[PATCH] D45544: [AAch64] Add the __ARM_FEATURE_DOTPROD macro definition

2018-04-11 Thread Junmo Park via Phabricator via cfe-commits 
flyingforyou created this revision.
flyingforyou added a reviewer: SjoerdMeijer.
Herald added subscribers: kristof.beyls, javed.absar.

This matches what GCC does.
https://github.com/gcc-mirror/gcc/blob/master/gcc/config/aarch64/aarch64-c.c


https://reviews.llvm.org/D45544

Files:
  lib/Basic/Targets/AArch64.cpp
  lib/Basic/Targets/AArch64.h
  test/Preprocessor/aarch64-target-features.c


Index: test/Preprocessor/aarch64-target-features.c
===
--- test/Preprocessor/aarch64-target-features.c
+++ test/Preprocessor/aarch64-target-features.c
@@ -89,6 +89,10 @@
 // RUN: %clang -target aarch64-none-linux-gnu -march=armv8-a+sve -x c -E -dM 
%s -o - | FileCheck --check-prefix=CHECK-SVE %s
 // CHECK-SVE: __ARM_FEATURE_SVE 1
 
+// RUN: %clang -target aarch64-none-linux-gnu -march=armv8.2a+dotprod -x c -E
+// -dM %s -o - | FileCheck --check-prefix=CHECK-DOTPROD %s
+// CHECK-DOTPROD: __ARM_FEATURE_DOTPROD 1
+
 // RUN: %clang -target aarch64-none-linux-gnueabi -march=armv8.2a+fp16 -x c -E 
-dM %s -o - | FileCheck -match-full-lines 
--check-prefix=CHECK-FULLFP16-VECTOR-SCALAR %s
 // CHECK-FULLFP16-VECTOR-SCALAR: #define __ARM_FEATURE_FP16_SCALAR_ARITHMETIC 1
 // CHECK-FULLFP16-VECTOR-SCALAR: #define __ARM_FEATURE_FP16_VECTOR_ARITHMETIC 1
Index: lib/Basic/Targets/AArch64.h
===
--- lib/Basic/Targets/AArch64.h
+++ lib/Basic/Targets/AArch64.h
@@ -33,6 +33,7 @@
   unsigned Crypto;
   unsigned Unaligned;
   unsigned HasFullFP16;
+  unsigned HasDotProd;
   llvm::AArch64::ArchKind ArchKind;
 
   static const Builtin::Info BuiltinInfo[];
Index: lib/Basic/Targets/AArch64.cpp
===
--- lib/Basic/Targets/AArch64.cpp
+++ lib/Basic/Targets/AArch64.cpp
@@ -193,6 +193,9 @@
   if (HasFullFP16)
Builder.defineMacro("__ARM_FEATURE_FP16_SCALAR_ARITHMETIC", "1");
 
+  if (HasDotProd)
+Builder.defineMacro("__ARM_FEATURE_DOTPROD", "1");
+
   switch (ArchKind) {
   default:
 break;
@@ -248,6 +251,8 @@
   ArchKind = llvm::AArch64::ArchKind::ARMV8_2A;
 if (Feature == "+fullfp16")
   HasFullFP16 = 1;
+if (Feature == "+dotprod")
+  HasDotProd = 1;
   }
 
   setDataLayout();


Index: test/Preprocessor/aarch64-target-features.c
===
--- test/Preprocessor/aarch64-target-features.c
+++ test/Preprocessor/aarch64-target-features.c
@@ -89,6 +89,10 @@
 // RUN: %clang -target aarch64-none-linux-gnu -march=armv8-a+sve -x c -E -dM %s -o - | FileCheck --check-prefix=CHECK-SVE %s
 // CHECK-SVE: __ARM_FEATURE_SVE 1
 
+// RUN: %clang -target aarch64-none-linux-gnu -march=armv8.2a+dotprod -x c -E
+// -dM %s -o - | FileCheck --check-prefix=CHECK-DOTPROD %s
+// CHECK-DOTPROD: __ARM_FEATURE_DOTPROD 1
+
 // RUN: %clang -target aarch64-none-linux-gnueabi -march=armv8.2a+fp16 -x c -E -dM %s -o - | FileCheck -match-full-lines --check-prefix=CHECK-FULLFP16-VECTOR-SCALAR %s
 // CHECK-FULLFP16-VECTOR-SCALAR: #define __ARM_FEATURE_FP16_SCALAR_ARITHMETIC 1
 // CHECK-FULLFP16-VECTOR-SCALAR: #define __ARM_FEATURE_FP16_VECTOR_ARITHMETIC 1
Index: lib/Basic/Targets/AArch64.h
===
--- lib/Basic/Targets/AArch64.h
+++ lib/Basic/Targets/AArch64.h
@@ -33,6 +33,7 @@
   unsigned Crypto;
   unsigned Unaligned;
   unsigned HasFullFP16;
+  unsigned HasDotProd;
   llvm::AArch64::ArchKind ArchKind;
 
   static const Builtin::Info BuiltinInfo[];
Index: lib/Basic/Targets/AArch64.cpp
===
--- lib/Basic/Targets/AArch64.cpp
+++ lib/Basic/Targets/AArch64.cpp
@@ -193,6 +193,9 @@
   if (HasFullFP16)
Builder.defineMacro("__ARM_FEATURE_FP16_SCALAR_ARITHMETIC", "1");
 
+  if (HasDotProd)
+Builder.defineMacro("__ARM_FEATURE_DOTPROD", "1");
+
   switch (ArchKind) {
   default:
 break;
@@ -248,6 +251,8 @@
   ArchKind = llvm::AArch64::ArchKind::ARMV8_2A;
 if (Feature == "+fullfp16")
   HasFullFP16 = 1;
+if (Feature == "+dotprod")
+  HasDotProd = 1;
   }
 
   setDataLayout();
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D45544: [AAch64] Add the __ARM_FEATURE_DOTPROD macro definition

2018-04-12 Thread Junmo Park via Phabricator via cfe-commits 
flyingforyou updated this revision to Diff 142265.
flyingforyou added a comment.

Addressed SjoerdMeijer' new comments. Thanks.

> Also wanted to confirm that macro __ARM_FEATURE_DOTPROD will defined/included 
> in the next release of the ACLE.

Could you check it, please? I saw several codes that use above definition which 
are written by ARM engineers.


https://reviews.llvm.org/D45544

Files:
  lib/Basic/Targets/AArch64.cpp
  lib/Basic/Targets/AArch64.h
  test/Preprocessor/aarch64-target-features.c


Index: test/Preprocessor/aarch64-target-features.c
===
--- test/Preprocessor/aarch64-target-features.c
+++ test/Preprocessor/aarch64-target-features.c
@@ -38,6 +38,7 @@
 // CHECK-NOT: __ARM_SIZEOF_MINIMAL_ENUM 1
 // CHECK-NOT: __ARM_SIZEOF_WCHAR_T 2
 // CHECK-NOT: __ARM_FEATURE_SVE
+// CHECK-NOT: __ARM_FEATURE_DOTPROD
 
 // RUN: %clang -target aarch64_be-eabi -x c -E -dM %s -o - | FileCheck %s 
-check-prefix CHECK-BIGENDIAN
 // CHECK-BIGENDIAN: __ARM_BIG_ENDIAN 1
@@ -89,6 +90,9 @@
 // RUN: %clang -target aarch64-none-linux-gnu -march=armv8-a+sve -x c -E -dM 
%s -o - | FileCheck --check-prefix=CHECK-SVE %s
 // CHECK-SVE: __ARM_FEATURE_SVE 1
 
+// RUN: %clang -target aarch64-none-linux-gnu -march=armv8.2a+dotprod -x c -E 
-dM %s -o - | FileCheck --check-prefix=CHECK-DOTPROD %s
+// CHECK-DOTPROD: __ARM_FEATURE_DOTPROD 1
+
 // RUN: %clang -target aarch64-none-linux-gnueabi -march=armv8.2a+fp16 -x c -E 
-dM %s -o - | FileCheck -match-full-lines 
--check-prefix=CHECK-FULLFP16-VECTOR-SCALAR %s
 // CHECK-FULLFP16-VECTOR-SCALAR: #define __ARM_FEATURE_FP16_SCALAR_ARITHMETIC 1
 // CHECK-FULLFP16-VECTOR-SCALAR: #define __ARM_FEATURE_FP16_VECTOR_ARITHMETIC 1
Index: lib/Basic/Targets/AArch64.h
===
--- lib/Basic/Targets/AArch64.h
+++ lib/Basic/Targets/AArch64.h
@@ -33,6 +33,7 @@
   unsigned Crypto;
   unsigned Unaligned;
   unsigned HasFullFP16;
+  unsigned HasDotProd;
   llvm::AArch64::ArchKind ArchKind;
 
   static const Builtin::Info BuiltinInfo[];
Index: lib/Basic/Targets/AArch64.cpp
===
--- lib/Basic/Targets/AArch64.cpp
+++ lib/Basic/Targets/AArch64.cpp
@@ -193,6 +193,9 @@
   if (HasFullFP16)
Builder.defineMacro("__ARM_FEATURE_FP16_SCALAR_ARITHMETIC", "1");
 
+  if (HasDotProd)
+Builder.defineMacro("__ARM_FEATURE_DOTPROD", "1");
+
   switch (ArchKind) {
   default:
 break;
@@ -229,6 +232,7 @@
   Crypto = 0;
   Unaligned = 1;
   HasFullFP16 = 0;
+  HasDotProd = 0;
   ArchKind = llvm::AArch64::ArchKind::ARMV8A;
 
   for (const auto &Feature : Features) {
@@ -248,6 +252,8 @@
   ArchKind = llvm::AArch64::ArchKind::ARMV8_2A;
 if (Feature == "+fullfp16")
   HasFullFP16 = 1;
+if (Feature == "+dotprod")
+  HasDotProd = 1;
   }
 
   setDataLayout();


Index: test/Preprocessor/aarch64-target-features.c
===
--- test/Preprocessor/aarch64-target-features.c
+++ test/Preprocessor/aarch64-target-features.c
@@ -38,6 +38,7 @@
 // CHECK-NOT: __ARM_SIZEOF_MINIMAL_ENUM 1
 // CHECK-NOT: __ARM_SIZEOF_WCHAR_T 2
 // CHECK-NOT: __ARM_FEATURE_SVE
+// CHECK-NOT: __ARM_FEATURE_DOTPROD
 
 // RUN: %clang -target aarch64_be-eabi -x c -E -dM %s -o - | FileCheck %s -check-prefix CHECK-BIGENDIAN
 // CHECK-BIGENDIAN: __ARM_BIG_ENDIAN 1
@@ -89,6 +90,9 @@
 // RUN: %clang -target aarch64-none-linux-gnu -march=armv8-a+sve -x c -E -dM %s -o - | FileCheck --check-prefix=CHECK-SVE %s
 // CHECK-SVE: __ARM_FEATURE_SVE 1
 
+// RUN: %clang -target aarch64-none-linux-gnu -march=armv8.2a+dotprod -x c -E -dM %s -o - | FileCheck --check-prefix=CHECK-DOTPROD %s
+// CHECK-DOTPROD: __ARM_FEATURE_DOTPROD 1
+
 // RUN: %clang -target aarch64-none-linux-gnueabi -march=armv8.2a+fp16 -x c -E -dM %s -o - | FileCheck -match-full-lines --check-prefix=CHECK-FULLFP16-VECTOR-SCALAR %s
 // CHECK-FULLFP16-VECTOR-SCALAR: #define __ARM_FEATURE_FP16_SCALAR_ARITHMETIC 1
 // CHECK-FULLFP16-VECTOR-SCALAR: #define __ARM_FEATURE_FP16_VECTOR_ARITHMETIC 1
Index: lib/Basic/Targets/AArch64.h
===
--- lib/Basic/Targets/AArch64.h
+++ lib/Basic/Targets/AArch64.h
@@ -33,6 +33,7 @@
   unsigned Crypto;
   unsigned Unaligned;
   unsigned HasFullFP16;
+  unsigned HasDotProd;
   llvm::AArch64::ArchKind ArchKind;
 
   static const Builtin::Info BuiltinInfo[];
Index: lib/Basic/Targets/AArch64.cpp
===
--- lib/Basic/Targets/AArch64.cpp
+++ lib/Basic/Targets/AArch64.cpp
@@ -193,6 +193,9 @@
   if (HasFullFP16)
Builder.defineMacro("__ARM_FEATURE_FP16_SCALAR_ARITHMETIC", "1");
 
+  if (HasDotProd)
+Builder.defineMacro("__ARM_FEATURE_DOTPROD", "1");
+
   switch (ArchKind) {
   default:
 break;
@@ -229,6 +232,7 @@
   Crypto = 0;
   Unaligned = 1;
   HasFullFP16 = 0;
+  HasDotProd = 0;
   ArchKin

[PATCH] D45544: [AAch64] Add the __ARM_FEATURE_DOTPROD macro definition

2018-04-17 Thread Junmo Park via Phabricator via cfe-commits 
This revision was automatically updated to reflect the committed changes.
Closed by commit rL330229: [AAch64] Add the __ARM_FEATURE_DOTPROD macro 
definition (authored by flyingforyou, committed by ).
Herald added subscribers: llvm-commits, chrib.

Changed prior to commit:
  https://reviews.llvm.org/D45544?vs=142265&id=142851#toc

Repository:
  rL LLVM

https://reviews.llvm.org/D45544

Files:
  cfe/trunk/lib/Basic/Targets/AArch64.cpp
  cfe/trunk/lib/Basic/Targets/AArch64.h
  cfe/trunk/test/Preprocessor/aarch64-target-features.c


Index: cfe/trunk/test/Preprocessor/aarch64-target-features.c
===
--- cfe/trunk/test/Preprocessor/aarch64-target-features.c
+++ cfe/trunk/test/Preprocessor/aarch64-target-features.c
@@ -38,6 +38,7 @@
 // CHECK-NOT: __ARM_SIZEOF_MINIMAL_ENUM 1
 // CHECK-NOT: __ARM_SIZEOF_WCHAR_T 2
 // CHECK-NOT: __ARM_FEATURE_SVE
+// CHECK-NOT: __ARM_FEATURE_DOTPROD
 
 // RUN: %clang -target aarch64_be-eabi -x c -E -dM %s -o - | FileCheck %s 
-check-prefix CHECK-BIGENDIAN
 // CHECK-BIGENDIAN: __ARM_BIG_ENDIAN 1
@@ -89,6 +90,9 @@
 // RUN: %clang -target aarch64-none-linux-gnu -march=armv8-a+sve -x c -E -dM 
%s -o - | FileCheck --check-prefix=CHECK-SVE %s
 // CHECK-SVE: __ARM_FEATURE_SVE 1
 
+// RUN: %clang -target aarch64-none-linux-gnu -march=armv8.2a+dotprod -x c -E 
-dM %s -o - | FileCheck --check-prefix=CHECK-DOTPROD %s
+// CHECK-DOTPROD: __ARM_FEATURE_DOTPROD 1
+
 // RUN: %clang -target aarch64-none-linux-gnueabi -march=armv8.2a+fp16 -x c -E 
-dM %s -o - | FileCheck -match-full-lines 
--check-prefix=CHECK-FULLFP16-VECTOR-SCALAR %s
 // CHECK-FULLFP16-VECTOR-SCALAR: #define __ARM_FEATURE_FP16_SCALAR_ARITHMETIC 1
 // CHECK-FULLFP16-VECTOR-SCALAR: #define __ARM_FEATURE_FP16_VECTOR_ARITHMETIC 1
Index: cfe/trunk/lib/Basic/Targets/AArch64.cpp
===
--- cfe/trunk/lib/Basic/Targets/AArch64.cpp
+++ cfe/trunk/lib/Basic/Targets/AArch64.cpp
@@ -193,6 +193,9 @@
   if (HasFullFP16)
Builder.defineMacro("__ARM_FEATURE_FP16_SCALAR_ARITHMETIC", "1");
 
+  if (HasDotProd)
+Builder.defineMacro("__ARM_FEATURE_DOTPROD", "1");
+
   switch (ArchKind) {
   default:
 break;
@@ -229,6 +232,7 @@
   Crypto = 0;
   Unaligned = 1;
   HasFullFP16 = 0;
+  HasDotProd = 0;
   ArchKind = llvm::AArch64::ArchKind::ARMV8A;
 
   for (const auto &Feature : Features) {
@@ -248,6 +252,8 @@
   ArchKind = llvm::AArch64::ArchKind::ARMV8_2A;
 if (Feature == "+fullfp16")
   HasFullFP16 = 1;
+if (Feature == "+dotprod")
+  HasDotProd = 1;
   }
 
   setDataLayout();
Index: cfe/trunk/lib/Basic/Targets/AArch64.h
===
--- cfe/trunk/lib/Basic/Targets/AArch64.h
+++ cfe/trunk/lib/Basic/Targets/AArch64.h
@@ -33,6 +33,7 @@
   unsigned Crypto;
   unsigned Unaligned;
   unsigned HasFullFP16;
+  unsigned HasDotProd;
   llvm::AArch64::ArchKind ArchKind;
 
   static const Builtin::Info BuiltinInfo[];


Index: cfe/trunk/test/Preprocessor/aarch64-target-features.c
===
--- cfe/trunk/test/Preprocessor/aarch64-target-features.c
+++ cfe/trunk/test/Preprocessor/aarch64-target-features.c
@@ -38,6 +38,7 @@
 // CHECK-NOT: __ARM_SIZEOF_MINIMAL_ENUM 1
 // CHECK-NOT: __ARM_SIZEOF_WCHAR_T 2
 // CHECK-NOT: __ARM_FEATURE_SVE
+// CHECK-NOT: __ARM_FEATURE_DOTPROD
 
 // RUN: %clang -target aarch64_be-eabi -x c -E -dM %s -o - | FileCheck %s -check-prefix CHECK-BIGENDIAN
 // CHECK-BIGENDIAN: __ARM_BIG_ENDIAN 1
@@ -89,6 +90,9 @@
 // RUN: %clang -target aarch64-none-linux-gnu -march=armv8-a+sve -x c -E -dM %s -o - | FileCheck --check-prefix=CHECK-SVE %s
 // CHECK-SVE: __ARM_FEATURE_SVE 1
 
+// RUN: %clang -target aarch64-none-linux-gnu -march=armv8.2a+dotprod -x c -E -dM %s -o - | FileCheck --check-prefix=CHECK-DOTPROD %s
+// CHECK-DOTPROD: __ARM_FEATURE_DOTPROD 1
+
 // RUN: %clang -target aarch64-none-linux-gnueabi -march=armv8.2a+fp16 -x c -E -dM %s -o - | FileCheck -match-full-lines --check-prefix=CHECK-FULLFP16-VECTOR-SCALAR %s
 // CHECK-FULLFP16-VECTOR-SCALAR: #define __ARM_FEATURE_FP16_SCALAR_ARITHMETIC 1
 // CHECK-FULLFP16-VECTOR-SCALAR: #define __ARM_FEATURE_FP16_VECTOR_ARITHMETIC 1
Index: cfe/trunk/lib/Basic/Targets/AArch64.cpp
===
--- cfe/trunk/lib/Basic/Targets/AArch64.cpp
+++ cfe/trunk/lib/Basic/Targets/AArch64.cpp
@@ -193,6 +193,9 @@
   if (HasFullFP16)
Builder.defineMacro("__ARM_FEATURE_FP16_SCALAR_ARITHMETIC", "1");
 
+  if (HasDotProd)
+Builder.defineMacro("__ARM_FEATURE_DOTPROD", "1");
+
   switch (ArchKind) {
   default:
 break;
@@ -229,6 +232,7 @@
   Crypto = 0;
   Unaligned = 1;
   HasFullFP16 = 0;
+  HasDotProd = 0;
   ArchKind = llvm::AArch64::ArchKind::ARMV8A;
 
   for (const auto &Feature : Features) {
@@ -248,6 +252,8 @@
   ArchKind = llvm::AArch64::ArchKind::ARMV8_2A;
 if (Feature == "+fullfp16

[PATCH] D46109: [ARM, AArch64] Add intrinsics for dot product instructions

2018-04-26 Thread Junmo Park via Phabricator via cfe-commits 
flyingforyou added inline comments.



Comment at: test/CodeGen/arm-neon-dot-product.c:1
+// RUN: %clang_cc1 -triple armv8-linux-gnueabihf -target-cpu cortex-a57 
-target-feature +dotprod \
+// RUN: -disable-O0-optnone  -emit-llvm -o - %s | opt -S -instcombine | 
FileCheck %s

I think proper target is cortex-a55 or cortex-a75.
Do we need  check routines for wrong target-cpu?


Repository:
  rL LLVM

https://reviews.llvm.org/D46109



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D46109: [ARM, AArch64] Add intrinsics for dot product instructions

2018-04-26 Thread Junmo Park via Phabricator via cfe-commits 
flyingforyou added inline comments.



Comment at: test/CodeGen/arm-neon-dot-product.c:1
+// RUN: %clang_cc1 -triple armv8-linux-gnueabihf -target-cpu cortex-a57 
-target-feature +dotprod \
+// RUN: -disable-O0-optnone  -emit-llvm -o - %s | opt -S -instcombine | 
FileCheck %s

efriedma wrote:
> flyingforyou wrote:
> > I think proper target is cortex-a55 or cortex-a75.
> > Do we need  check routines for wrong target-cpu?
> This is working as intended, I think: target-feature overrides target-cpu.
dotprod is ARMv8.2's addition feature not ARMv8. Cortex-a57 only supports ARMv8 
which couldn't have dotprod feature. Am I missing something?


Repository:
  rL LLVM

https://reviews.llvm.org/D46109



___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits