[BOLT] Add profile density computation

Reuse the definition of profile density from llvm-profgen (#92144):
- the density is computed in perf2bolt using raw samples (perf.data or
  pre-aggregated data),
- function density is the ratio of dynamically executed function bytes
  to the static function size in bytes,
- profile density:
  - functions are sorted by density in decreasing order, accumulating
    their respective sample counts,
  - profile density is the smallest density covering 99% of total sample
    count.

In other words, BOLT binary profile density is the minimum amount of
profile information per function (excluding functions in tail 1% sample
count) which is sufficient to optimize the binary well.

The density threshold of 60 was determined through experiments with
large binaries by reducing the sample count and checking resulting
profile density and performance. The threshold is conservative.

perf2bolt would print the warning if the density is below the threshold
and suggest to increase the sampling duration and/or frequency to reach
a given density, e.g.:
```
BOLT-WARNING: BOLT is estimated to optimize better with 2.8x more samples.
```

Test Plan: updated pre-aggregated-perf.test

Reviewers: maksfb, wlei-llvm, rafaelauler, ayermolo, dcci, WenleiHe

Reviewed By: WenleiHe, wlei-llvm

Pull Request: https://github.com/llvm/llvm-project/pull/101094
This commit is contained in:
Amir Ayupov
2024-10-24 18:30:59 -07:00
committed by GitHub
parent 08916cef7e
commit 6ee5ff95ab
7 changed files with 106 additions and 3 deletions

View File

@@ -849,6 +849,12 @@ bool DataAggregator::doTrace(const LBREntry &First, const LBREntry &Second,
return false;
}
// Set ParentFunc to BAT parent function or FromFunc itself.
BinaryFunction *ParentFunc = getBATParentFunction(*FromFunc);
if (!ParentFunc)
ParentFunc = FromFunc;
ParentFunc->SampleCountInBytes += Count * (Second.From - First.To);
std::optional<BoltAddressTranslation::FallthroughListTy> FTs =
BAT ? BAT->getFallthroughsInTrace(FromFunc->getAddress(), First.To,
Second.From)
@@ -868,13 +874,12 @@ bool DataAggregator::doTrace(const LBREntry &First, const LBREntry &Second,
<< FromFunc->getPrintName() << ":"
<< Twine::utohexstr(First.To) << " to "
<< Twine::utohexstr(Second.From) << ".\n");
BinaryFunction *ParentFunc = getBATParentFunction(*FromFunc);
for (auto [From, To] : *FTs) {
if (BAT) {
From = BAT->translate(FromFunc->getAddress(), From, /*IsBranchSrc=*/true);
To = BAT->translate(FromFunc->getAddress(), To, /*IsBranchSrc=*/false);
}
doIntraBranch(ParentFunc ? *ParentFunc : *FromFunc, From, To, Count, false);
doIntraBranch(*ParentFunc, From, To, Count, false);
}
return true;