7a0f7dbf2d
Reapply of a22d1c2225. Using this PR for
pre-merge CI.
Instead of relying on any pass manager to schedule Polly's passes, add
Polly's own pipeline manager which is seen as a monolithic pass in
LLVM's pass manager. Polly's former passes are now phases of the new
PhaseManager component.
Relying on LLVM's pass manager (the legacy as well as the New Pass
Manager) to manage Polly's phases never was a good fit that the
PhaseManager resolves:
* Polly passes were modifying analysis results, in particular RegionInfo
and ScopInfo. This means that there was not just one unique and
"definite" analysis result, the actual result depended on which analyses
ran prior, and the pass manager was not allowed to throw away cached
analyses or prior SCoP optimizations would have been forgotten. The LLVM
pass manger's persistance of analysis results is not contractual but
designed for caching.
* Polly depends on a particular execution order of passes and regions
(e.g. regression tests, invalidation of consecutive SCoPs). LLVM's pass
manager does not guarantee any excecution order.
* Polly does not completely preserve DominatorTree, RegionInfo,
LoopInfo, or ScalarEvolution, but only as-needed for Polly's own uses.
Because the ScopDetection object stores references to those analyses, it
still had to lie to the pass manager that they would be preserved, or
the pass manager would have released and recomputed the invalidated
analysis objects that ScopDetection/ScopInfo was still referencing. To
ensure that no non-Polly pass would see these not-completely-preserved
analyses, all analyses still had to be thrown away after the
ScopPassManager, respectively with a BarrierNoopPass in case of the LPM.
* The NPM's PassInstrumentation wraps the IR unit into an `llvm::Any`
object, but implementations such as PrintIRInstrumentation call
llvm_unreachable on encountering an unknown IR unit, such as SCoPs, with
no extension points to add support. Hence LLVM crashes when dumping IR
between SCoP passes (such as `-print-before-changed` with Polly being
active).
The new PhaseManager uses some command line options that previously
belonged to Polly's legacy passes, such as `-polly-print-detect` (so the
option will continue to work). Hence the LPM support is incompatible
with the new approach and support for it is removed.
138 lines
3.9 KiB
LLVM
138 lines
3.9 KiB
LLVM
; RUN: opt %loadNPMPolly '-passes=polly-custom<ast>' -polly-print-ast -disable-output < %s | FileCheck %s
|
|
|
|
;#include <string.h>
|
|
;#define N 1024
|
|
;
|
|
;int A[N];
|
|
;
|
|
;void sequential_loops() {
|
|
; int i;
|
|
; for (i = 0; i < N/2; i++) {
|
|
; A[i] = 1;
|
|
; }
|
|
; for (i = N/2 ; i < N; i++) {
|
|
; A[i] = 2;
|
|
; }
|
|
;}
|
|
;
|
|
;int main () {
|
|
; int i;
|
|
; memset(A, 0, sizeof(int) * N);
|
|
;
|
|
; sequential_loops();
|
|
;
|
|
; for (i = 0; i < N; i++) {
|
|
; if (A[i] != 1 && i < N/2)
|
|
; return 1;
|
|
; if (A[i] != 2 && i >= N/2)
|
|
; return 1;
|
|
; }
|
|
;
|
|
; return 0;
|
|
;}
|
|
|
|
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"
|
|
|
|
@A = common global [1024 x i32] zeroinitializer, align 4 ; <ptr> [#uses=5]
|
|
|
|
define void @sequential_loops() nounwind {
|
|
bb:
|
|
br label %bb1
|
|
|
|
bb1: ; preds = %bb3, %bb
|
|
%indvar1 = phi i64 [ %indvar.next2, %bb3 ], [ 0, %bb ]
|
|
%scevgep4 = getelementptr [1024 x i32], ptr @A, i64 0, i64 %indvar1
|
|
%exitcond3 = icmp ne i64 %indvar1, 512
|
|
br i1 %exitcond3, label %bb2, label %bb4
|
|
|
|
bb2: ; preds = %bb1
|
|
store i32 1, ptr %scevgep4
|
|
br label %bb3
|
|
|
|
bb3: ; preds = %bb2
|
|
%indvar.next2 = add i64 %indvar1, 1
|
|
br label %bb1
|
|
|
|
bb4: ; preds = %bb1
|
|
br label %bb5
|
|
|
|
bb5: ; preds = %bb7, %bb4
|
|
%indvar = phi i64 [ %indvar.next, %bb7 ], [ 0, %bb4 ]
|
|
%tmp = add i64 %indvar, 512
|
|
%scevgep = getelementptr [1024 x i32], ptr @A, i64 0, i64 %tmp
|
|
%exitcond = icmp ne i64 %indvar, 512
|
|
br i1 %exitcond, label %bb6, label %bb8
|
|
|
|
bb6: ; preds = %bb5
|
|
store i32 2, ptr %scevgep
|
|
br label %bb7
|
|
|
|
bb7: ; preds = %bb6
|
|
%indvar.next = add i64 %indvar, 1
|
|
br label %bb5
|
|
|
|
bb8: ; preds = %bb5
|
|
ret void
|
|
}
|
|
|
|
define i32 @main() nounwind {
|
|
bb:
|
|
call void @llvm.memset.p0.i64(ptr @A, i8 0, i64 4096, i32 1, i1 false)
|
|
call void @sequential_loops()
|
|
br label %bb1
|
|
|
|
bb1: ; preds = %bb15, %bb
|
|
%indvar = phi i64 [ %indvar.next, %bb15 ], [ 0, %bb ]
|
|
%i.0 = trunc i64 %indvar to i32
|
|
%scevgep = getelementptr [1024 x i32], ptr @A, i64 0, i64 %indvar
|
|
%tmp = icmp slt i32 %i.0, 1024
|
|
br i1 %tmp, label %bb2, label %bb16
|
|
|
|
bb2: ; preds = %bb1
|
|
%tmp3 = load i32, ptr %scevgep
|
|
%tmp4 = icmp ne i32 %tmp3, 1
|
|
br i1 %tmp4, label %bb5, label %bb8
|
|
|
|
bb5: ; preds = %bb2
|
|
%tmp6 = icmp slt i32 %i.0, 512
|
|
br i1 %tmp6, label %bb7, label %bb8
|
|
|
|
bb7: ; preds = %bb5
|
|
br label %bb17
|
|
|
|
bb8: ; preds = %bb5, %bb2
|
|
%tmp9 = load i32, ptr %scevgep
|
|
%tmp10 = icmp ne i32 %tmp9, 2
|
|
br i1 %tmp10, label %bb11, label %bb14
|
|
|
|
bb11: ; preds = %bb8
|
|
%tmp12 = icmp sge i32 %i.0, 512
|
|
br i1 %tmp12, label %bb13, label %bb14
|
|
|
|
bb13: ; preds = %bb11
|
|
br label %bb17
|
|
|
|
bb14: ; preds = %bb11, %bb8
|
|
br label %bb15
|
|
|
|
bb15: ; preds = %bb14
|
|
%indvar.next = add i64 %indvar, 1
|
|
br label %bb1
|
|
|
|
bb16: ; preds = %bb1
|
|
br label %bb17
|
|
|
|
bb17: ; preds = %bb16, %bb13, %bb7
|
|
%.0 = phi i32 [ 1, %bb7 ], [ 1, %bb13 ], [ 0, %bb16 ]
|
|
ret i32 %.0
|
|
}
|
|
|
|
declare void @llvm.memset.p0.i64(ptr nocapture, i8, i64, i32, i1) nounwind
|
|
|
|
; CHECK: {
|
|
; CHECK: for (int c0 = 0; c0 <= 511; c0 += 1)
|
|
; CHECK: Stmt_bb2(c0);
|
|
; CHECK: for (int c0 = 0; c0 <= 511; c0 += 1)
|
|
; CHECK: Stmt_bb6(c0);
|
|
; CHECK: }
|