leandrof/linux
1
0
Fork 0
forked from Minki/linux
Code Issues Pull Requests Packages Projects Releases Wiki Activity

powerpc/perf: Use Instruction Counter value

Browse Source 
Since PM_INST_DISP include speculative instruction,
based on the workload the dispatch count could vary
considerably. Hence as an alternative, for completed
instruction counting, program the PM_INST_DISP event
to the MMCR* but use Instruction Counter register value.

Signed-off-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This commit is contained in:
Madhavan Srinivasan 2017-02-12 22:33:14 +05:30 committed by Michael Ellerman
parent 8a088542c8
commit 356d8ce3d0
1 changed files with 36 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

36
arch/powerpc/perf/core-book3s.c
View File

@ -57,6 +57,7 @@ struct cpu_hw_events {
void *bhrb_context; void *bhrb_context;
struct perf_branch_stack bhrb_stack; struct perf_branch_stack bhrb_stack;
struct perf_branch_entry bhrb_entries[BHRB_MAX_ENTRIES]; struct perf_branch_entry bhrb_entries[BHRB_MAX_ENTRIES];
u64 ic_init;
}; };
static DEFINE_PER_CPU(struct cpu_hw_events, cpu_hw_events); static DEFINE_PER_CPU(struct cpu_hw_events, cpu_hw_events);
@ -127,6 +128,10 @@ static inline void power_pmu_bhrb_disable(struct perf_event *event) {}
static void power_pmu_sched_task(struct perf_event_context *ctx, bool sched_in) {} static void power_pmu_sched_task(struct perf_event_context *ctx, bool sched_in) {}
static inline void power_pmu_bhrb_read(struct cpu_hw_events *cpuhw) {} static inline void power_pmu_bhrb_read(struct cpu_hw_events *cpuhw) {}
static void pmao_restore_workaround(bool ebb) { } static void pmao_restore_workaround(bool ebb) { }
static bool use_ic(u64 event)
{
return false;
}
#endif /* CONFIG_PPC32 */ #endif /* CONFIG_PPC32 */
static bool regs_use_siar(struct pt_regs *regs) static bool regs_use_siar(struct pt_regs *regs)
@ -688,6 +693,15 @@ static void pmao_restore_workaround(bool ebb)
mtspr(SPRN_PMC5, pmcs[4]); mtspr(SPRN_PMC5, pmcs[4]);
mtspr(SPRN_PMC6, pmcs[5]); mtspr(SPRN_PMC6, pmcs[5]);
} }
static bool use_ic(u64 event)
{
if (cpu_has_feature(CPU_FTR_POWER9_DD1) &&
(event == 0x200f2 || event == 0x300f2))
return true;
return false;
}
#endif /* CONFIG_PPC64 */ #endif /* CONFIG_PPC64 */
static void perf_event_interrupt(struct pt_regs *regs); static void perf_event_interrupt(struct pt_regs *regs);
@ -1007,6 +1021,7 @@ static u64 check_and_compute_delta(u64 prev, u64 val)
static void power_pmu_read(struct perf_event *event) static void power_pmu_read(struct perf_event *event)
{ {
s64 val, delta, prev; s64 val, delta, prev;
struct cpu_hw_events *cpuhw = this_cpu_ptr(&cpu_hw_events);
if (event->hw.state & PERF_HES_STOPPED) if (event->hw.state & PERF_HES_STOPPED)
return; return;
@ -1016,6 +1031,13 @@ static void power_pmu_read(struct perf_event *event)
if (is_ebb_event(event)) { if (is_ebb_event(event)) {
val = read_pmc(event->hw.idx); val = read_pmc(event->hw.idx);
if (use_ic(event->attr.config)) {
val = mfspr(SPRN_IC);
if (val > cpuhw->ic_init)
val = val - cpuhw->ic_init;
else
val = val + (0 - cpuhw->ic_init);
}
local64_set(&event->hw.prev_count, val); local64_set(&event->hw.prev_count, val);
return; return;
} }
@ -1029,6 +1051,13 @@ static void power_pmu_read(struct perf_event *event)
prev = local64_read(&event->hw.prev_count); prev = local64_read(&event->hw.prev_count);
barrier(); barrier();
val = read_pmc(event->hw.idx); val = read_pmc(event->hw.idx);
if (use_ic(event->attr.config)) {
val = mfspr(SPRN_IC);
if (val > cpuhw->ic_init)
val = val - cpuhw->ic_init;
else
val = val + (0 - cpuhw->ic_init);
}
delta = check_and_compute_delta(prev, val); delta = check_and_compute_delta(prev, val);
if (!delta) if (!delta)
return; return;
@ -1466,6 +1495,13 @@ nocheck:
event->attr.branch_sample_type); event->attr.branch_sample_type);
} }
/*
* Workaround for POWER9 DD1 to use the Instruction Counter
* register value for instruction counting
*/
if (use_ic(event->attr.config))
cpuhw->ic_init = mfspr(SPRN_IC);
perf_pmu_enable(event->pmu); perf_pmu_enable(event->pmu);
local_irq_restore(flags); local_irq_restore(flags);
return ret; return ret;