maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/clang/test/CodeGen/aarch64-fix-cortex-a53-8357...

28 lines
1.1 KiB
C
Raw Normal View History

[AArch64] Add workaround for Cortex-A53 erratum (835769) Some early revisions of the Cortex-A53 have an erratum (835769) whereby it is possible for a 64-bit multiply-accumulate instruction in AArch64 state to generate an incorrect result. The details are quite complex and hard to determine statically, since branches in the code may exist in some circumstances, but all cases end with a memory (load, store, or prefetch) instruction followed immediately by the multiply-accumulate operation. The safest work-around for this issue is to make the compiler avoid emitting multiply-accumulate instructions immediately after memory instructions and the simplest way to do this is to insert a NOP. This patch implements clang options to enable this workaround in the backend. The work-around code generation is not enabled by default. llvm-svn: 219604
2014-10-13 18:16:06 +08:00
// REQUIRES: aarch64-registered-target
[AArch64] Enable A53 erratum workaround (835769) by default for Android targets llvm-svn: 219933
2014-10-17 00:35:14 +08:00
[AArch64] Add workaround for Cortex-A53 erratum (835769) Some early revisions of the Cortex-A53 have an erratum (835769) whereby it is possible for a 64-bit multiply-accumulate instruction in AArch64 state to generate an incorrect result. The details are quite complex and hard to determine statically, since branches in the code may exist in some circumstances, but all cases end with a memory (load, store, or prefetch) instruction followed immediately by the multiply-accumulate operation. The safest work-around for this issue is to make the compiler avoid emitting multiply-accumulate instructions immediately after memory instructions and the simplest way to do this is to insert a NOP. This patch implements clang options to enable this workaround in the backend. The work-around code generation is not enabled by default. llvm-svn: 219604
2014-10-13 18:16:06 +08:00
// RUN: %clang -O3 -target aarch64-linux-eabi %s -S -o- \
// RUN: | FileCheck --check-prefix=CHECK-NO --check-prefix=CHECK %s
// RUN: %clang -O3 -target aarch64-linux-eabi -mfix-cortex-a53-835769 %s -S -o- 2>&1 \
// RUN: | FileCheck --check-prefix=CHECK-YES --check-prefix=CHECK %s
// RUN: %clang -O3 -target aarch64-linux-eabi -mno-fix-cortex-a53-835769 %s -S -o- 2>&1 \
// RUN: | FileCheck --check-prefix=CHECK-NO --check-prefix=CHECK %s
[AArch64] Enable A53 erratum workaround (835769) by default for Android targets llvm-svn: 219933
2014-10-17 00:35:14 +08:00
// RUN: %clang -O3 -target aarch64-android-eabi %s -S -o- \
// RUN: | FileCheck --check-prefix=CHECK-YES --check-prefix=CHECK %s
// RUN: %clang -O3 -target aarch64-android-eabi -mfix-cortex-a53-835769 %s -S -o- \
// RUN: | FileCheck --check-prefix=CHECK-YES --check-prefix=CHECK %s
// RUN: %clang -O3 -target aarch64-android-eabi -mno-fix-cortex-a53-835769 %s -S -o- \
// RUN: | FileCheck --check-prefix=CHECK-NO --check-prefix=CHECK %s
[AArch64] Fixup test from A53 erratum patch after buildbot failures Don't include stdint.h directly, instead typedef int64_t which is all we need. llvm-svn: 219608
2014-10-13 19:18:05 +08:00
typedef long int64_t;
[AArch64] Add workaround for Cortex-A53 erratum (835769) Some early revisions of the Cortex-A53 have an erratum (835769) whereby it is possible for a 64-bit multiply-accumulate instruction in AArch64 state to generate an incorrect result. The details are quite complex and hard to determine statically, since branches in the code may exist in some circumstances, but all cases end with a memory (load, store, or prefetch) instruction followed immediately by the multiply-accumulate operation. The safest work-around for this issue is to make the compiler avoid emitting multiply-accumulate instructions immediately after memory instructions and the simplest way to do this is to insert a NOP. This patch implements clang options to enable this workaround in the backend. The work-around code generation is not enabled by default. llvm-svn: 219604
2014-10-13 18:16:06 +08:00
int64_t f_load_madd_64(int64_t a, int64_t b, int64_t *c) {
int64_t result = a+b*(*c);
return result;
}
// CHECK: ldr
// CHECK-YES-NEXT: nop
Fix Clang tests that used CHECK-NEXT-NOT and CHECK-DAG-NOT. FileCheck actually doesn't support combo suffixes. Differential Revision: http://reviews.llvm.org/D17589 llvm-svn: 262052
2016-02-27 03:34:01 +08:00
// CHECK-NO-NOT: nop
[AArch64] Add workaround for Cortex-A53 erratum (835769) Some early revisions of the Cortex-A53 have an erratum (835769) whereby it is possible for a 64-bit multiply-accumulate instruction in AArch64 state to generate an incorrect result. The details are quite complex and hard to determine statically, since branches in the code may exist in some circumstances, but all cases end with a memory (load, store, or prefetch) instruction followed immediately by the multiply-accumulate operation. The safest work-around for this issue is to make the compiler avoid emitting multiply-accumulate instructions immediately after memory instructions and the simplest way to do this is to insert a NOP. This patch implements clang options to enable this workaround in the backend. The work-around code generation is not enabled by default. llvm-svn: 219604
2014-10-13 18:16:06 +08:00
// CHECK-NEXT: madd