpe_id())) | M_PERFCTL_TC;
+
 	/*
 	 * We do not actually let the counter run. Leave it until start().
 	 */
···
 	[PERF_COUNT_HW_BUS_CYCLES] = { 0x25, CNTR_ALL },
 };
 
+static const struct mips_perf_event bmips5000_event_map
+				[PERF_COUNT_HW_MAX] = {
+	[PERF_COUNT_HW_CPU_CYCLES] = { 0x00, CNTR_EVEN | CNTR_ODD, T },
+	[PERF_COUNT_HW_INSTRUCTIONS] = { 0x01, CNTR_EVEN | CNTR_ODD, T },
+	[PERF_COUNT_HW_BRANCH_MISSES] = { 0x02, CNTR_ODD, T },
+};
+
 /* 24K/34K/1004K cores can share the same cache event map. */
 static const struct mips_perf_event mipsxxcore_cache_map
 				[PERF_COUNT_HW_CACHE_MAX]
···
 	[C(OP_WRITE)] = {
 		[C(RESULT_ACCESS)]	= { 0x27, CNTR_EVEN, T },
 		[C(RESULT_MISS)]	= { 0x27, CNTR_ODD, T },
+	},
+},
+};
+
+/* BMIPS5000 */
+static const struct mips_perf_event bmips5000_cache_map
+				[PERF_COUNT_HW_CACHE_MAX]
+				[PERF_COUNT_HW_CACHE_OP_MAX]
+				[PERF_COUNT_HW_CACHE_RESULT_MAX] = {
+[C(L1D)] = {
+	/*
+	 * Like some other architectures (e.g. ARM), the performance
+	 * counters don't differentiate between read and write
+	 * accesses/misses, so this isn't strictly correct, but it's the
+	 * best we can do. Writes and reads get combined.
+	 */
+	[C(OP_READ)] = {
+		[C(RESULT_ACCESS)]	= { 12, CNTR_EVEN, T },
+		[C(RESULT_MISS)]	= { 12, CNTR_ODD, T },
+	},
+	[C(OP_WRITE)] = {
+		[C(RESULT_ACCESS)]	= { 12, CNTR_EVEN, T },
+		[C(RESULT_MISS)]	= { 12, CNTR_ODD, T },
+	},
+},
+[C(L1I)] = {
+	[C(OP_READ)] = {
+		[C(RESULT_ACCESS)]	= { 10, CNTR_EVEN, T },
+		[C(RESULT_MISS)]	= { 10, CNTR_ODD, T },
+	},
+	[C(OP_WRITE)] = {
+		[C(RESULT_ACCESS)]	= { 10, CNTR_EVEN, T },
+		[C(RESULT_MISS)]	= { 10, CNTR_ODD, T },
+	},
+	[C(OP_PREFETCH)] = {
+		[C(RESULT_ACCESS)]	= { 23, CNTR_EVEN, T },
+		/*
+		 * Note that MIPS has only "hit" events countable for
+		 * the prefetch operation.
+		 */
+	},
+},
+[C(LL)] = {
+	[C(OP_READ)] = {
+		[C(RESULT_ACCESS)]	= { 28, CNTR_EVEN, P },
+		[C(RESULT_MISS)]	= { 28, CNTR_ODD, P },
+	},
+	[C(OP_WRITE)] = {
+		[C(RESULT_ACCESS)]	= { 28, CNTR_EVEN, P },
+		[C(RESULT_MISS)]	= { 28, CNTR_ODD, P },
+	},
+},
+[C(BPU)] = {
+	/* Using the same code for *HW_BRANCH* */
+	[C(OP_READ)] = {
+		[C(RESULT_MISS)]	= { 0x02, CNTR_ODD, T },
+	},
+	[C(OP_WRITE)] = {
+		[C(RESULT_MISS)]	= { 0x02, CNTR_ODD, T },
 	},
 },
 };
···
 #define IS_RANGE_V_1004K_EVENT(r)	((r) == 47)
 #endif
 
+/* BMIPS5000 */
+#define IS_BOTH_COUNTERS_BMIPS5000_EVENT(b)				\
+	((b) == 0 || (b) == 1)
+
+
 /*
  * User can use 0-255 raw events, where 0-127 for the events of even
  * counters, and 128-255 for odd counters. Note that bit 7 is used to
···
 			raw_event.range = T;
 #endif
 		break;
+	case CPU_BMIPS5000:
+		if (IS_BOTH_COUNTERS_BMIPS5000_EVENT(base_id))
+			raw_event.cntr_mask = CNTR_EVEN | CNTR_ODD;
+		else
+			raw_event.cntr_mask =
+				raw_id > 127 ? CNTR_ODD : CNTR_EVEN;
 	}
 
 	return &raw_event;
···
 		mipspmu.general_event_map = &octeon_event_map;
 		mipspmu.cache_event_map = &octeon_cache_map;
 		mipspmu.map_raw_event = octeon_pmu_map_raw_event;
+		break;
+	case CPU_BMIPS5000:
+		mipspmu.name = "BMIPS5000";
+		mipspmu.general_event_map = &bmips5000_event_map;
+		mipspmu.cache_event_map = &bmips5000_cache_map;
 		break;
 	default:
 		pr_cont("Either hardware does not support performance "