zahiraam created this revision.
zahiraam added reviewers: bader, andrew.w.kaylor.
Herald added subscribers: Naghasan, Anastasia, ebevhan, yaxunl.
Herald added a project: All.
zahiraam requested review of this revision.
Herald added a project: clang.
__arithmetic_fence enforces ordering on expression evaluation when fast math is
enabled.
In fast math mode some floating-point optimizations are performed such as
reassociation and distribution.
For example, the compiler may transform (a+b)+c into a+(b+c). Although these
two expressions are
equivalent in integer arithmetic, they may not be in floating-point arithmetic.
The builtin tells the compiler that
the expression in parenthesis can’t be re-associated or distributed.
__arithmetic_fence(a+b)+c is not equivalent to a+(b+c).
This patch adds the support of the builtin to SPIR-V.
Repository:
rG LLVM Github Monorepo
https://reviews.llvm.org/D142583
Files:
clang/lib/Basic/Targets/SPIR.h
clang/test/CodeGen/arithmetic-fence-builtin.c
clang/test/Sema/arithmetic-fence-builtin.c
Index: clang/test/Sema/arithmetic-fence-builtin.c
===================================================================
--- clang/test/Sema/arithmetic-fence-builtin.c
+++ clang/test/Sema/arithmetic-fence-builtin.c
@@ -2,6 +2,8 @@
// RUN: %clang_cc1 -triple ppc64le -DPPC -emit-llvm -o - -verify -x c++ %s
// RUN: not %clang_cc1 -triple ppc64le -DPPC -emit-llvm -o - -x c++ %s \
// RUN: -fprotect-parens 2>&1 | FileCheck -check-prefix=PPC %s
+// RUN: %clang_cc1 -triple spir64 -emit-llvm -fsycl-is-device \
+// RUN: -o - -verify -x c++ %s
#ifndef PPC
int v;
template <typename T> T addT(T a, T b) {
Index: clang/test/CodeGen/arithmetic-fence-builtin.c
===================================================================
--- clang/test/CodeGen/arithmetic-fence-builtin.c
+++ clang/test/CodeGen/arithmetic-fence-builtin.c
@@ -1,17 +1,23 @@
// Test with fast math
// RUN: %clang_cc1 -triple i386-pc-linux-gnu -emit-llvm -DFAST \
-// RUN: -mreassociate \
-// RUN: -o - %s | FileCheck --check-prefixes CHECK,CHECKFAST,CHECKNP %s
+// RUN: -mreassociate -o - %s \
+// RUN: | FileCheck --check-prefixes CHECK,CHECKFAST,CHECKPRECISE,CHECKNP %s
//
// Test with fast math and fprotect-parens
// RUN: %clang_cc1 -triple i386-pc-linux-gnu -emit-llvm -DFAST \
-// RUN: -mreassociate -fprotect-parens -ffp-contract=on\
-// RUN: -o - %s | FileCheck --check-prefixes CHECK,CHECKFAST,CHECKPP %s
+// RUN: -mreassociate -fprotect-parens -ffp-contract=on -o - %s \
+// RUN: | FileCheck --check-prefixes CHECK,CHECKFAST,CHECKPRECISE,CHECKPP %s
//
// Test without fast math: llvm intrinsic not created
// RUN: %clang_cc1 -triple i386-pc-linux-gnu -emit-llvm -fprotect-parens\
// RUN: -o - %s | FileCheck --implicit-check-not="llvm.arithmetic.fence" %s
//
+// Test with fast math on spir target
+// RUN: %clang_cc1 -triple spir64 -emit-llvm -fsycl-is-device \
+// RUN: -mreassociate -o - %s \
+// RUN: | FileCheck --check-prefixes CHECK,CHECKFAST,CHECKNP %s
+//
+
int v;
int addit(float a, float b) {
// CHECK: define {{.*}}@addit(float noundef %a, float noundef %b) #0 {
@@ -65,7 +71,7 @@
#ifdef FAST
#pragma float_control(precise, on)
int subit(float a, float b, float *fp) {
- // CHECKFAST: define {{.*}}@subit(float noundef %a, float noundef %b{{.*}}
+ // CHECKPRECISE: define {{.*}}@subit(float noundef %a, float noundef %b{{.*}}
*fp = __arithmetic_fence(a - b);
*fp = (a + b);
// CHECK-NOT: call{{.*}} float @llvm.arithmetic.fence.f32(float noundef %add)
Index: clang/lib/Basic/Targets/SPIR.h
===================================================================
--- clang/lib/Basic/Targets/SPIR.h
+++ clang/lib/Basic/Targets/SPIR.h
@@ -191,6 +191,8 @@
bool hasFeature(StringRef Feature) const override {
return Feature == "spir";
}
+
+ bool checkArithmeticFenceSupported() const override { return true; }
};
class LLVM_LIBRARY_VISIBILITY SPIR32TargetInfo : public SPIRTargetInfo {
Index: clang/test/Sema/arithmetic-fence-builtin.c
===================================================================
--- clang/test/Sema/arithmetic-fence-builtin.c
+++ clang/test/Sema/arithmetic-fence-builtin.c
@@ -2,6 +2,8 @@
// RUN: %clang_cc1 -triple ppc64le -DPPC -emit-llvm -o - -verify -x c++ %s
// RUN: not %clang_cc1 -triple ppc64le -DPPC -emit-llvm -o - -x c++ %s \
// RUN: -fprotect-parens 2>&1 | FileCheck -check-prefix=PPC %s
+// RUN: %clang_cc1 -triple spir64 -emit-llvm -fsycl-is-device \
+// RUN: -o - -verify -x c++ %s
#ifndef PPC
int v;
template <typename T> T addT(T a, T b) {
Index: clang/test/CodeGen/arithmetic-fence-builtin.c
===================================================================
--- clang/test/CodeGen/arithmetic-fence-builtin.c
+++ clang/test/CodeGen/arithmetic-fence-builtin.c
@@ -1,17 +1,23 @@
// Test with fast math
// RUN: %clang_cc1 -triple i386-pc-linux-gnu -emit-llvm -DFAST \
-// RUN: -mreassociate \
-// RUN: -o - %s | FileCheck --check-prefixes CHECK,CHECKFAST,CHECKNP %s
+// RUN: -mreassociate -o - %s \
+// RUN: | FileCheck --check-prefixes CHECK,CHECKFAST,CHECKPRECISE,CHECKNP %s
//
// Test with fast math and fprotect-parens
// RUN: %clang_cc1 -triple i386-pc-linux-gnu -emit-llvm -DFAST \
-// RUN: -mreassociate -fprotect-parens -ffp-contract=on\
-// RUN: -o - %s | FileCheck --check-prefixes CHECK,CHECKFAST,CHECKPP %s
+// RUN: -mreassociate -fprotect-parens -ffp-contract=on -o - %s \
+// RUN: | FileCheck --check-prefixes CHECK,CHECKFAST,CHECKPRECISE,CHECKPP %s
//
// Test without fast math: llvm intrinsic not created
// RUN: %clang_cc1 -triple i386-pc-linux-gnu -emit-llvm -fprotect-parens\
// RUN: -o - %s | FileCheck --implicit-check-not="llvm.arithmetic.fence" %s
//
+// Test with fast math on spir target
+// RUN: %clang_cc1 -triple spir64 -emit-llvm -fsycl-is-device \
+// RUN: -mreassociate -o - %s \
+// RUN: | FileCheck --check-prefixes CHECK,CHECKFAST,CHECKNP %s
+//
+
int v;
int addit(float a, float b) {
// CHECK: define {{.*}}@addit(float noundef %a, float noundef %b) #0 {
@@ -65,7 +71,7 @@
#ifdef FAST
#pragma float_control(precise, on)
int subit(float a, float b, float *fp) {
- // CHECKFAST: define {{.*}}@subit(float noundef %a, float noundef %b{{.*}}
+ // CHECKPRECISE: define {{.*}}@subit(float noundef %a, float noundef %b{{.*}}
*fp = __arithmetic_fence(a - b);
*fp = (a + b);
// CHECK-NOT: call{{.*}} float @llvm.arithmetic.fence.f32(float noundef %add)
Index: clang/lib/Basic/Targets/SPIR.h
===================================================================
--- clang/lib/Basic/Targets/SPIR.h
+++ clang/lib/Basic/Targets/SPIR.h
@@ -191,6 +191,8 @@
bool hasFeature(StringRef Feature) const override {
return Feature == "spir";
}
+
+ bool checkArithmeticFenceSupported() const override { return true; }
};
class LLVM_LIBRARY_VISIBILITY SPIR32TargetInfo : public SPIRTargetInfo {
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
