6e2cea6f0c
Code in CGCall.cpp that loads up function arguments that need to be coerced to a different type may in some cases ignore the fact that the source of the argument is not naturally aligned. This may cause incorrect code to be generated. In some places in CreateCoercedLoad, we already have setAlignment calls to address this, but I ran into one where it was missing, causing wrong code generation on SystemZ. However, in that location, we do not actually know what alignment of the source location we can rely on; the callers do not pass anything to this routine. This is already an issue in other places in CreateCoercedLoad; and the same problem exists for CreateCoercedStore. To avoid pessimising code, and to fix the FIXMEs already in place, this patch also adds an alignment argument to the CreateCoerced* routines and uses it instead of forcing an alignment of 1. The callers are changed to pass in the best information they have. This actually requires changes in a number of existing test cases since we now get better alignment in many places. Differential Revision: http://reviews.llvm.org/D11033 llvm-svn: 241898
50 lines
2.5 KiB
C
50 lines
2.5 KiB
C
// REQUIRES: powerpc-registered-target
|
|
// RUN: %clang_cc1 -triple powerpc64-unknown-linux-gnu -emit-llvm -o - %s | FileCheck %s
|
|
|
|
typedef struct s1 { float f; } Sf;
|
|
typedef struct s2 { double d; } Sd;
|
|
typedef struct s4 { Sf fs; } SSf;
|
|
typedef struct s5 { Sd ds; } SSd;
|
|
|
|
void bar(Sf a, Sd b, SSf d, SSd e) {}
|
|
|
|
// CHECK-LABEL: define void @bar
|
|
// CHECK: %a = alloca %struct.s1, align 4
|
|
// CHECK: %b = alloca %struct.s2, align 8
|
|
// CHECK: %d = alloca %struct.s4, align 4
|
|
// CHECK: %e = alloca %struct.s5, align 8
|
|
// CHECK: %{{[a-zA-Z0-9.]+}} = getelementptr %struct.s1, %struct.s1* %a, i32 0, i32 0
|
|
// CHECK: store float %a.coerce, float* %{{[a-zA-Z0-9.]+}}, align 4
|
|
// CHECK: %{{[a-zA-Z0-9.]+}} = getelementptr %struct.s2, %struct.s2* %b, i32 0, i32 0
|
|
// CHECK: store double %b.coerce, double* %{{[a-zA-Z0-9.]+}}, align 8
|
|
// CHECK: %{{[a-zA-Z0-9.]+}} = getelementptr %struct.s4, %struct.s4* %d, i32 0, i32 0
|
|
// CHECK: %{{[a-zA-Z0-9.]+}} = getelementptr %struct.s1, %struct.s1* %{{[a-zA-Z0-9.]+}}, i32 0, i32 0
|
|
// CHECK: store float %d.coerce, float* %{{[a-zA-Z0-9.]+}}, align 4
|
|
// CHECK: %{{[a-zA-Z0-9.]+}} = getelementptr %struct.s5, %struct.s5* %e, i32 0, i32 0
|
|
// CHECK: %{{[a-zA-Z0-9.]+}} = getelementptr %struct.s2, %struct.s2* %{{[a-zA-Z0-9.]+}}, i32 0, i32 0
|
|
// CHECK: store double %e.coerce, double* %{{[a-zA-Z0-9.]+}}, align 8
|
|
// CHECK: ret void
|
|
|
|
void foo(void)
|
|
{
|
|
Sf p1 = { 22.63f };
|
|
Sd p2 = { 19.47 };
|
|
SSf p4 = { { 22.63f } };
|
|
SSd p5 = { { 19.47 } };
|
|
bar(p1, p2, p4, p5);
|
|
}
|
|
|
|
// CHECK-LABEL: define void @foo
|
|
// CHECK: %{{[a-zA-Z0-9.]+}} = getelementptr %struct.s1, %struct.s1* %p1, i32 0, i32 0
|
|
// CHECK: %{{[0-9]+}} = load float, float* %{{[a-zA-Z0-9.]+}}, align 4
|
|
// CHECK: %{{[a-zA-Z0-9.]+}} = getelementptr %struct.s2, %struct.s2* %p2, i32 0, i32 0
|
|
// CHECK: %{{[0-9]+}} = load double, double* %{{[a-zA-Z0-9.]+}}, align 8
|
|
// CHECK: %{{[a-zA-Z0-9.]+}} = getelementptr %struct.s4, %struct.s4* %p4, i32 0, i32 0
|
|
// CHECK: %{{[a-zA-Z0-9.]+}} = getelementptr %struct.s1, %struct.s1* %{{[a-zA-Z0-9.]+}}, i32 0, i32 0
|
|
// CHECK: %{{[0-9]+}} = load float, float* %{{[a-zA-Z0-9.]+}}, align 4
|
|
// CHECK: %{{[a-zA-Z0-9.]+}} = getelementptr %struct.s5, %struct.s5* %p5, i32 0, i32 0
|
|
// CHECK: %{{[a-zA-Z0-9.]+}} = getelementptr %struct.s2, %struct.s2* %{{[a-zA-Z0-9.]+}}, i32 0, i32 0
|
|
// CHECK: %{{[0-9]+}} = load double, double* %{{[a-zA-Z0-9.]+}}, align 8
|
|
// CHECK: call void @bar(float inreg %{{[0-9]+}}, double inreg %{{[0-9]+}}, float inreg %{{[0-9]+}}, double inreg %{{[0-9]+}})
|
|
// CHECK: ret void
|