Time Spent

The profiled code ran for {{TotalTime}}ms. Of this, {{GCOverheadTime}}ms were spent on garbage collection (that's {{GCOverheadTimePercent}}%).

The dynamic optimizer was active for {{SpeshTimePercent}}% of the program's run time.

Executing Code	{{ExecutingTimePercent}}% ({{ExecutingTime}}ms)
Garbage Collection	{{GCTimePercent}}% ({{GCTime}}ms)
Dynamic Optimization	{{SpeshTimePercent}}% ({{SpeshTime}}ms)

Call Frames

In total, {{EntriesWithoutInline}} call frames were entered and exited by the profiled code. Inlining eliminated the need to create {{EntriesInline}} call frames (that's {{InlinePercent}}%).

Interpreted Frames	{{InterpFramesPercent}}% ({{InterpFrames}})
Specialized Frames	{{SpeshFramesPercent}}% ({{SpeshFrames}})
JIT-Compiled Frames	{{JITFramesPercent}}% ({{JITFrames}})

Garbage Collection

The profiled code did {{GCRuns}} garbage collections. There were {{FullGCRuns}} full collections involving the entire heap.

The average nursery collection time was {{NurseryAverage}}ms. The average full collection time was {{FullAverage}}ms. On average, {{AverageParticipants}} threads participated in GC.

Scalar replacement eliminated {{ScalarReplacements}} allocations (that's {{ScalarReplacementPercent}}%).

Dynamic Optimization

Of {{OptimizedFrames}} specialized or JIT-compiled frames, there were {{DeoptOnes}} deoptimizations (that's {{DeoptOnePercent}}% of all optimized frames).

There was no global deoptimization triggered by the profiled code.

There was one global deoptimization triggered by the profiled code.

There were {{DeoptAlls}} global deoptimization triggered by the profiled code.

There was no On Stack Replacement performed while executing the profiled code (normal if the code lacks long-running loops with many iterations).

There was one On Stack Replacement performed while executing the profiled code.

There were {{OSRs}} On Stack Replacements performed while executing the profiled code.

Name	Entries	Inclusive Time	Exclusive Time	Interp / Spesh / JIT
{{routine.Name}} {{routine.File}}:{{routine.Line}}	{{routine.Entries}}	{{routine.InclusiveTimePercent}}% ({{routine.InclusiveTime}}ms)	{{routine.ExclusiveTimePercent}}% ({{routine.ExclusiveTime}}ms)	OSR

1. Callers:
2. None
3. {{caller.name == '' ? '<anon>' : caller.name}}

{{Current.name == '' ? '<anon>' : Current.name}}
{{File}}:{{Line}}

Calls (Inlined)	{{Entries}} {{Entries}} + {{InlineEntries}} ({{InlinePercent}}%)
Interpreted Calls	{{InterpPercent}}% ({{InterpEntries}})
Specialized Calls	{{SpeshPercent}}% ({{SpeshEntries}})
JIT-Compiled Calls	{{JITPercent}}% ({{JITEntries}})

Name	Allocations	Allocating Routines
{{alloc.Name}}	{{alloc.Allocations}}	View

Start Time	Run	Full	Time	Nursery: Retained / Promoted / Freed
{{gc.StartTime}}ms	{{gc.Run}}		{{gc.Time}}ms	{{gc.RetainedKilobytes}}KB / {{gc.PromotedKilobytes}}KB / {{gc.ClearedKilobytes}}KB ... {{gc.Gen2Roots}} gen2 roots

OSR

On Stack Replacement detects routines containing hot loops that are being interpreted, and replaces them with specialized or JIT-compiled code.

Routine	On Stack Replacements
{{osr.Name}} {{osr.File}}:{{osr.Line}}	{{osr.Count}}

No OSR was performed during this profile.

Local Deoptimization

Local deoptimization happens when a guard in specialized or JIT-compiled code fails. Since the code was produced assuming the guard would hold, the VM falls back to running the safe, but slower, interpreted code.

Routine	Deoptimizations
{{deopt.Name}} {{deopt.File}}:{{deopt.Line}}	{{deopt.Count}}

No local deoptimizations occurred during this profile.

Global Deoptimization

Global deoptimization happens when an event occurs that renders all currently type-specialized or JIT-compiled code on the call stack potentially invalid. Mixins - changing the type of an object in place - are a common reason.

Routine	Deoptimizations
{{deopt.Name}} {{deopt.File}}:{{deopt.Line}}	{{deopt.Count}}

No global deoptimizations occurred during this profile.