mirrors/llvm-project
0
0
Fork 0
mirror of https://github.com/llvm/llvm-project.git synced 2025-04-25 23:36:06 +00:00
Code Issues Projects Releases Wiki Activity
llvm-project/clang/test/OpenMP/tile_codegen_for_dependent.cpp
Michael Kruse 9120562dfc
[Clang][OpenMP] Enable tile/unroll on iterator- and foreach-loops (#91459) 
OpenMP loop transformation did not work on a for-loop using an iterator
or range-based for-loops. The first reason is that it combined the
iterator's type for generated loops with the type of `NumIterations` as
generated for any `OMPLoopBasedDirective` which is an integer. Fixed by
basing all generated loop variables on `NumIterations`.

Second, C++11 range-based for-loops include syntactic sugar that needs
to be executed before the loop. This additional code is now added to the
construct's Pre-Init lists.

Third, C++20 added an initializer statement to range-based for-loops
which is also added to the pre-init statement. PreInits used to be a
`DeclStmt` which made it difficult to add arbitrary statements from
`CXXRangeForStmt`'s syntactic sugar, especially the for-loops init
statement which does not need to be a declaration. Change it to be a
general `Stmt` that can be a `CompoundStmt` to hold arbitrary Stmts,
including DeclStmts. This also avoids the `PointerUnion` workaround used
by `checkTransformableLoopNest`.

End-to-end tests are added to verify the expected number and order of
loop execution and evaluations of expressions (such as iterator
dereference). The order and number of evaluations of expressions in
canonical loops is explicitly undefined by OpenMP but checked here for
clarification and for changes to be noticed.
2024-05-22 14:30:31 +02:00

194 lines
9.9 KiB
C++
Raw Blame History

// Check code generation
// RUN: %clang_cc1 -verify -triple x86_64-pc-linux-gnu -fopenmp -emit-llvm %s -o - | FileCheck %s --check-prefix=IR
// Check same results after serialization round-trip
// RUN: %clang_cc1 -verify -triple x86_64-pc-linux-gnu -fopenmp -emit-pch -o %t %s
// RUN: %clang_cc1 -verify -triple x86_64-pc-linux-gnu -fopenmp -include-pch %t -emit-llvm %s -o - | FileCheck %s --check-prefix=IR
// expected-no-diagnostics
// The loop trip count used by #pragma omp for depends on code generated
// by #pragma omp file. Check that theses PreInits are emitted before
// the code generated by #pragma omp for.
#ifndef HEADER
#define HEADER
// placeholder for loop body code.
extern "C" void body(...) {}
// IR-LABEL: define {{.*}}@func(
// IR-NEXT: [[ENTRY:.*]]:
// IR-NEXT: %[[START_ADDR:.+]] = alloca i32, align 4
// IR-NEXT: %[[END_ADDR:.+]] = alloca i32, align 4
// IR-NEXT: %[[STEP_ADDR:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTOMP_IV:.+]] = alloca i32, align 4
// IR-NEXT: %[[TMP:.+]] = alloca i32, align 4
// IR-NEXT: %[[I:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTCAPTURE_EXPR_:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTCAPTURE_EXPR_1:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTNEW_STEP:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTCAPTURE_EXPR_2:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTCAPTURE_EXPR_5:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTCAPTURE_EXPR_7:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTFLOOR_0_IV_I:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTOMP_LB:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTOMP_UB:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTOMP_STRIDE:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTOMP_IS_LAST:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTFLOOR_0_IV_I11:.+]] = alloca i32, align 4
// IR-NEXT: %[[DOTTILE_0_IV_I:.+]] = alloca i32, align 4
// IR-NEXT: %[[TMP0:.+]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB2:.+]])
// IR-NEXT: store i32 %[[START:.+]], ptr %[[START_ADDR]], align 4
// IR-NEXT: store i32 %[[END:.+]], ptr %[[END_ADDR]], align 4
// IR-NEXT: store i32 %[[STEP:.+]], ptr %[[STEP_ADDR]], align 4
// IR-NEXT: %[[TMP1:.+]] = load i32, ptr %[[START_ADDR]], align 4
// IR-NEXT: store i32 %[[TMP1]], ptr %[[I]], align 4
// IR-NEXT: %[[TMP2:.+]] = load i32, ptr %[[START_ADDR]], align 4
// IR-NEXT: store i32 %[[TMP2]], ptr %[[DOTCAPTURE_EXPR_]], align 4
// IR-NEXT: %[[TMP3:.+]] = load i32, ptr %[[END_ADDR]], align 4
// IR-NEXT: store i32 %[[TMP3]], ptr %[[DOTCAPTURE_EXPR_1]], align 4
// IR-NEXT: %[[TMP4:.+]] = load i32, ptr %[[STEP_ADDR]], align 4
// IR-NEXT: store i32 %[[TMP4]], ptr %[[DOTNEW_STEP]], align 4
// IR-NEXT: %[[TMP5:.+]] = load i32, ptr %[[DOTCAPTURE_EXPR_1]], align 4
// IR-NEXT: %[[TMP6:.+]] = load i32, ptr %[[DOTCAPTURE_EXPR_]], align 4
// IR-NEXT: %[[SUB:.+]] = sub i32 %[[TMP5]], %[[TMP6]]
// IR-NEXT: %[[SUB3:.+]] = sub i32 %[[SUB]], 1
// IR-NEXT: %[[TMP7:.+]] = load i32, ptr %[[DOTNEW_STEP]], align 4
// IR-NEXT: %[[ADD:.+]] = add i32 %[[SUB3]], %[[TMP7]]
// IR-NEXT: %[[TMP8:.+]] = load i32, ptr %[[DOTNEW_STEP]], align 4
// IR-NEXT: %[[DIV:.+]] = udiv i32 %[[ADD]], %[[TMP8]]
// IR-NEXT: %[[SUB4:.+]] = sub i32 %[[DIV]], 1
// IR-NEXT: store i32 %[[SUB4]], ptr %[[DOTCAPTURE_EXPR_2]], align 4
// IR-NEXT: %[[TMP9:.+]] = load i32, ptr %[[DOTCAPTURE_EXPR_2]], align 4
// IR-NEXT: %[[ADD6:.+]] = add i32 %[[TMP9]], 1
// IR-NEXT: store i32 %[[ADD6]], ptr %[[DOTCAPTURE_EXPR_5]], align 4
// IR-NEXT: %[[TMP10:.+]] = load i32, ptr %[[DOTCAPTURE_EXPR_5]], align 4
// IR-NEXT: %[[SUB8:.+]] = sub i32 %[[TMP10]], -3
// IR-NEXT: %[[DIV9:.+]] = udiv i32 %[[SUB8]], 4
// IR-NEXT: %[[SUB10:.+]] = sub i32 %[[DIV9]], 1
// IR-NEXT: store i32 %[[SUB10]], ptr %[[DOTCAPTURE_EXPR_7]], align 4
// IR-NEXT: store i32 0, ptr %[[DOTFLOOR_0_IV_I]], align 4
// IR-NEXT: %[[TMP11:.+]] = load i32, ptr %[[DOTCAPTURE_EXPR_5]], align 4
// IR-NEXT: %[[CMP:.+]] = icmp ult i32 0, %[[TMP11]]
// IR-NEXT: br i1 %[[CMP]], label %[[OMP_PRECOND_THEN:.+]], label %[[OMP_PRECOND_END:.+]]
// IR-EMPTY:
// IR-NEXT: [[OMP_PRECOND_THEN]]:
// IR-NEXT: store i32 0, ptr %[[DOTOMP_LB]], align 4
// IR-NEXT: %[[TMP12:.+]] = load i32, ptr %[[DOTCAPTURE_EXPR_7]], align 4
// IR-NEXT: store i32 %[[TMP12]], ptr %[[DOTOMP_UB]], align 4
// IR-NEXT: store i32 1, ptr %[[DOTOMP_STRIDE]], align 4
// IR-NEXT: store i32 0, ptr %[[DOTOMP_IS_LAST]], align 4
// IR-NEXT: call void @__kmpc_for_static_init_4u(ptr @[[GLOB1:.+]], i32 %[[TMP0]], i32 34, ptr %[[DOTOMP_IS_LAST]], ptr %[[DOTOMP_LB]], ptr %[[DOTOMP_UB]], ptr %[[DOTOMP_STRIDE]], i32 1, i32 1)
// IR-NEXT: %[[TMP13:.+]] = load i32, ptr %[[DOTOMP_UB]], align 4
// IR-NEXT: %[[TMP14:.+]] = load i32, ptr %[[DOTCAPTURE_EXPR_7]], align 4
// IR-NEXT: %[[CMP12:.+]] = icmp ugt i32 %[[TMP13]], %[[TMP14]]
// IR-NEXT: br i1 %[[CMP12]], label %[[COND_TRUE:.+]], label %[[COND_FALSE:.+]]
// IR-EMPTY:
// IR-NEXT: [[COND_TRUE]]:
// IR-NEXT: %[[TMP15:.+]] = load i32, ptr %[[DOTCAPTURE_EXPR_7]], align 4
// IR-NEXT: br label %[[COND_END:.+]]
// IR-EMPTY:
// IR-NEXT: [[COND_FALSE]]:
// IR-NEXT: %[[TMP16:.+]] = load i32, ptr %[[DOTOMP_UB]], align 4
// IR-NEXT: br label %[[COND_END]]
// IR-EMPTY:
// IR-NEXT: [[COND_END]]:
// IR-NEXT: %[[COND:.+]] = phi i32 [ %[[TMP15]], %[[COND_TRUE]] ], [ %[[TMP16]], %[[COND_FALSE]] ]
// IR-NEXT: store i32 %[[COND]], ptr %[[DOTOMP_UB]], align 4
// IR-NEXT: %[[TMP17:.+]] = load i32, ptr %[[DOTOMP_LB]], align 4
// IR-NEXT: store i32 %[[TMP17]], ptr %[[DOTOMP_IV]], align 4
// IR-NEXT: br label %[[OMP_INNER_FOR_COND:.+]]
// IR-EMPTY:
// IR-NEXT: [[OMP_INNER_FOR_COND]]:
// IR-NEXT: %[[TMP18:.+]] = load i32, ptr %[[DOTOMP_IV]], align 4
// IR-NEXT: %[[TMP19:.+]] = load i32, ptr %[[DOTOMP_UB]], align 4
// IR-NEXT: %[[ADD13:.+]] = add i32 %[[TMP19]], 1
// IR-NEXT: %[[CMP14:.+]] = icmp ult i32 %[[TMP18]], %[[ADD13]]
// IR-NEXT: br i1 %[[CMP14]], label %[[OMP_INNER_FOR_BODY:.+]], label %[[OMP_INNER_FOR_END:.+]]
// IR-EMPTY:
// IR-NEXT: [[OMP_INNER_FOR_BODY]]:
// IR-NEXT: %[[TMP20:.+]] = load i32, ptr %[[DOTOMP_IV]], align 4
// IR-NEXT: %[[MUL:.+]] = mul i32 %[[TMP20]], 4
// IR-NEXT: %[[ADD15:.+]] = add i32 0, %[[MUL]]
// IR-NEXT: store i32 %[[ADD15]], ptr %[[DOTFLOOR_0_IV_I11]], align 4
// IR-NEXT: %[[TMP21:.+]] = load i32, ptr %[[DOTFLOOR_0_IV_I11]], align 4
// IR-NEXT: store i32 %[[TMP21]], ptr %[[DOTTILE_0_IV_I]], align 4
// IR-NEXT: br label %[[FOR_COND:.+]]
// IR-EMPTY:
// IR-NEXT: [[FOR_COND]]:
// IR-NEXT: %[[TMP22:.+]] = load i32, ptr %[[DOTTILE_0_IV_I]], align 4
// IR-NEXT: %[[TMP23:.+]] = load i32, ptr %[[DOTCAPTURE_EXPR_2]], align 4
// IR-NEXT: %[[ADD16:.+]] = add i32 %[[TMP23]], 1
// IR-NEXT: %[[TMP24:.+]] = load i32, ptr %[[DOTFLOOR_0_IV_I11]], align 4
// IR-NEXT: %[[ADD17:.+]] = add i32 %[[TMP24]], 4
// IR-NEXT: %[[CMP18:.+]] = icmp ult i32 %[[ADD16]], %[[ADD17]]
// IR-NEXT: br i1 %[[CMP18]], label %[[COND_TRUE19:.+]], label %[[COND_FALSE21:.+]]
// IR-EMPTY:
// IR-NEXT: [[COND_TRUE19]]:
// IR-NEXT: %[[TMP25:.+]] = load i32, ptr %[[DOTCAPTURE_EXPR_2]], align 4
// IR-NEXT: %[[ADD20:.+]] = add i32 %[[TMP25]], 1
// IR-NEXT: br label %[[COND_END23:.+]]
// IR-EMPTY:
// IR-NEXT: [[COND_FALSE21]]:
// IR-NEXT: %[[TMP26:.+]] = load i32, ptr %[[DOTFLOOR_0_IV_I11]], align 4
// IR-NEXT: %[[ADD22:.+]] = add i32 %[[TMP26]], 4
// IR-NEXT: br label %[[COND_END23]]
// IR-EMPTY:
// IR-NEXT: [[COND_END23]]:
// IR-NEXT: %[[COND24:.+]] = phi i32 [ %[[ADD20]], %[[COND_TRUE19]] ], [ %[[ADD22]], %[[COND_FALSE21]] ]
// IR-NEXT: %[[CMP25:.+]] = icmp ult i32 %[[TMP22]], %[[COND24]]
// IR-NEXT: br i1 %[[CMP25]], label %[[FOR_BODY:.+]], label %[[FOR_END:.+]]
// IR-EMPTY:
// IR-NEXT: [[FOR_BODY]]:
// IR-NEXT: %[[TMP27:.+]] = load i32, ptr %[[DOTCAPTURE_EXPR_]], align 4
// IR-NEXT: %[[TMP28:.+]] = load i32, ptr %[[DOTTILE_0_IV_I]], align 4
// IR-NEXT: %[[TMP29:.+]] = load i32, ptr %[[DOTNEW_STEP]], align 4
// IR-NEXT: %[[MUL26:.+]] = mul i32 %[[TMP28]], %[[TMP29]]
// IR-NEXT: %[[ADD27:.+]] = add i32 %[[TMP27]], %[[MUL26]]
// IR-NEXT: store i32 %[[ADD27]], ptr %[[I]], align 4
// IR-NEXT: %[[TMP30:.+]] = load i32, ptr %[[START_ADDR]], align 4
// IR-NEXT: %[[TMP31:.+]] = load i32, ptr %[[END_ADDR]], align 4
// IR-NEXT: %[[TMP32:.+]] = load i32, ptr %[[STEP_ADDR]], align 4
// IR-NEXT: %[[TMP33:.+]] = load i32, ptr %[[I]], align 4
// IR-NEXT: call void (...) @body(i32 noundef %[[TMP30]], i32 noundef %[[TMP31]], i32 noundef %[[TMP32]], i32 noundef %[[TMP33]])
// IR-NEXT: br label %[[FOR_INC:.+]]
// IR-EMPTY:
// IR-NEXT: [[FOR_INC]]:
// IR-NEXT: %[[TMP34:.+]] = load i32, ptr %[[DOTTILE_0_IV_I]], align 4
// IR-NEXT: %[[INC:.+]] = add i32 %[[TMP34]], 1
// IR-NEXT: store i32 %[[INC]], ptr %[[DOTTILE_0_IV_I]], align 4
// IR-NEXT: br label %[[FOR_COND]], !llvm.loop ![[LOOP3:[0-9]+]]
// IR-EMPTY:
// IR-NEXT: [[FOR_END]]:
// IR-NEXT: br label %[[OMP_BODY_CONTINUE:.+]]
// IR-EMPTY:
// IR-NEXT: [[OMP_BODY_CONTINUE]]:
// IR-NEXT: br label %[[OMP_INNER_FOR_INC:.+]]
// IR-EMPTY:
// IR-NEXT: [[OMP_INNER_FOR_INC]]:
// IR-NEXT: %[[TMP35:.+]] = load i32, ptr %[[DOTOMP_IV]], align 4
// IR-NEXT: %[[ADD28:.+]] = add i32 %[[TMP35]], 1
// IR-NEXT: store i32 %[[ADD28]], ptr %[[DOTOMP_IV]], align 4
// IR-NEXT: br label %[[OMP_INNER_FOR_COND]]
// IR-EMPTY:
// IR-NEXT: [[OMP_INNER_FOR_END]]:
// IR-NEXT: br label %[[OMP_LOOP_EXIT:.+]]
// IR-EMPTY:
// IR-NEXT: [[OMP_LOOP_EXIT]]:
// IR-NEXT: call void @__kmpc_for_static_fini(ptr @[[GLOB1]], i32 %[[TMP0]])
// IR-NEXT: br label %[[OMP_PRECOND_END]]
// IR-EMPTY:
// IR-NEXT: [[OMP_PRECOND_END]]:
// IR-NEXT: call void @__kmpc_barrier(ptr @[[GLOB3:.+]], i32 %[[TMP0]])
// IR-NEXT: ret void
// IR-NEXT: }
extern "C" void func(int start, int end, int step) {
#pragma omp for
#pragma omp tile sizes(4)
for (int i = start; i < end; i += step)
body(start, end, step, i);
}
#endif /* HEADER */
Reference in New Issue View Git Blame Copy Permalink