Magic-Mirror/llvm-capstone
1
0
Fork 0
mirror of https://github.com/capstone-engine/llvm-capstone.git synced 2025-02-05 00:49:43 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

[Detect] Consider nested loop profitable if entry block is not in loop

Browse Source 
In cases where the entry block of a scop was not contained in a loop that was
part of the scop region and at the same time there was a loop surrounding the
scop, we missed to count the loops in the scop and consequently did not consider
the scop profitable. We correct this by only moving to the loop parent, in case
the current loop is loop contained in the scop.

This increases the number of loops in COSMO which we assume to be profitable
from 3974 to 4981.

llvm-svn: 311863
This commit is contained in:
Tobias Grosser 2017-08-27 21:39:25 +00:00
parent f11daad18d
commit 93ab558d2e
2 changed files with 101 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
polly/lib/Analysis/ScopDetection.cpp
View File

@ -1312,8 +1312,13 @@ ScopDetection::countBeneficialLoops(Region *R, ScalarEvolution &SE,
int MaxLoopDepth = 0;
auto L = LI.getLoopFor(R->getEntry());
L = L ? R->outermostLoopInRegion(L) : nullptr;
L = L ? L->getParentLoop() : nullptr;
// If L is fully contained in R, move to first loop surrounding R. Otherwise,
// L is either nullptr or already surrounding R.
if (L && R->contains(L)) {
L = R->outermostLoopInRegion(L);
L = L->getParentLoop();
}
auto SubLoops =
L ? L->getSubLoopsVector() : std::vector<Loop *>(LI.begin(), LI.end());

94
polly/test/ScopDetect/profitability-two-nested-loops.ll Normal file
View File

@ -0,0 +1,94 @@
; RUN: opt %loadPolly -polly-detect -analyze \
; RUN: -polly-process-unprofitable=false < %s | FileCheck %s
; CHECK: Valid Region for Scop: next => bb3
;
; void foo(float A[], long p) {
; for (long x = 0; x < 1024; x++) {
; __sync_synchronize();
; if (p >= 0) {
; for (long i = 0; i < 1024; i++)
; for (long j = 0; j < 1024; j++)
; A[i + j] += j;
; }
; }
; }
;
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-pc-linux-gnu"
define void @foo(float* %A, i64 %p) {
bb:
br label %bb3
bb3: ; preds = %bb25, %bb
%x.0 = phi i64 [ 0, %bb ], [ %tmp26, %bb25 ]
%exitcond2 = icmp ne i64 %x.0, 1024
br i1 %exitcond2, label %bb5, label %bb4
bb4: ; preds = %bb3
br label %bb27
bb5: ; preds = %bb3
fence seq_cst
br label %next
next:
%tmp = icmp sgt i64 %p, -1
br i1 %tmp, label %bb6, label %bb24
bb6: ; preds = %bb5
br label %bb7
bb7: ; preds = %bb21, %bb6
%i.0 = phi i64 [ 0, %bb6 ], [ %tmp22, %bb21 ]
%exitcond1 = icmp ne i64 %i.0, 1024
br i1 %exitcond1, label %bb9, label %bb8
bb8: ; preds = %bb7
br label %bb23
bb9: ; preds = %bb7
br label %bb10
bb10: ; preds = %bb18, %bb9
%j.0 = phi i64 [ 0, %bb9 ], [ %tmp19, %bb18 ]
%exitcond = icmp ne i64 %j.0, 1024
br i1 %exitcond, label %bb12, label %bb11
bb11: ; preds = %bb10
br label %bb20
bb12: ; preds = %bb10
%tmp13 = sitofp i64 %j.0 to float
%tmp14 = add nuw nsw i64 %i.0, %j.0
%tmp15 = getelementptr inbounds float, float* %A, i64 %tmp14
%tmp16 = load float, float* %tmp15, align 4
%tmp17 = fadd float %tmp16, %tmp13
store float %tmp17, float* %tmp15, align 4
br label %bb18
bb18: ; preds = %bb12
%tmp19 = add nuw nsw i64 %j.0, 1
br label %bb10
bb20: ; preds = %bb11
br label %bb21
bb21: ; preds = %bb20
%tmp22 = add nuw nsw i64 %i.0, 1
br label %bb7
bb23: ; preds = %bb8
br label %bb24
bb24: ; preds = %bb23, %bb5
br label %bb25
bb25: ; preds = %bb24
%tmp26 = add nuw nsw i64 %x.0, 1
br label %bb3
bb27: ; preds = %bb4
ret void
}