456ec1c2f4
We are not handling 'S' scalar dependencies correctly and have at least the following miscompiles related to that: [LoopInterchange] incorrect handling of scalar dependencies and dependence vectors starting with ">" #54176 [LoopInterchange] Interchange breaks program correctness #46867 [LoopInterchange] Loops should not interchanged due to dependencies #47259 [LoopInterchange] Loops should not interchanged due to control flow #47401 This patch does no longer insert the "S" dependency/direction into the dependency matrix, so a dependency is never "S". We seem to have forgotten what the exact meaning is of this dependency type, and don't see why it should be treated differently. We prefer correctness over incorrect and more aggressive results. I.e., this prevents the miscompiles at the expense of handling less cases, i.e. making interchange more pessimistic. However, some of the cases that are now rejected for dependence analysis reasons, were rejected before too but for other reasons (e.g. profitability). So at least for the llvm regression tests, the number of regression are very reasonable. This should be a stopgap. We would like to get interchange enabled by default and thus prefer correctness over unsafe transforms, and later see if we can get solve the regressions.
162 lines
4.4 KiB
LLVM
162 lines
4.4 KiB
LLVM
; Remove 'S' Scalar Dependencies #119345
|
|
; Scalar dependencies are not handled correctly, so they were removed to avoid
|
|
; miscompiles. The loop nest in this test case used to be interchanged, but it's
|
|
; no longer triggering. XFAIL'ing this test to indicate that this test should
|
|
; interchanged if scalar deps are handled correctly.
|
|
;
|
|
; XFAIL: *
|
|
|
|
; RUN: opt < %s -passes=loop-interchange -pass-remarks-output=%t -disable-output
|
|
; RUN: FileCheck -input-file %t %s
|
|
|
|
@f = dso_local local_unnamed_addr global [4 x [9 x i32]] [[9 x i32] [i32 5, i32 3, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0], [9 x i32] zeroinitializer, [9 x i32] zeroinitializer, [9 x i32] zeroinitializer], align 4
|
|
@g = common dso_local local_unnamed_addr global i32 0, align 4
|
|
|
|
; int32_t f[3][3];
|
|
; int32_t g;
|
|
;
|
|
; int32_t test1(_Bool cond) {
|
|
; for (int64_t i = 0; i < 3; i++) {
|
|
; for (int64_t j = 0; j < 3; j++) {
|
|
; int32_t val = f[i][j];
|
|
; if (val == 0)
|
|
; if (!cond)
|
|
; g++;
|
|
; else
|
|
; g = 3;
|
|
; if (!cond)
|
|
; g++;
|
|
; }
|
|
; }
|
|
; return g;
|
|
; }
|
|
;
|
|
define dso_local i32 @test1(i1 %cond) {
|
|
;
|
|
; FIXME: if there's an output dependency inside the loop and Src doesn't
|
|
; dominate Dst, we should not interchange. Thus, this currently miscompiles.
|
|
;
|
|
; CHECK: --- !Passed
|
|
; CHECK-NEXT: Pass: loop-interchange
|
|
; CHECK-NEXT: Name: Interchanged
|
|
; CHECK-NEXT: Function: test1
|
|
; CHECK-NEXT: Args:
|
|
; CHECK-NEXT: - String: Loop interchanged with enclosing loop.
|
|
; CHECK-NEXT: ...
|
|
;
|
|
for.preheader:
|
|
br label %outerloop
|
|
|
|
outerloop:
|
|
%i = phi i64 [ 0, %for.preheader ], [ %indvars.iv.next21.i, %for.latch ]
|
|
br label %innerloop
|
|
|
|
innerloop:
|
|
%j = phi i64 [ 0, %outerloop ], [ %j.next, %if.end ]
|
|
%arrayidx6.i = getelementptr inbounds [4 x [9 x i32]], [4 x [9 x i32]]* @f, i64 0, i64 %j, i64 %i
|
|
%i1 = load i32, i32* %arrayidx6.i, align 4
|
|
%tobool.i = icmp eq i32 %i1, 0
|
|
br i1 %tobool.i, label %land.end, label %land.rhs
|
|
|
|
land.rhs:
|
|
store i32 3, i32* @g, align 4
|
|
br label %land.end
|
|
|
|
land.end:
|
|
br i1 %cond, label %if.end, label %if.then
|
|
|
|
if.then:
|
|
%i2 = load i32, i32* @g, align 4
|
|
%inc.i = add i32 %i2, 1
|
|
store i32 %inc.i, i32* @g, align 4
|
|
br label %if.end
|
|
|
|
if.end:
|
|
%j.next = add nuw nsw i64 %j, 1
|
|
%exitcond.i = icmp eq i64 %j.next, 3
|
|
br i1 %exitcond.i, label %for.latch, label %innerloop
|
|
|
|
for.latch:
|
|
%indvars.iv.next21.i = add nsw i64 %i, 1
|
|
%cmp.i = icmp slt i64 %i, 2
|
|
br i1 %cmp.i, label %outerloop, label %exit
|
|
|
|
exit:
|
|
%i3 = load i32, i32* @g, align 4
|
|
ret i32 %i3
|
|
}
|
|
|
|
; int32_t f[3][3];
|
|
; int32_t g;
|
|
;
|
|
; int32_t test2(_Bool cond) {
|
|
; for (int64_t i = 0; i < 3; i++) {
|
|
; for (int64_t j = 0; j < 3; j++) {
|
|
; int32_t val = f[i][j];
|
|
; g = 3;
|
|
; if (val == 0)
|
|
; if (!cond)
|
|
; g++;
|
|
; else
|
|
; if (!cond)
|
|
; g++;
|
|
; }
|
|
; }
|
|
; return g;
|
|
; }
|
|
;
|
|
define dso_local i32 @test2(i1 %cond) {
|
|
;
|
|
; FIXME: if there's an output dependency inside the loop and Src doesn't
|
|
; dominate Dst, we should not interchange. Thus, this currently miscompiles.
|
|
;
|
|
; CHECK: --- !Passed
|
|
; CHECK-NEXT: Pass: loop-interchange
|
|
; CHECK-NEXT: Name: Interchanged
|
|
; CHECK-NEXT: Function: test2
|
|
; CHECK-NEXT: Args:
|
|
; CHECK-NEXT: - String: Loop interchanged with enclosing loop.
|
|
; CHECK-NEXT: ...
|
|
;
|
|
for.preheader:
|
|
br label %outerloop
|
|
|
|
outerloop:
|
|
%i = phi i64 [ 0, %for.preheader ], [ %indvars.iv.next21.i, %for.latch ]
|
|
br label %innerloop
|
|
|
|
innerloop:
|
|
%j = phi i64 [ 0, %outerloop ], [ %j.next, %if.end ]
|
|
%arrayidx6.i = getelementptr inbounds [4 x [9 x i32]], [4 x [9 x i32]]* @f, i64 0, i64 %j, i64 %i
|
|
%i1 = load i32, i32* %arrayidx6.i, align 4
|
|
%tobool.i = icmp eq i32 %i1, 0
|
|
store i32 3, i32* @g, align 4
|
|
br i1 %tobool.i, label %land.end, label %land.rhs
|
|
|
|
land.rhs:
|
|
br label %land.end
|
|
|
|
land.end:
|
|
br i1 %cond, label %if.end, label %if.then
|
|
|
|
if.then:
|
|
%i2 = load i32, i32* @g, align 4
|
|
%inc.i = add i32 %i2, 1
|
|
store i32 %inc.i, i32* @g, align 4
|
|
br label %if.end
|
|
|
|
if.end:
|
|
%j.next = add nuw nsw i64 %j, 1
|
|
%exitcond.i = icmp eq i64 %j.next, 3
|
|
br i1 %exitcond.i, label %for.latch, label %innerloop
|
|
|
|
for.latch:
|
|
%indvars.iv.next21.i = add nsw i64 %i, 1
|
|
%cmp.i = icmp slt i64 %i, 2
|
|
br i1 %cmp.i, label %outerloop, label %exit
|
|
|
|
exit:
|
|
%i3 = load i32, i32* @g, align 4
|
|
ret i32 %i3
|
|
}
|