diff options
Diffstat (limited to 'arch/x86/oprofile/op_model_amd.c')
-rw-r--r-- | arch/x86/oprofile/op_model_amd.c | 267 |
1 files changed, 118 insertions, 149 deletions
diff --git a/arch/x86/oprofile/op_model_amd.c b/arch/x86/oprofile/op_model_amd.c index 8fdf06e4edf..cc930467575 100644 --- a/arch/x86/oprofile/op_model_amd.c +++ b/arch/x86/oprofile/op_model_amd.c @@ -25,43 +25,28 @@ #define NUM_COUNTERS 4 #define NUM_CONTROLS 4 +#define OP_EVENT_MASK 0x0FFF +#define OP_CTR_OVERFLOW (1ULL<<31) -#define CTR_IS_RESERVED(msrs, c) (msrs->counters[(c)].addr ? 1 : 0) -#define CTR_READ(l, h, msrs, c) do {rdmsr(msrs->counters[(c)].addr, (l), (h)); } while (0) -#define CTR_WRITE(l, msrs, c) do {wrmsr(msrs->counters[(c)].addr, -(unsigned int)(l), -1); } while (0) -#define CTR_OVERFLOWED(n) (!((n) & (1U<<31))) - -#define CTRL_IS_RESERVED(msrs, c) (msrs->controls[(c)].addr ? 1 : 0) -#define CTRL_READ(l, h, msrs, c) do {rdmsr(msrs->controls[(c)].addr, (l), (h)); } while (0) -#define CTRL_WRITE(l, h, msrs, c) do {wrmsr(msrs->controls[(c)].addr, (l), (h)); } while (0) -#define CTRL_SET_ACTIVE(n) (n |= (1<<22)) -#define CTRL_SET_INACTIVE(n) (n &= ~(1<<22)) -#define CTRL_CLEAR_LO(x) (x &= (1<<21)) -#define CTRL_CLEAR_HI(x) (x &= 0xfffffcf0) -#define CTRL_SET_ENABLE(val) (val |= 1<<20) -#define CTRL_SET_USR(val, u) (val |= ((u & 1) << 16)) -#define CTRL_SET_KERN(val, k) (val |= ((k & 1) << 17)) -#define CTRL_SET_UM(val, m) (val |= (m << 8)) -#define CTRL_SET_EVENT_LOW(val, e) (val |= (e & 0xff)) -#define CTRL_SET_EVENT_HIGH(val, e) (val |= ((e >> 8) & 0xf)) -#define CTRL_SET_HOST_ONLY(val, h) (val |= ((h & 1) << 9)) -#define CTRL_SET_GUEST_ONLY(val, h) (val |= ((h & 1) << 8)) +#define MSR_AMD_EVENTSEL_RESERVED ((0xFFFFFCF0ULL<<32)|(1ULL<<21)) static unsigned long reset_value[NUM_COUNTERS]; #ifdef CONFIG_OPROFILE_IBS /* IbsFetchCtl bits/masks */ -#define IBS_FETCH_HIGH_VALID_BIT (1UL << 17) /* bit 49 */ -#define IBS_FETCH_HIGH_ENABLE (1UL << 16) /* bit 48 */ -#define IBS_FETCH_LOW_MAX_CNT_MASK 0x0000FFFFUL /* MaxCnt mask */ +#define IBS_FETCH_RAND_EN (1ULL<<57) +#define IBS_FETCH_VAL (1ULL<<49) +#define IBS_FETCH_ENABLE (1ULL<<48) +#define IBS_FETCH_CNT_MASK 0xFFFF0000ULL /*IbsOpCtl bits */ -#define IBS_OP_LOW_VALID_BIT (1ULL<<18) /* bit 18 */ -#define IBS_OP_LOW_ENABLE (1ULL<<17) /* bit 17 */ +#define IBS_OP_CNT_CTL (1ULL<<19) +#define IBS_OP_VAL (1ULL<<18) +#define IBS_OP_ENABLE (1ULL<<17) -#define IBS_FETCH_SIZE 6 -#define IBS_OP_SIZE 12 +#define IBS_FETCH_SIZE 6 +#define IBS_OP_SIZE 12 static int has_ibs; /* AMD Family10h and later */ @@ -99,49 +84,38 @@ static void op_amd_fill_in_addresses(struct op_msrs * const msrs) } } - -static void op_amd_setup_ctrs(struct op_msrs const * const msrs) +static void op_amd_setup_ctrs(struct op_x86_model_spec const *model, + struct op_msrs const * const msrs) { - unsigned int low, high; + u64 val; int i; /* clear all counters */ for (i = 0 ; i < NUM_CONTROLS; ++i) { - if (unlikely(!CTRL_IS_RESERVED(msrs, i))) + if (unlikely(!msrs->controls[i].addr)) continue; - CTRL_READ(low, high, msrs, i); - CTRL_CLEAR_LO(low); - CTRL_CLEAR_HI(high); - CTRL_WRITE(low, high, msrs, i); + rdmsrl(msrs->controls[i].addr, val); + val &= model->reserved; + wrmsrl(msrs->controls[i].addr, val); } /* avoid a false detection of ctr overflows in NMI handler */ for (i = 0; i < NUM_COUNTERS; ++i) { - if (unlikely(!CTR_IS_RESERVED(msrs, i))) + if (unlikely(!msrs->counters[i].addr)) continue; - CTR_WRITE(1, msrs, i); + wrmsrl(msrs->counters[i].addr, -1LL); } /* enable active counters */ for (i = 0; i < NUM_COUNTERS; ++i) { - if ((counter_config[i].enabled) && (CTR_IS_RESERVED(msrs, i))) { + if (counter_config[i].enabled && msrs->counters[i].addr) { reset_value[i] = counter_config[i].count; - - CTR_WRITE(counter_config[i].count, msrs, i); - - CTRL_READ(low, high, msrs, i); - CTRL_CLEAR_LO(low); - CTRL_CLEAR_HI(high); - CTRL_SET_ENABLE(low); - CTRL_SET_USR(low, counter_config[i].user); - CTRL_SET_KERN(low, counter_config[i].kernel); - CTRL_SET_UM(low, counter_config[i].unit_mask); - CTRL_SET_EVENT_LOW(low, counter_config[i].event); - CTRL_SET_EVENT_HIGH(high, counter_config[i].event); - CTRL_SET_HOST_ONLY(high, 0); - CTRL_SET_GUEST_ONLY(high, 0); - - CTRL_WRITE(low, high, msrs, i); + wrmsrl(msrs->counters[i].addr, + -(s64)counter_config[i].count); + rdmsrl(msrs->controls[i].addr, val); + val &= model->reserved; + val |= op_x86_get_ctrl(model, &counter_config[i]); + wrmsrl(msrs->controls[i].addr, val); } else { reset_value[i] = 0; } @@ -154,93 +128,116 @@ static inline int op_amd_handle_ibs(struct pt_regs * const regs, struct op_msrs const * const msrs) { - u32 low, high; - u64 msr; + u64 val, ctl; struct op_entry entry; if (!has_ibs) return 1; if (ibs_config.fetch_enabled) { - rdmsr(MSR_AMD64_IBSFETCHCTL, low, high); - if (high & IBS_FETCH_HIGH_VALID_BIT) { - rdmsrl(MSR_AMD64_IBSFETCHLINAD, msr); - oprofile_write_reserve(&entry, regs, msr, + rdmsrl(MSR_AMD64_IBSFETCHCTL, ctl); + if (ctl & IBS_FETCH_VAL) { + rdmsrl(MSR_AMD64_IBSFETCHLINAD, val); + oprofile_write_reserve(&entry, regs, val, IBS_FETCH_CODE, IBS_FETCH_SIZE); - oprofile_add_data(&entry, (u32)msr); - oprofile_add_data(&entry, (u32)(msr >> 32)); - oprofile_add_data(&entry, low); - oprofile_add_data(&entry, high); - rdmsrl(MSR_AMD64_IBSFETCHPHYSAD, msr); - oprofile_add_data(&entry, (u32)msr); - oprofile_add_data(&entry, (u32)(msr >> 32)); + oprofile_add_data64(&entry, val); + oprofile_add_data64(&entry, ctl); + rdmsrl(MSR_AMD64_IBSFETCHPHYSAD, val); + oprofile_add_data64(&entry, val); oprofile_write_commit(&entry); /* reenable the IRQ */ - high &= ~IBS_FETCH_HIGH_VALID_BIT; - high |= IBS_FETCH_HIGH_ENABLE; - low &= IBS_FETCH_LOW_MAX_CNT_MASK; - wrmsr(MSR_AMD64_IBSFETCHCTL, low, high); + ctl &= ~(IBS_FETCH_VAL | IBS_FETCH_CNT_MASK); + ctl |= IBS_FETCH_ENABLE; + wrmsrl(MSR_AMD64_IBSFETCHCTL, ctl); } } if (ibs_config.op_enabled) { - rdmsr(MSR_AMD64_IBSOPCTL, low, high); - if (low & IBS_OP_LOW_VALID_BIT) { - rdmsrl(MSR_AMD64_IBSOPRIP, msr); - oprofile_write_reserve(&entry, regs, msr, + rdmsrl(MSR_AMD64_IBSOPCTL, ctl); + if (ctl & IBS_OP_VAL) { + rdmsrl(MSR_AMD64_IBSOPRIP, val); + oprofile_write_reserve(&entry, regs, val, IBS_OP_CODE, IBS_OP_SIZE); - oprofile_add_data(&entry, (u32)msr); - oprofile_add_data(&entry, (u32)(msr >> 32)); - rdmsrl(MSR_AMD64_IBSOPDATA, msr); - oprofile_add_data(&entry, (u32)msr); - oprofile_add_data(&entry, (u32)(msr >> 32)); - rdmsrl(MSR_AMD64_IBSOPDATA2, msr); - oprofile_add_data(&entry, (u32)msr); - oprofile_add_data(&entry, (u32)(msr >> 32)); - rdmsrl(MSR_AMD64_IBSOPDATA3, msr); - oprofile_add_data(&entry, (u32)msr); - oprofile_add_data(&entry, (u32)(msr >> 32)); - rdmsrl(MSR_AMD64_IBSDCLINAD, msr); - oprofile_add_data(&entry, (u32)msr); - oprofile_add_data(&entry, (u32)(msr >> 32)); - rdmsrl(MSR_AMD64_IBSDCPHYSAD, msr); - oprofile_add_data(&entry, (u32)msr); - oprofile_add_data(&entry, (u32)(msr >> 32)); + oprofile_add_data64(&entry, val); + rdmsrl(MSR_AMD64_IBSOPDATA, val); + oprofile_add_data64(&entry, val); + rdmsrl(MSR_AMD64_IBSOPDATA2, val); + oprofile_add_data64(&entry, val); + rdmsrl(MSR_AMD64_IBSOPDATA3, val); + oprofile_add_data64(&entry, val); + rdmsrl(MSR_AMD64_IBSDCLINAD, val); + oprofile_add_data64(&entry, val); + rdmsrl(MSR_AMD64_IBSDCPHYSAD, val); + oprofile_add_data64(&entry, val); oprofile_write_commit(&entry); /* reenable the IRQ */ - high = 0; - low &= ~IBS_OP_LOW_VALID_BIT; - low |= IBS_OP_LOW_ENABLE; - wrmsr(MSR_AMD64_IBSOPCTL, low, high); + ctl &= ~IBS_OP_VAL & 0xFFFFFFFF; + ctl |= IBS_OP_ENABLE; + wrmsrl(MSR_AMD64_IBSOPCTL, ctl); } } return 1; } +static inline void op_amd_start_ibs(void) +{ + u64 val; + if (has_ibs && ibs_config.fetch_enabled) { + val = (ibs_config.max_cnt_fetch >> 4) & 0xFFFF; + val |= ibs_config.rand_en ? IBS_FETCH_RAND_EN : 0; + val |= IBS_FETCH_ENABLE; + wrmsrl(MSR_AMD64_IBSFETCHCTL, val); + } + + if (has_ibs && ibs_config.op_enabled) { + val = (ibs_config.max_cnt_op >> 4) & 0xFFFF; + val |= ibs_config.dispatched_ops ? IBS_OP_CNT_CTL : 0; + val |= IBS_OP_ENABLE; + wrmsrl(MSR_AMD64_IBSOPCTL, val); + } +} + +static void op_amd_stop_ibs(void) +{ + if (has_ibs && ibs_config.fetch_enabled) + /* clear max count and enable */ + wrmsrl(MSR_AMD64_IBSFETCHCTL, 0); + + if (has_ibs && ibs_config.op_enabled) + /* clear max count and enable */ + wrmsrl(MSR_AMD64_IBSOPCTL, 0); +} + +#else + +static inline int op_amd_handle_ibs(struct pt_regs * const regs, + struct op_msrs const * const msrs) { } +static inline void op_amd_start_ibs(void) { } +static inline void op_amd_stop_ibs(void) { } + #endif static int op_amd_check_ctrs(struct pt_regs * const regs, struct op_msrs const * const msrs) { - unsigned int low, high; + u64 val; int i; for (i = 0 ; i < NUM_COUNTERS; ++i) { if (!reset_value[i]) continue; - CTR_READ(low, high, msrs, i); - if (CTR_OVERFLOWED(low)) { - oprofile_add_sample(regs, i); - CTR_WRITE(reset_value[i], msrs, i); - } + rdmsrl(msrs->counters[i].addr, val); + /* bit is clear if overflowed: */ + if (val & OP_CTR_OVERFLOW) + continue; + oprofile_add_sample(regs, i); + wrmsrl(msrs->counters[i].addr, -(s64)reset_value[i]); } -#ifdef CONFIG_OPROFILE_IBS op_amd_handle_ibs(regs, msrs); -#endif /* See op_model_ppro.c */ return 1; @@ -248,38 +245,22 @@ static int op_amd_check_ctrs(struct pt_regs * const regs, static void op_amd_start(struct op_msrs const * const msrs) { - unsigned int low, high; + u64 val; int i; for (i = 0 ; i < NUM_COUNTERS ; ++i) { if (reset_value[i]) { - CTRL_READ(low, high, msrs, i); - CTRL_SET_ACTIVE(low); - CTRL_WRITE(low, high, msrs, i); + rdmsrl(msrs->controls[i].addr, val); + val |= ARCH_PERFMON_EVENTSEL0_ENABLE; + wrmsrl(msrs->controls[i].addr, val); } } -#ifdef CONFIG_OPROFILE_IBS - if (has_ibs && ibs_config.fetch_enabled) { - low = (ibs_config.max_cnt_fetch >> 4) & 0xFFFF; - high = ((ibs_config.rand_en & 0x1) << 25) /* bit 57 */ - + IBS_FETCH_HIGH_ENABLE; - wrmsr(MSR_AMD64_IBSFETCHCTL, low, high); - } - - if (has_ibs && ibs_config.op_enabled) { - low = ((ibs_config.max_cnt_op >> 4) & 0xFFFF) - + ((ibs_config.dispatched_ops & 0x1) << 19) /* bit 19 */ - + IBS_OP_LOW_ENABLE; - high = 0; - wrmsr(MSR_AMD64_IBSOPCTL, low, high); - } -#endif + op_amd_start_ibs(); } - static void op_amd_stop(struct op_msrs const * const msrs) { - unsigned int low, high; + u64 val; int i; /* @@ -289,26 +270,12 @@ static void op_amd_stop(struct op_msrs const * const msrs) for (i = 0 ; i < NUM_COUNTERS ; ++i) { if (!reset_value[i]) continue; - CTRL_READ(low, high, msrs, i); - CTRL_SET_INACTIVE(low); - CTRL_WRITE(low, high, msrs, i); - } - -#ifdef CONFIG_OPROFILE_IBS - if (has_ibs && ibs_config.fetch_enabled) { - /* clear max count and enable */ - low = 0; - high = 0; - wrmsr(MSR_AMD64_IBSFETCHCTL, low, high); + rdmsrl(msrs->controls[i].addr, val); + val &= ~ARCH_PERFMON_EVENTSEL0_ENABLE; + wrmsrl(msrs->controls[i].addr, val); } - if (has_ibs && ibs_config.op_enabled) { - /* clear max count and enable */ - low = 0; - high = 0; - wrmsr(MSR_AMD64_IBSOPCTL, low, high); - } -#endif + op_amd_stop_ibs(); } static void op_amd_shutdown(struct op_msrs const * const msrs) @@ -316,11 +283,11 @@ static void op_amd_shutdown(struct op_msrs const * const msrs) int i; for (i = 0 ; i < NUM_COUNTERS ; ++i) { - if (CTR_IS_RESERVED(msrs, i)) + if (msrs->counters[i].addr) release_perfctr_nmi(MSR_K7_PERFCTR0 + i); } for (i = 0 ; i < NUM_CONTROLS ; ++i) { - if (CTRL_IS_RESERVED(msrs, i)) + if (msrs->controls[i].addr) release_evntsel_nmi(MSR_K7_EVNTSEL0 + i); } } @@ -491,14 +458,16 @@ static void op_amd_exit(void) {} #endif /* CONFIG_OPROFILE_IBS */ struct op_x86_model_spec const op_amd_spec = { - .init = op_amd_init, - .exit = op_amd_exit, .num_counters = NUM_COUNTERS, .num_controls = NUM_CONTROLS, + .reserved = MSR_AMD_EVENTSEL_RESERVED, + .event_mask = OP_EVENT_MASK, + .init = op_amd_init, + .exit = op_amd_exit, .fill_in_addresses = &op_amd_fill_in_addresses, .setup_ctrs = &op_amd_setup_ctrs, .check_ctrs = &op_amd_check_ctrs, .start = &op_amd_start, .stop = &op_amd_stop, - .shutdown = &op_amd_shutdown + .shutdown = &op_amd_shutdown, }; |