[LoopVectorize] Use AA to partition potential dependency checks
Prior to this change, the loop vectorizer did not make use of the alias
analysis infrastructure. Instead, it performed memory dependence analysis using
ScalarEvolution-based linear dependence checks within equivalence classes
derived from the results of ValueTracking's GetUnderlyingObjects.
Unfortunately, this meant that:
1. The loop vectorizer had logic that essentially duplicated that in BasicAA
for aliasing based on identified objects.
2. The loop vectorizer could not partition the space of dependency checks
based on information only easily available from within AA (TBAA metadata is
currently the prime example).
This means, for example, regardless of whether -fno-strict-aliasing was
provided, the vectorizer would only vectorize this loop with a runtime
memory-overlap check:
void foo(int *a, float *b) {
for (int i = 0; i < 1600; ++i)
a[i] = b[i];
}
This is suboptimal because the TBAA metadata already provides the information
necessary to show that this check unnecessary. Of course, the vectorizer has a
limit on the number of such checks it will insert, so in practice, ignoring
TBAA means not vectorizing more-complicated loops that we should.
This change causes the vectorizer to use an AliasSetTracker to keep track of
the pointers in the loop. The resulting alias sets are then used to partition
the space of dependency checks, and potential runtime checks; this results in
more-efficient vectorizations.
When pointer locations are added to the AliasSetTracker, two things are done:
1. The location size is set to UnknownSize (otherwise you'd not catch
inter-iteration dependencies)
2. For instructions in blocks that would need to be predicated, TBAA is
removed (because the metadata might have a control dependency on the condition
being speculated).
For non-predicated blocks, you can leave the TBAA metadata. This is safe
because you can't have an iteration dependency on the TBAA metadata (if you
did, and you unrolled sufficiently, you'd end up with the same pointer value
used by two accesses that TBAA says should not alias, and that would yield
undefined behavior).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@213486 91177308-0d34-0410-b5e6-96231b3b80d8
2014-07-20 23:07:52 +00:00
|
|
|
; RUN: opt -S -basicaa -loop-vectorize -force-vector-unroll=1 -force-vector-width=4 -dce -instcombine < %s | FileCheck %s
|
2013-06-27 00:45:41 +00:00
|
|
|
|
|
|
|
|
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"
|
|
|
|
|
target triple = "x86_64-apple-macosx10.8.0"
|
|
|
|
|
|
|
|
|
|
@uf = common global [100 x i32] zeroinitializer, align 16
|
|
|
|
|
@xi = common global [100 x i32] zeroinitializer, align 16
|
|
|
|
|
@q = common global [100 x i32] zeroinitializer, align 16
|
|
|
|
|
|
|
|
|
|
; PR16455
|
|
|
|
|
|
|
|
|
|
|
2013-06-27 01:01:11 +00:00
|
|
|
; Due to a bug in the way we handled reverse induction stores we would generate
|
2013-06-27 00:45:41 +00:00
|
|
|
; a shuffle too many.
|
|
|
|
|
|
|
|
|
|
define void @t() {
|
|
|
|
|
entry:
|
|
|
|
|
br label %for.body
|
|
|
|
|
|
2013-07-14 01:42:54 +00:00
|
|
|
; CHECK-LABEL: @t(
|
2013-06-27 00:45:41 +00:00
|
|
|
; CHECK: vector.body:
|
2014-05-11 08:46:12 +00:00
|
|
|
; CHECK: [[VAR1:%[a-zA-Z0-9.]+]] = load <4 x i32>
|
|
|
|
|
; CHECK: [[VAR2:%[a-zA-Z0-9.]+]] = load <4 x i32>
|
2013-06-27 00:45:41 +00:00
|
|
|
; CHECK: [[VAR3:%[a-zA-Z0-9]+]] = add nsw <4 x i32> [[VAR2]], [[VAR1]]
|
2014-05-11 08:46:12 +00:00
|
|
|
; CHECK: store <4 x i32> [[VAR3]]
|
|
|
|
|
; CHECK: [[VAR4:%[a-zA-Z0-9.]+]] = load <4 x i32>
|
|
|
|
|
; CHECK: add nsw <4 x i32> [[VAR3]], [[VAR4]]
|
|
|
|
|
; CHECK-NOT: shufflevector
|
2013-06-27 00:45:41 +00:00
|
|
|
|
|
|
|
|
for.body:
|
|
|
|
|
%indvars.iv = phi i64 [ 93, %entry ], [ %indvars.iv.next, %for.body ]
|
|
|
|
|
%0 = add i64 %indvars.iv, 1
|
|
|
|
|
%arrayidx = getelementptr inbounds [100 x i32]* @uf, i64 0, i64 %0
|
|
|
|
|
%arrayidx3 = getelementptr inbounds [100 x i32]* @xi, i64 0, i64 %0
|
|
|
|
|
%1 = load i32* %arrayidx3, align 4
|
|
|
|
|
%2 = load i32* %arrayidx, align 4
|
|
|
|
|
%add4 = add nsw i32 %2, %1
|
|
|
|
|
store i32 %add4, i32* %arrayidx, align 4
|
|
|
|
|
%arrayidx7 = getelementptr inbounds [100 x i32]* @q, i64 0, i64 %0
|
|
|
|
|
%3 = load i32* %arrayidx7, align 4
|
|
|
|
|
%add8 = add nsw i32 %add4, %3
|
|
|
|
|
store i32 %add8, i32* %arrayidx, align 4
|
|
|
|
|
%indvars.iv.next = add i64 %indvars.iv, -1
|
|
|
|
|
%4 = trunc i64 %indvars.iv.next to i32
|
|
|
|
|
%cmp = icmp ugt i32 %4, 2
|
|
|
|
|
br i1 %cmp, label %for.body, label %for.end
|
|
|
|
|
|
|
|
|
|
for.end:
|
|
|
|
|
ret void
|
|
|
|
|
}
|
