diff options
author | Nadav Rotem <nrotem@apple.com> | 2013-05-03 17:42:55 +0000 |
---|---|---|
committer | Nadav Rotem <nrotem@apple.com> | 2013-05-03 17:42:55 +0000 |
commit | 4bcd5f888fa762613cf8096a79ba7b8a72665de2 (patch) | |
tree | 1dfc096d2c67d5b3fbf51252b403f5ce3ebede39 /test | |
parent | 19301d5d1234d032d42f20deb6f3076c972fd5f4 (diff) | |
download | external_llvm-4bcd5f888fa762613cf8096a79ba7b8a72665de2.zip external_llvm-4bcd5f888fa762613cf8096a79ba7b8a72665de2.tar.gz external_llvm-4bcd5f888fa762613cf8096a79ba7b8a72665de2.tar.bz2 |
LoopVectorizer: Add support for if-conversion of PHINodes with 3+ incoming values.
By supporting the vectorization of PHINodes with more than two incoming values we can increase the complexity of nested if statements.
We can now vectorize this loop:
int foo(int *A, int *B, int n) {
for (int i=0; i < n; i++) {
int x = 9;
if (A[i] > B[i]) {
if (A[i] > 19) {
x = 3;
} else if (B[i] < 4 ) {
x = 4;
} else {
x = 5;
}
}
A[i] = x;
}
}
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@181037 91177308-0d34-0410-b5e6-96231b3b80d8
Diffstat (limited to 'test')
-rw-r--r-- | test/Transforms/LoopVectorize/if-conversion-nest.ll | 48 |
1 files changed, 48 insertions, 0 deletions
diff --git a/test/Transforms/LoopVectorize/if-conversion-nest.ll b/test/Transforms/LoopVectorize/if-conversion-nest.ll new file mode 100644 index 0000000..f44862a --- /dev/null +++ b/test/Transforms/LoopVectorize/if-conversion-nest.ll @@ -0,0 +1,48 @@ +; RUN: opt < %s -loop-vectorize -force-vector-unroll=1 -force-vector-width=4 -enable-if-conversion -dce -instcombine -S | FileCheck %s + +target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128" + +;CHECK: @foo +;CHECK: icmp sgt +;CHECK: icmp sgt +;CHECK: icmp slt +;CHECK: select <4 x i1> +;CHECK: %[[P1:.*]] = select <4 x i1> +;CHECK: xor <4 x i1> +;CHECK: and <4 x i1> +;CHECK: select <4 x i1> %{{.*}}, <4 x i32> %{{.*}}, <4 x i32> %[[P1]] +;CHECK: ret +define i32 @foo(i32* nocapture %A, i32* nocapture %B, i32 %n) { +entry: + %cmp26 = icmp sgt i32 %n, 0 + br i1 %cmp26, label %for.body, label %for.end + +for.body: + %indvars.iv = phi i64 [ %indvars.iv.next, %if.end14 ], [ 0, %entry ] + %arrayidx = getelementptr inbounds i32* %A, i64 %indvars.iv + %0 = load i32* %arrayidx, align 4 + %arrayidx2 = getelementptr inbounds i32* %B, i64 %indvars.iv + %1 = load i32* %arrayidx2, align 4 + %cmp3 = icmp sgt i32 %0, %1 + br i1 %cmp3, label %if.then, label %if.end14 + +if.then: + %cmp6 = icmp sgt i32 %0, 19 + br i1 %cmp6, label %if.end14, label %if.else + +if.else: + %cmp10 = icmp slt i32 %1, 4 + %. = select i1 %cmp10, i32 4, i32 5 + br label %if.end14 + +if.end14: + %x.0 = phi i32 [ 9, %for.body ], [ 3, %if.then ], [ %., %if.else ] ; <------------- A PHI with 3 entries that we can still vectorize. + store i32 %x.0, i32* %arrayidx, align 4 + %indvars.iv.next = add i64 %indvars.iv, 1 + %lftr.wideiv = trunc i64 %indvars.iv.next to i32 + %exitcond = icmp eq i32 %lftr.wideiv, %n + br i1 %exitcond, label %for.end, label %for.body + +for.end: + ret i32 undef +} |