[clang] [X86] Finally handle target of __builtin_ia32_cmp[p|s][s|d] from avx into sse/sse2/avx (PR #84136)
Freddy Ye via cfe-commits
cfe-commits at lists.llvm.org
Fri Mar 8 21:47:01 PST 2024
https://github.com/FreddyLeaf updated https://github.com/llvm/llvm-project/pull/84136
>From 2ba698b222fa1dc963d21850d5931562aa65533d Mon Sep 17 00:00:00 2001
From: Freddy Ye <freddy.ye at intel.com>
Date: Tue, 26 Sep 2023 16:44:01 +0800
Subject: [PATCH 1/6] [X86] Change target of __builtin_ia32_cmp[p|s][s|d] from
avx into sse/sse2
---
clang/include/clang/Basic/BuiltinsX86.def | 8 +-
clang/lib/Headers/avxintrin.h | 247 ------------------
clang/lib/Headers/emmintrin.h | 121 +++++++++
clang/lib/Headers/xmmintrin.h | 128 +++++++++
clang/test/CodeGen/X86/avx-builtins.c | 96 -------
.../test/CodeGen/X86/cmp-avx-builtins-error.c | 22 ++
clang/test/CodeGen/X86/sse-builtins.c | 54 ++++
clang/test/CodeGen/X86/sse2-builtins.c | 54 ++++
clang/test/CodeGen/target-features-error-2.c | 4 +-
9 files changed, 385 insertions(+), 349 deletions(-)
create mode 100644 clang/test/CodeGen/X86/cmp-avx-builtins-error.c
diff --git a/clang/include/clang/Basic/BuiltinsX86.def b/clang/include/clang/Basic/BuiltinsX86.def
index 207cde0414b54e..eafcc219c10966 100644
--- a/clang/include/clang/Basic/BuiltinsX86.def
+++ b/clang/include/clang/Basic/BuiltinsX86.def
@@ -226,6 +226,8 @@ TARGET_BUILTIN(__builtin_ia32_minps, "V4fV4fV4f", "ncV:128:", "sse")
TARGET_BUILTIN(__builtin_ia32_maxps, "V4fV4fV4f", "ncV:128:", "sse")
TARGET_BUILTIN(__builtin_ia32_minss, "V4fV4fV4f", "ncV:128:", "sse")
TARGET_BUILTIN(__builtin_ia32_maxss, "V4fV4fV4f", "ncV:128:", "sse")
+TARGET_BUILTIN(__builtin_ia32_cmpps, "V4fV4fV4fIc", "ncV:128:", "sse")
+TARGET_BUILTIN(__builtin_ia32_cmpss, "V4fV4fV4fIc", "ncV:128:", "sse")
TARGET_BUILTIN(__builtin_ia32_cmpeqpd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_cmpltpd, "V2dV2dV2d", "ncV:128:", "sse2")
@@ -243,6 +245,8 @@ TARGET_BUILTIN(__builtin_ia32_cmpneqsd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_cmpnltsd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_cmpnlesd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_cmpordsd, "V2dV2dV2d", "ncV:128:", "sse2")
+TARGET_BUILTIN(__builtin_ia32_cmpsd, "V2dV2dV2dIc", "ncV:128:", "sse2")
+TARGET_BUILTIN(__builtin_ia32_cmppd, "V2dV2dV2dIc", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_minpd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_maxpd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_minsd, "V2dV2dV2d", "ncV:128:", "sse2")
@@ -462,12 +466,8 @@ TARGET_BUILTIN(__builtin_ia32_blendvps256, "V8fV8fV8fV8f", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_shufpd256, "V4dV4dV4dIi", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_shufps256, "V8fV8fV8fIi", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_dpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmppd, "V2dV2dV2dIc", "ncV:128:", "avx")
TARGET_BUILTIN(__builtin_ia32_cmppd256, "V4dV4dV4dIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpps, "V4fV4fV4fIc", "ncV:128:", "avx")
TARGET_BUILTIN(__builtin_ia32_cmpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpsd, "V2dV2dV2dIc", "ncV:128:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpss, "V4fV4fV4fIc", "ncV:128:", "avx")
TARGET_BUILTIN(__builtin_ia32_vextractf128_pd256, "V2dV4dIi", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_vextractf128_ps256, "V4fV8fIi", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_vextractf128_si256, "V4iV8iIi", "ncV:256:", "avx")
diff --git a/clang/lib/Headers/avxintrin.h b/clang/lib/Headers/avxintrin.h
index f116d8bc3a94c7..d6192518ea24ba 100644
--- a/clang/lib/Headers/avxintrin.h
+++ b/clang/lib/Headers/avxintrin.h
@@ -1573,15 +1573,6 @@ _mm256_blendv_ps(__m256 __a, __m256 __b, __m256 __c)
((__m256d)__builtin_ia32_shufpd256((__v4df)(__m256d)(a), \
(__v4df)(__m256d)(b), (int)(mask)))
-/* Compare */
-#define _CMP_EQ_OQ 0x00 /* Equal (ordered, non-signaling) */
-#define _CMP_LT_OS 0x01 /* Less-than (ordered, signaling) */
-#define _CMP_LE_OS 0x02 /* Less-than-or-equal (ordered, signaling) */
-#define _CMP_UNORD_Q 0x03 /* Unordered (non-signaling) */
-#define _CMP_NEQ_UQ 0x04 /* Not-equal (unordered, non-signaling) */
-#define _CMP_NLT_US 0x05 /* Not-less-than (unordered, signaling) */
-#define _CMP_NLE_US 0x06 /* Not-less-than-or-equal (unordered, signaling) */
-#define _CMP_ORD_Q 0x07 /* Ordered (non-signaling) */
#define _CMP_EQ_UQ 0x08 /* Equal (unordered, non-signaling) */
#define _CMP_NGE_US 0x09 /* Not-greater-than-or-equal (unordered, signaling) */
#define _CMP_NGT_US 0x0a /* Not-greater-than (unordered, signaling) */
@@ -1607,126 +1598,6 @@ _mm256_blendv_ps(__m256 __a, __m256 __b, __m256 __c)
#define _CMP_GT_OQ 0x1e /* Greater-than (ordered, non-signaling) */
#define _CMP_TRUE_US 0x1f /* True (unordered, signaling) */
-/// Compares each of the corresponding double-precision values of two
-/// 128-bit vectors of [2 x double], using the operation specified by the
-/// immediate integer operand.
-///
-/// Returns a [2 x double] vector consisting of two doubles corresponding to
-/// the two comparison results: zero if the comparison is false, and all 1's
-/// if the comparison is true.
-///
-/// \headerfile <x86intrin.h>
-///
-/// \code
-/// __m128d _mm_cmp_pd(__m128d a, __m128d b, const int c);
-/// \endcode
-///
-/// This intrinsic corresponds to the <c> VCMPPD </c> instruction.
-///
-/// \param a
-/// A 128-bit vector of [2 x double].
-/// \param b
-/// A 128-bit vector of [2 x double].
-/// \param c
-/// An immediate integer operand, with bits [4:0] specifying which comparison
-/// operation to use: \n
-/// 0x00: Equal (ordered, non-signaling) \n
-/// 0x01: Less-than (ordered, signaling) \n
-/// 0x02: Less-than-or-equal (ordered, signaling) \n
-/// 0x03: Unordered (non-signaling) \n
-/// 0x04: Not-equal (unordered, non-signaling) \n
-/// 0x05: Not-less-than (unordered, signaling) \n
-/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
-/// 0x07: Ordered (non-signaling) \n
-/// 0x08: Equal (unordered, non-signaling) \n
-/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
-/// 0x0A: Not-greater-than (unordered, signaling) \n
-/// 0x0B: False (ordered, non-signaling) \n
-/// 0x0C: Not-equal (ordered, non-signaling) \n
-/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
-/// 0x0E: Greater-than (ordered, signaling) \n
-/// 0x0F: True (unordered, non-signaling) \n
-/// 0x10: Equal (ordered, signaling) \n
-/// 0x11: Less-than (ordered, non-signaling) \n
-/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
-/// 0x13: Unordered (signaling) \n
-/// 0x14: Not-equal (unordered, signaling) \n
-/// 0x15: Not-less-than (unordered, non-signaling) \n
-/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
-/// 0x17: Ordered (signaling) \n
-/// 0x18: Equal (unordered, signaling) \n
-/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
-/// 0x1A: Not-greater-than (unordered, non-signaling) \n
-/// 0x1B: False (ordered, signaling) \n
-/// 0x1C: Not-equal (ordered, signaling) \n
-/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
-/// 0x1E: Greater-than (ordered, non-signaling) \n
-/// 0x1F: True (unordered, signaling)
-/// \returns A 128-bit vector of [2 x double] containing the comparison results.
-#define _mm_cmp_pd(a, b, c) \
- ((__m128d)__builtin_ia32_cmppd((__v2df)(__m128d)(a), \
- (__v2df)(__m128d)(b), (c)))
-
-/// Compares each of the corresponding values of two 128-bit vectors of
-/// [4 x float], using the operation specified by the immediate integer
-/// operand.
-///
-/// Returns a [4 x float] vector consisting of four floats corresponding to
-/// the four comparison results: zero if the comparison is false, and all 1's
-/// if the comparison is true.
-///
-/// \headerfile <x86intrin.h>
-///
-/// \code
-/// __m128 _mm_cmp_ps(__m128 a, __m128 b, const int c);
-/// \endcode
-///
-/// This intrinsic corresponds to the <c> VCMPPS </c> instruction.
-///
-/// \param a
-/// A 128-bit vector of [4 x float].
-/// \param b
-/// A 128-bit vector of [4 x float].
-/// \param c
-/// An immediate integer operand, with bits [4:0] specifying which comparison
-/// operation to use: \n
-/// 0x00: Equal (ordered, non-signaling) \n
-/// 0x01: Less-than (ordered, signaling) \n
-/// 0x02: Less-than-or-equal (ordered, signaling) \n
-/// 0x03: Unordered (non-signaling) \n
-/// 0x04: Not-equal (unordered, non-signaling) \n
-/// 0x05: Not-less-than (unordered, signaling) \n
-/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
-/// 0x07: Ordered (non-signaling) \n
-/// 0x08: Equal (unordered, non-signaling) \n
-/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
-/// 0x0A: Not-greater-than (unordered, signaling) \n
-/// 0x0B: False (ordered, non-signaling) \n
-/// 0x0C: Not-equal (ordered, non-signaling) \n
-/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
-/// 0x0E: Greater-than (ordered, signaling) \n
-/// 0x0F: True (unordered, non-signaling) \n
-/// 0x10: Equal (ordered, signaling) \n
-/// 0x11: Less-than (ordered, non-signaling) \n
-/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
-/// 0x13: Unordered (signaling) \n
-/// 0x14: Not-equal (unordered, signaling) \n
-/// 0x15: Not-less-than (unordered, non-signaling) \n
-/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
-/// 0x17: Ordered (signaling) \n
-/// 0x18: Equal (unordered, signaling) \n
-/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
-/// 0x1A: Not-greater-than (unordered, non-signaling) \n
-/// 0x1B: False (ordered, signaling) \n
-/// 0x1C: Not-equal (ordered, signaling) \n
-/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
-/// 0x1E: Greater-than (ordered, non-signaling) \n
-/// 0x1F: True (unordered, signaling)
-/// \returns A 128-bit vector of [4 x float] containing the comparison results.
-#define _mm_cmp_ps(a, b, c) \
- ((__m128)__builtin_ia32_cmpps((__v4sf)(__m128)(a), \
- (__v4sf)(__m128)(b), (c)))
-
/// Compares each of the corresponding double-precision values of two
/// 256-bit vectors of [4 x double], using the operation specified by the
/// immediate integer operand.
@@ -1847,124 +1718,6 @@ _mm256_blendv_ps(__m256 __a, __m256 __b, __m256 __c)
((__m256)__builtin_ia32_cmpps256((__v8sf)(__m256)(a), \
(__v8sf)(__m256)(b), (c)))
-/// Compares each of the corresponding scalar double-precision values of
-/// two 128-bit vectors of [2 x double], using the operation specified by the
-/// immediate integer operand.
-///
-/// If the result is true, all 64 bits of the destination vector are set;
-/// otherwise they are cleared.
-///
-/// \headerfile <x86intrin.h>
-///
-/// \code
-/// __m128d _mm_cmp_sd(__m128d a, __m128d b, const int c);
-/// \endcode
-///
-/// This intrinsic corresponds to the <c> VCMPSD </c> instruction.
-///
-/// \param a
-/// A 128-bit vector of [2 x double].
-/// \param b
-/// A 128-bit vector of [2 x double].
-/// \param c
-/// An immediate integer operand, with bits [4:0] specifying which comparison
-/// operation to use: \n
-/// 0x00: Equal (ordered, non-signaling) \n
-/// 0x01: Less-than (ordered, signaling) \n
-/// 0x02: Less-than-or-equal (ordered, signaling) \n
-/// 0x03: Unordered (non-signaling) \n
-/// 0x04: Not-equal (unordered, non-signaling) \n
-/// 0x05: Not-less-than (unordered, signaling) \n
-/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
-/// 0x07: Ordered (non-signaling) \n
-/// 0x08: Equal (unordered, non-signaling) \n
-/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
-/// 0x0A: Not-greater-than (unordered, signaling) \n
-/// 0x0B: False (ordered, non-signaling) \n
-/// 0x0C: Not-equal (ordered, non-signaling) \n
-/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
-/// 0x0E: Greater-than (ordered, signaling) \n
-/// 0x0F: True (unordered, non-signaling) \n
-/// 0x10: Equal (ordered, signaling) \n
-/// 0x11: Less-than (ordered, non-signaling) \n
-/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
-/// 0x13: Unordered (signaling) \n
-/// 0x14: Not-equal (unordered, signaling) \n
-/// 0x15: Not-less-than (unordered, non-signaling) \n
-/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
-/// 0x17: Ordered (signaling) \n
-/// 0x18: Equal (unordered, signaling) \n
-/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
-/// 0x1A: Not-greater-than (unordered, non-signaling) \n
-/// 0x1B: False (ordered, signaling) \n
-/// 0x1C: Not-equal (ordered, signaling) \n
-/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
-/// 0x1E: Greater-than (ordered, non-signaling) \n
-/// 0x1F: True (unordered, signaling)
-/// \returns A 128-bit vector of [2 x double] containing the comparison results.
-#define _mm_cmp_sd(a, b, c) \
- ((__m128d)__builtin_ia32_cmpsd((__v2df)(__m128d)(a), \
- (__v2df)(__m128d)(b), (c)))
-
-/// Compares each of the corresponding scalar values of two 128-bit
-/// vectors of [4 x float], using the operation specified by the immediate
-/// integer operand.
-///
-/// If the result is true, all 32 bits of the destination vector are set;
-/// otherwise they are cleared.
-///
-/// \headerfile <x86intrin.h>
-///
-/// \code
-/// __m128 _mm_cmp_ss(__m128 a, __m128 b, const int c);
-/// \endcode
-///
-/// This intrinsic corresponds to the <c> VCMPSS </c> instruction.
-///
-/// \param a
-/// A 128-bit vector of [4 x float].
-/// \param b
-/// A 128-bit vector of [4 x float].
-/// \param c
-/// An immediate integer operand, with bits [4:0] specifying which comparison
-/// operation to use: \n
-/// 0x00: Equal (ordered, non-signaling) \n
-/// 0x01: Less-than (ordered, signaling) \n
-/// 0x02: Less-than-or-equal (ordered, signaling) \n
-/// 0x03: Unordered (non-signaling) \n
-/// 0x04: Not-equal (unordered, non-signaling) \n
-/// 0x05: Not-less-than (unordered, signaling) \n
-/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
-/// 0x07: Ordered (non-signaling) \n
-/// 0x08: Equal (unordered, non-signaling) \n
-/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
-/// 0x0A: Not-greater-than (unordered, signaling) \n
-/// 0x0B: False (ordered, non-signaling) \n
-/// 0x0C: Not-equal (ordered, non-signaling) \n
-/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
-/// 0x0E: Greater-than (ordered, signaling) \n
-/// 0x0F: True (unordered, non-signaling) \n
-/// 0x10: Equal (ordered, signaling) \n
-/// 0x11: Less-than (ordered, non-signaling) \n
-/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
-/// 0x13: Unordered (signaling) \n
-/// 0x14: Not-equal (unordered, signaling) \n
-/// 0x15: Not-less-than (unordered, non-signaling) \n
-/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
-/// 0x17: Ordered (signaling) \n
-/// 0x18: Equal (unordered, signaling) \n
-/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
-/// 0x1A: Not-greater-than (unordered, non-signaling) \n
-/// 0x1B: False (ordered, signaling) \n
-/// 0x1C: Not-equal (ordered, signaling) \n
-/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
-/// 0x1E: Greater-than (ordered, non-signaling) \n
-/// 0x1F: True (unordered, signaling)
-/// \returns A 128-bit vector of [4 x float] containing the comparison results.
-#define _mm_cmp_ss(a, b, c) \
- ((__m128)__builtin_ia32_cmpss((__v4sf)(__m128)(a), \
- (__v4sf)(__m128)(b), (c)))
-
/// Takes a [8 x i32] vector and returns the vector element value
/// indexed by the immediate constant operand.
///
diff --git a/clang/lib/Headers/emmintrin.h b/clang/lib/Headers/emmintrin.h
index 1d451b5f5b25de..db6d43ea4ad6fe 100644
--- a/clang/lib/Headers/emmintrin.h
+++ b/clang/lib/Headers/emmintrin.h
@@ -4745,6 +4745,127 @@ static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_castsi128_pd(__m128i __a) {
return (__m128d)__a;
}
+/// Compares each of the corresponding double-precision values of two
+/// 128-bit vectors of [2 x double], using the operation specified by the
+/// immediate integer operand.
+///
+/// Returns a [2 x double] vector consisting of two doubles corresponding to
+/// the two comparison results: zero if the comparison is false, and all 1's
+/// if the comparison is true.
+///
+/// \headerfile <x86intrin.h>
+///
+/// \code
+/// __m128d _mm_cmp_pd(__m128d a, __m128d b, const int c);
+/// \endcode
+///
+/// This intrinsic corresponds to the <c> (V)CMPPD </c> instruction.
+///
+/// \param a
+/// A 128-bit vector of [2 x double].
+/// \param b
+/// A 128-bit vector of [2 x double].
+/// \param c
+/// An immediate integer operand, with bits [4:0] specifying which comparison
+/// operation to use: \n
+/// (Note that without avx enabled, only bits [2:0] are supported) \n
+/// 0x00: Equal (ordered, non-signaling) \n
+/// 0x01: Less-than (ordered, signaling) \n
+/// 0x02: Less-than-or-equal (ordered, signaling) \n
+/// 0x03: Unordered (non-signaling) \n
+/// 0x04: Not-equal (unordered, non-signaling) \n
+/// 0x05: Not-less-than (unordered, signaling) \n
+/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
+/// 0x07: Ordered (non-signaling) \n
+/// 0x08: Equal (unordered, non-signaling) \n
+/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
+/// 0x0A: Not-greater-than (unordered, signaling) \n
+/// 0x0B: False (ordered, non-signaling) \n
+/// 0x0C: Not-equal (ordered, non-signaling) \n
+/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
+/// 0x0E: Greater-than (ordered, signaling) \n
+/// 0x0F: True (unordered, non-signaling) \n
+/// 0x10: Equal (ordered, signaling) \n
+/// 0x11: Less-than (ordered, non-signaling) \n
+/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
+/// 0x13: Unordered (signaling) \n
+/// 0x14: Not-equal (unordered, signaling) \n
+/// 0x15: Not-less-than (unordered, non-signaling) \n
+/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
+/// 0x17: Ordered (signaling) \n
+/// 0x18: Equal (unordered, signaling) \n
+/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
+/// 0x1A: Not-greater-than (unordered, non-signaling) \n
+/// 0x1B: False (ordered, signaling) \n
+/// 0x1C: Not-equal (ordered, signaling) \n
+/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
+/// 0x1E: Greater-than (ordered, non-signaling) \n
+/// 0x1F: True (unordered, signaling)
+/// \returns A 128-bit vector of [2 x double] containing the comparison results.
+#define _mm_cmp_pd(a, b, c) \
+ ((__m128d)__builtin_ia32_cmppd((__v2df)(__m128d)(a), (__v2df)(__m128d)(b), \
+ (c)))
+
+/// Compares each of the corresponding scalar double-precision values of
+/// two 128-bit vectors of [2 x double], using the operation specified by the
+/// immediate integer operand.
+///
+/// If the result is true, all 64 bits of the destination vector are set;
+/// otherwise they are cleared.
+///
+/// \headerfile <x86intrin.h>
+///
+/// \code
+/// __m128d _mm_cmp_sd(__m128d a, __m128d b, const int c);
+/// \endcode
+///
+/// This intrinsic corresponds to the <c> (V)CMPSD </c> instruction.
+///
+/// \param a
+/// A 128-bit vector of [2 x double].
+/// \param b
+/// A 128-bit vector of [2 x double].
+/// \param c
+/// An immediate integer operand, with bits [4:0] specifying which comparison
+/// operation to use: \n
+/// (Note that without avx enabled, only bits [2:0] are supported) \n
+/// 0x00: Equal (ordered, non-signaling) \n
+/// 0x01: Less-than (ordered, signaling) \n
+/// 0x02: Less-than-or-equal (ordered, signaling) \n
+/// 0x03: Unordered (non-signaling) \n
+/// 0x04: Not-equal (unordered, non-signaling) \n
+/// 0x05: Not-less-than (unordered, signaling) \n
+/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
+/// 0x07: Ordered (non-signaling) \n
+/// 0x08: Equal (unordered, non-signaling) \n
+/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
+/// 0x0A: Not-greater-than (unordered, signaling) \n
+/// 0x0B: False (ordered, non-signaling) \n
+/// 0x0C: Not-equal (ordered, non-signaling) \n
+/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
+/// 0x0E: Greater-than (ordered, signaling) \n
+/// 0x0F: True (unordered, non-signaling) \n
+/// 0x10: Equal (ordered, signaling) \n
+/// 0x11: Less-than (ordered, non-signaling) \n
+/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
+/// 0x13: Unordered (signaling) \n
+/// 0x14: Not-equal (unordered, signaling) \n
+/// 0x15: Not-less-than (unordered, non-signaling) \n
+/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
+/// 0x17: Ordered (signaling) \n
+/// 0x18: Equal (unordered, signaling) \n
+/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
+/// 0x1A: Not-greater-than (unordered, non-signaling) \n
+/// 0x1B: False (ordered, signaling) \n
+/// 0x1C: Not-equal (ordered, signaling) \n
+/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
+/// 0x1E: Greater-than (ordered, non-signaling) \n
+/// 0x1F: True (unordered, signaling)
+/// \returns A 128-bit vector of [2 x double] containing the comparison results.
+#define _mm_cmp_sd(a, b, c) \
+ ((__m128d)__builtin_ia32_cmpsd((__v2df)(__m128d)(a), (__v2df)(__m128d)(b), \
+ (c)))
+
#if defined(__cplusplus)
extern "C" {
#endif
diff --git a/clang/lib/Headers/xmmintrin.h b/clang/lib/Headers/xmmintrin.h
index 47368f3c23d2d6..f9894c23ce25bf 100644
--- a/clang/lib/Headers/xmmintrin.h
+++ b/clang/lib/Headers/xmmintrin.h
@@ -2940,6 +2940,134 @@ _mm_movemask_ps(__m128 __a)
return __builtin_ia32_movmskps((__v4sf)__a);
}
+/* Compare */
+#define _CMP_EQ_OQ 0x00 /* Equal (ordered, non-signaling) */
+#define _CMP_LT_OS 0x01 /* Less-than (ordered, signaling) */
+#define _CMP_LE_OS 0x02 /* Less-than-or-equal (ordered, signaling) */
+#define _CMP_UNORD_Q 0x03 /* Unordered (non-signaling) */
+#define _CMP_NEQ_UQ 0x04 /* Not-equal (unordered, non-signaling) */
+#define _CMP_NLT_US 0x05 /* Not-less-than (unordered, signaling) */
+#define _CMP_NLE_US 0x06 /* Not-less-than-or-equal (unordered, signaling) */
+#define _CMP_ORD_Q 0x07 /* Ordered (non-signaling) */
+
+/// Compares each of the corresponding values of two 128-bit vectors of
+/// [4 x float], using the operation specified by the immediate integer
+/// operand.
+///
+/// Returns a [4 x float] vector consisting of four floats corresponding to
+/// the four comparison results: zero if the comparison is false, and all 1's
+/// if the comparison is true.
+///
+/// \headerfile <x86intrin.h>
+///
+/// \code
+/// __m128 _mm_cmp_ps(__m128 a, __m128 b, const int c);
+/// \endcode
+///
+/// This intrinsic corresponds to the <c> (V)CMPPS </c> instruction.
+///
+/// \param a
+/// A 128-bit vector of [4 x float].
+/// \param b
+/// A 128-bit vector of [4 x float].
+/// \param c
+/// An immediate integer operand, with bits [4:0] specifying which comparison
+/// operation to use: \n
+/// (Note that without avx enabled, only bits [2:0] are supported) \n
+/// 0x00: Equal (ordered, non-signaling) \n
+/// 0x01: Less-than (ordered, signaling) \n
+/// 0x02: Less-than-or-equal (ordered, signaling) \n
+/// 0x03: Unordered (non-signaling) \n
+/// 0x04: Not-equal (unordered, non-signaling) \n
+/// 0x05: Not-less-than (unordered, signaling) \n
+/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
+/// 0x07: Ordered (non-signaling) \n
+/// 0x08: Equal (unordered, non-signaling) \n
+/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
+/// 0x0A: Not-greater-than (unordered, signaling) \n
+/// 0x0B: False (ordered, non-signaling) \n
+/// 0x0C: Not-equal (ordered, non-signaling) \n
+/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
+/// 0x0E: Greater-than (ordered, signaling) \n
+/// 0x0F: True (unordered, non-signaling) \n
+/// 0x10: Equal (ordered, signaling) \n
+/// 0x11: Less-than (ordered, non-signaling) \n
+/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
+/// 0x13: Unordered (signaling) \n
+/// 0x14: Not-equal (unordered, signaling) \n
+/// 0x15: Not-less-than (unordered, non-signaling) \n
+/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
+/// 0x17: Ordered (signaling) \n
+/// 0x18: Equal (unordered, signaling) \n
+/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
+/// 0x1A: Not-greater-than (unordered, non-signaling) \n
+/// 0x1B: False (ordered, signaling) \n
+/// 0x1C: Not-equal (ordered, signaling) \n
+/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
+/// 0x1E: Greater-than (ordered, non-signaling) \n
+/// 0x1F: True (unordered, signaling)
+/// \returns A 128-bit vector of [4 x float] containing the comparison results.
+#define _mm_cmp_ps(a, b, c) \
+ ((__m128)__builtin_ia32_cmpps((__v4sf)(__m128)(a), (__v4sf)(__m128)(b), (c)))
+
+/// Compares each of the corresponding scalar values of two 128-bit
+/// vectors of [4 x float], using the operation specified by the immediate
+/// integer operand.
+///
+/// If the result is true, all 32 bits of the destination vector are set;
+/// otherwise they are cleared.
+///
+/// \headerfile <x86intrin.h>
+///
+/// \code
+/// __m128 _mm_cmp_ss(__m128 a, __m128 b, const int c);
+/// \endcode
+///
+/// This intrinsic corresponds to the <c> (V)CMPSS </c> instruction.
+///
+/// \param a
+/// A 128-bit vector of [4 x float].
+/// \param b
+/// A 128-bit vector of [4 x float].
+/// \param c
+/// An immediate integer operand, with bits [4:0] specifying which comparison
+/// operation to use: \n
+/// (Note that without avx enabled, only bits [2:0] are supported) \n
+/// 0x00: Equal (ordered, non-signaling) \n
+/// 0x01: Less-than (ordered, signaling) \n
+/// 0x02: Less-than-or-equal (ordered, signaling) \n
+/// 0x03: Unordered (non-signaling) \n
+/// 0x04: Not-equal (unordered, non-signaling) \n
+/// 0x05: Not-less-than (unordered, signaling) \n
+/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
+/// 0x07: Ordered (non-signaling) \n
+/// 0x08: Equal (unordered, non-signaling) \n
+/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
+/// 0x0A: Not-greater-than (unordered, signaling) \n
+/// 0x0B: False (ordered, non-signaling) \n
+/// 0x0C: Not-equal (ordered, non-signaling) \n
+/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
+/// 0x0E: Greater-than (ordered, signaling) \n
+/// 0x0F: True (unordered, non-signaling) \n
+/// 0x10: Equal (ordered, signaling) \n
+/// 0x11: Less-than (ordered, non-signaling) \n
+/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
+/// 0x13: Unordered (signaling) \n
+/// 0x14: Not-equal (unordered, signaling) \n
+/// 0x15: Not-less-than (unordered, non-signaling) \n
+/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
+/// 0x17: Ordered (signaling) \n
+/// 0x18: Equal (unordered, signaling) \n
+/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
+/// 0x1A: Not-greater-than (unordered, non-signaling) \n
+/// 0x1B: False (ordered, signaling) \n
+/// 0x1C: Not-equal (ordered, signaling) \n
+/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
+/// 0x1E: Greater-than (ordered, non-signaling) \n
+/// 0x1F: True (unordered, signaling)
+/// \returns A 128-bit vector of [4 x float] containing the comparison results.
+#define _mm_cmp_ss(a, b, c) \
+ ((__m128)__builtin_ia32_cmpss((__v4sf)(__m128)(a), (__v4sf)(__m128)(b), (c)))
#define _MM_ALIGN16 __attribute__((aligned(16)))
diff --git a/clang/test/CodeGen/X86/avx-builtins.c b/clang/test/CodeGen/X86/avx-builtins.c
index d50366c3a022cc..4bf1213d9fca97 100644
--- a/clang/test/CodeGen/X86/avx-builtins.c
+++ b/clang/test/CodeGen/X86/avx-builtins.c
@@ -596,54 +596,6 @@ __m256 test_mm256_cmp_ps_true_us(__m256 a, __m256 b) {
return _mm256_cmp_ps(a, b, _CMP_TRUE_US);
}
-__m128d test_mm_cmp_pd_eq_oq(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_eq_oq
- // CHECK: fcmp oeq <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_EQ_OQ);
-}
-
-__m128d test_mm_cmp_pd_lt_os(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_lt_os
- // CHECK: fcmp olt <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_LT_OS);
-}
-
-__m128d test_mm_cmp_pd_le_os(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_le_os
- // CHECK: fcmp ole <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_LE_OS);
-}
-
-__m128d test_mm_cmp_pd_unord_q(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_unord_q
- // CHECK: fcmp uno <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_UNORD_Q);
-}
-
-__m128d test_mm_cmp_pd_neq_uq(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_neq_uq
- // CHECK: fcmp une <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_NEQ_UQ);
-}
-
-__m128d test_mm_cmp_pd_nlt_us(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_nlt_us
- // CHECK: fcmp uge <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_NLT_US);
-}
-
-__m128d test_mm_cmp_pd_nle_us(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_nle_us
- // CHECK: fcmp ugt <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_NLE_US);
-}
-
-__m128d test_mm_cmp_pd_ord_q(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_ord_q
- // CHECK: fcmp ord <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_ORD_Q);
-}
-
__m128d test_mm_cmp_pd_eq_uq(__m128d a, __m128d b) {
// CHECK-LABEL: test_mm_cmp_pd_eq_uq
// CHECK: fcmp ueq <2 x double> %{{.*}}, %{{.*}}
@@ -788,54 +740,6 @@ __m128d test_mm_cmp_pd_true_us(__m128d a, __m128d b) {
return _mm_cmp_pd(a, b, _CMP_TRUE_US);
}
-__m128 test_mm_cmp_ps_eq_oq(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_eq_oq
- // CHECK: fcmp oeq <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_EQ_OQ);
-}
-
-__m128 test_mm_cmp_ps_lt_os(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_lt_os
- // CHECK: fcmp olt <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_LT_OS);
-}
-
-__m128 test_mm_cmp_ps_le_os(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_le_os
- // CHECK: fcmp ole <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_LE_OS);
-}
-
-__m128 test_mm_cmp_ps_unord_q(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_unord_q
- // CHECK: fcmp uno <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_UNORD_Q);
-}
-
-__m128 test_mm_cmp_ps_neq_uq(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_neq_uq
- // CHECK: fcmp une <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_NEQ_UQ);
-}
-
-__m128 test_mm_cmp_ps_nlt_us(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_nlt_us
- // CHECK: fcmp uge <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_NLT_US);
-}
-
-__m128 test_mm_cmp_ps_nle_us(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_nle_us
- // CHECK: fcmp ugt <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_NLE_US);
-}
-
-__m128 test_mm_cmp_ps_ord_q(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_ord_q
- // CHECK: fcmp ord <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_ORD_Q);
-}
-
__m128 test_mm_cmp_ps_eq_uq(__m128 a, __m128 b) {
// CHECK-LABEL: test_mm_cmp_ps_eq_uq
// CHECK: fcmp ueq <4 x float> %{{.*}}, %{{.*}}
diff --git a/clang/test/CodeGen/X86/cmp-avx-builtins-error.c b/clang/test/CodeGen/X86/cmp-avx-builtins-error.c
new file mode 100644
index 00000000000000..2b35aa84492b64
--- /dev/null
+++ b/clang/test/CodeGen/X86/cmp-avx-builtins-error.c
@@ -0,0 +1,22 @@
+// RUN: %clang_cc1 %s -ffreestanding -triple=x86_64-unknown-unknown \
+// RUN: -target-feature +avx -emit-llvm -fsyntax-only -verify
+// RUN: %clang_cc1 %s -ffreestanding -triple=i386-unknown-unknown \
+// RUN: -target-feature +avx -emit-llvm -fsyntax-only -verify
+
+#include <immintrin.h>
+
+__m128d test_mm_cmp_pd(__m128d a, __m128d b) {
+ return _mm_cmp_pd(a, b, 32); // expected-error {{argument value 32 is outside the valid range [0, 31]}}
+}
+
+__m128d test_mm_cmp_sd(__m128d a, __m128d b) {
+ return _mm_cmp_sd(a, b, 32); // expected-error {{argument value 32 is outside the valid range [0, 31]}}
+}
+
+__m128 test_mm_cmp_ps(__m128 a, __m128 b) {
+ return _mm_cmp_pd(a, b, 32); // expected-error {{argument value 32 is outside the valid range [0, 31]}}
+}
+
+__m128 test_mm_cmp_ss(__m128 a, __m128 b) {
+ return _mm_cmp_sd(a, b, 32); // expected-error {{argument value 32 is outside the valid range [0, 31]}}
+}
diff --git a/clang/test/CodeGen/X86/sse-builtins.c b/clang/test/CodeGen/X86/sse-builtins.c
index 830081296e52fb..9494fa46d54aed 100644
--- a/clang/test/CodeGen/X86/sse-builtins.c
+++ b/clang/test/CodeGen/X86/sse-builtins.c
@@ -813,3 +813,57 @@ __m128 test_mm_xor_ps(__m128 A, __m128 B) {
// CHECK: xor <4 x i32>
return _mm_xor_ps(A, B);
}
+
+__m128 test_mm_cmp_ps_eq_oq(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_eq_oq
+ // CHECK: fcmp oeq <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_EQ_OQ);
+}
+
+__m128 test_mm_cmp_ps_lt_os(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_lt_os
+ // CHECK: fcmp olt <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_LT_OS);
+}
+
+__m128 test_mm_cmp_ps_le_os(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_le_os
+ // CHECK: fcmp ole <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_LE_OS);
+}
+
+__m128 test_mm_cmp_ps_unord_q(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_unord_q
+ // CHECK: fcmp uno <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_UNORD_Q);
+}
+
+__m128 test_mm_cmp_ps_neq_uq(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_neq_uq
+ // CHECK: fcmp une <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_NEQ_UQ);
+}
+
+__m128 test_mm_cmp_ps_nlt_us(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_nlt_us
+ // CHECK: fcmp uge <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_NLT_US);
+}
+
+__m128 test_mm_cmp_ps_nle_us(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_nle_us
+ // CHECK: fcmp ugt <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_NLE_US);
+}
+
+__m128 test_mm_cmp_ps_ord_q(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_ord_q
+ // CHECK: fcmp ord <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_ORD_Q);
+}
+
+__m128 test_mm_cmp_ss(__m128 A, __m128 B) {
+ // CHECK-LABEL: test_mm_cmp_ss
+ // CHECK: call <4 x float> @llvm.x86.sse.cmp.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 7)
+ return _mm_cmp_ss(A, B, _CMP_ORD_Q);
+}
diff --git a/clang/test/CodeGen/X86/sse2-builtins.c b/clang/test/CodeGen/X86/sse2-builtins.c
index 44b2e485a5bd85..1189fb964dff07 100644
--- a/clang/test/CodeGen/X86/sse2-builtins.c
+++ b/clang/test/CodeGen/X86/sse2-builtins.c
@@ -1719,3 +1719,57 @@ __m128i test_mm_xor_si128(__m128i A, __m128i B) {
// CHECK: xor <2 x i64> %{{.*}}, %{{.*}}
return _mm_xor_si128(A, B);
}
+
+__m128d test_mm_cmp_pd_eq_oq(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_eq_oq
+ // CHECK: fcmp oeq <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_EQ_OQ);
+}
+
+__m128d test_mm_cmp_pd_lt_os(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_lt_os
+ // CHECK: fcmp olt <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_LT_OS);
+}
+
+__m128d test_mm_cmp_pd_le_os(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_le_os
+ // CHECK: fcmp ole <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_LE_OS);
+}
+
+__m128d test_mm_cmp_pd_unord_q(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_unord_q
+ // CHECK: fcmp uno <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_UNORD_Q);
+}
+
+__m128d test_mm_cmp_pd_neq_uq(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_neq_uq
+ // CHECK: fcmp une <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_NEQ_UQ);
+}
+
+__m128d test_mm_cmp_pd_nlt_us(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_nlt_us
+ // CHECK: fcmp uge <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_NLT_US);
+}
+
+__m128d test_mm_cmp_pd_nle_us(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_nle_us
+ // CHECK: fcmp ugt <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_NLE_US);
+}
+
+__m128d test_mm_cmp_pd_ord_q(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_ord_q
+ // CHECK: fcmp ord <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_ORD_Q);
+}
+
+__m128d test_mm_cmp_sd(__m128d A, __m128d B) {
+ // CHECK-LABEL: test_mm_cmp_sd
+ // CHECK: call <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 7)
+ return _mm_cmp_sd(A, B, _CMP_ORD_Q);
+}
diff --git a/clang/test/CodeGen/target-features-error-2.c b/clang/test/CodeGen/target-features-error-2.c
index 60586fb57f1c04..4f8bc8712aa51d 100644
--- a/clang/test/CodeGen/target-features-error-2.c
+++ b/clang/test/CodeGen/target-features-error-2.c
@@ -14,8 +14,8 @@ int baz(__m256i a) {
#endif
#if NEED_AVX_2
-__m128 need_avx(__m128 a, __m128 b) {
- return _mm_cmp_ps(a, b, 0); // expected-error {{'__builtin_ia32_cmpps' needs target feature avx}}
+__m256 need_avx(__m256 a, __m256 b) {
+ return _mm256_cmp_ps(a, b, 0); // expected-error {{'__builtin_ia32_cmpps256' needs target feature avx}}
}
#endif
>From 54c78b5ad694946186ba115576a6eea9c8992349 Mon Sep 17 00:00:00 2001
From: Freddy Ye <freddy.ye at intel.com>
Date: Mon, 26 Feb 2024 09:51:07 +0800
Subject: [PATCH 2/6] Handle attribute/target
---
clang/include/clang/Basic/BuiltinsX86.def | 8 ++++----
clang/lib/CodeGen/CodeGenFunction.cpp | 19 +++++++++++++++++++
.../CodeGen/X86/attribute-cmpsd-no-error.c | 11 +++++++++++
clang/test/CodeGen/target-features-error-2.c | 4 ++--
4 files changed, 36 insertions(+), 6 deletions(-)
create mode 100644 clang/test/CodeGen/X86/attribute-cmpsd-no-error.c
diff --git a/clang/include/clang/Basic/BuiltinsX86.def b/clang/include/clang/Basic/BuiltinsX86.def
index eafcc219c10966..d141d51a509e8d 100644
--- a/clang/include/clang/Basic/BuiltinsX86.def
+++ b/clang/include/clang/Basic/BuiltinsX86.def
@@ -226,8 +226,6 @@ TARGET_BUILTIN(__builtin_ia32_minps, "V4fV4fV4f", "ncV:128:", "sse")
TARGET_BUILTIN(__builtin_ia32_maxps, "V4fV4fV4f", "ncV:128:", "sse")
TARGET_BUILTIN(__builtin_ia32_minss, "V4fV4fV4f", "ncV:128:", "sse")
TARGET_BUILTIN(__builtin_ia32_maxss, "V4fV4fV4f", "ncV:128:", "sse")
-TARGET_BUILTIN(__builtin_ia32_cmpps, "V4fV4fV4fIc", "ncV:128:", "sse")
-TARGET_BUILTIN(__builtin_ia32_cmpss, "V4fV4fV4fIc", "ncV:128:", "sse")
TARGET_BUILTIN(__builtin_ia32_cmpeqpd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_cmpltpd, "V2dV2dV2d", "ncV:128:", "sse2")
@@ -245,8 +243,6 @@ TARGET_BUILTIN(__builtin_ia32_cmpneqsd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_cmpnltsd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_cmpnlesd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_cmpordsd, "V2dV2dV2d", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_cmpsd, "V2dV2dV2dIc", "ncV:128:", "sse2")
-TARGET_BUILTIN(__builtin_ia32_cmppd, "V2dV2dV2dIc", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_minpd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_maxpd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_minsd, "V2dV2dV2d", "ncV:128:", "sse2")
@@ -466,8 +462,12 @@ TARGET_BUILTIN(__builtin_ia32_blendvps256, "V8fV8fV8fV8f", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_shufpd256, "V4dV4dV4dIi", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_shufps256, "V8fV8fV8fIi", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_dpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
+TARGET_BUILTIN(__builtin_ia32_cmppd, "V2dV2dV2dIc", "ncV:128:", "avx|sse2")
TARGET_BUILTIN(__builtin_ia32_cmppd256, "V4dV4dV4dIc", "ncV:256:", "avx")
+TARGET_BUILTIN(__builtin_ia32_cmpps, "V4fV4fV4fIc", "ncV:128:", "avx|sse")
TARGET_BUILTIN(__builtin_ia32_cmpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
+TARGET_BUILTIN(__builtin_ia32_cmpsd, "V2dV2dV2dIc", "ncV:128:", "avx|sse2")
+TARGET_BUILTIN(__builtin_ia32_cmpss, "V4fV4fV4fIc", "ncV:128:", "avx|sse")
TARGET_BUILTIN(__builtin_ia32_vextractf128_pd256, "V2dV4dIi", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_vextractf128_ps256, "V4fV8fIi", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_vextractf128_si256, "V4iV8iIi", "ncV:256:", "avx")
diff --git a/clang/lib/CodeGen/CodeGenFunction.cpp b/clang/lib/CodeGen/CodeGenFunction.cpp
index b87fc86f4e635f..f6544452ffb7c1 100644
--- a/clang/lib/CodeGen/CodeGenFunction.cpp
+++ b/clang/lib/CodeGen/CodeGenFunction.cpp
@@ -31,6 +31,7 @@
#include "clang/AST/StmtObjC.h"
#include "clang/Basic/Builtins.h"
#include "clang/Basic/CodeGenOptions.h"
+#include "clang/Basic/TargetBuiltins.h"
#include "clang/Basic/TargetInfo.h"
#include "clang/CodeGen/CGFunctionInfo.h"
#include "clang/Frontend/FrontendDiagnostic.h"
@@ -2613,6 +2614,24 @@ void CGBuilderInserter::InsertHelper(
// called function.
void CodeGenFunction::checkTargetFeatures(const CallExpr *E,
const FunctionDecl *TargetDecl) {
+ // SemaCheking cannot handle below x86 builtins because they have different
+ // parameter ranges with different TargetAttribute of caller.
+ if (CGM.getContext().getTargetInfo().getTriple().isX86()) {
+ unsigned BuiltinID = TargetDecl->getBuiltinID();
+ if (BuiltinID == X86::BI__builtin_ia32_cmpps ||
+ BuiltinID == X86::BI__builtin_ia32_cmpss ||
+ BuiltinID == X86::BI__builtin_ia32_cmppd ||
+ BuiltinID == X86::BI__builtin_ia32_cmpsd) {
+ const FunctionDecl *FD = dyn_cast_or_null<FunctionDecl>(CurCodeDecl);
+ llvm::StringMap<bool> TargetFetureMap;
+ CGM.getContext().getFunctionFeatureMap(TargetFetureMap, FD);
+ llvm::APSInt Result =
+ *(E->getArg(2)->getIntegerConstantExpr(CGM.getContext()));
+ if (Result.getSExtValue() > 7 && !TargetFetureMap.lookup("avx"))
+ CGM.getDiags().Report(E->getBeginLoc(), diag::err_builtin_needs_feature)
+ << TargetDecl->getDeclName() << "avx";
+ }
+ }
return checkTargetFeatures(E->getBeginLoc(), TargetDecl);
}
diff --git a/clang/test/CodeGen/X86/attribute-cmpsd-no-error.c b/clang/test/CodeGen/X86/attribute-cmpsd-no-error.c
new file mode 100644
index 00000000000000..a73d998155d096
--- /dev/null
+++ b/clang/test/CodeGen/X86/attribute-cmpsd-no-error.c
@@ -0,0 +1,11 @@
+// RUN: %clang_cc1 %s -ffreestanding -triple=x86_64-unknown-unknown-emit-llvm -o /dev/null -verify
+// RUN: %clang_cc1 %s -ffreestanding -triple=i386-unknown-unknown-emit-llvm -o /dev/null -verify
+
+// expected-no-diagnostics
+
+#include <immintrin.h>
+
+__attribute__((target("avx")))
+__m128 test(__m128 a, __m128 b) {
+ return _mm_cmp_ps(a, b, 14);
+}
diff --git a/clang/test/CodeGen/target-features-error-2.c b/clang/test/CodeGen/target-features-error-2.c
index 4f8bc8712aa51d..1599b0135b747f 100644
--- a/clang/test/CodeGen/target-features-error-2.c
+++ b/clang/test/CodeGen/target-features-error-2.c
@@ -14,8 +14,8 @@ int baz(__m256i a) {
#endif
#if NEED_AVX_2
-__m256 need_avx(__m256 a, __m256 b) {
- return _mm256_cmp_ps(a, b, 0); // expected-error {{'__builtin_ia32_cmpps256' needs target feature avx}}
+__m128 need_avx(__m128 a, __m128 b) {
+ return _mm_cmp_ps(a, b, 8); // expected-error {{'__builtin_ia32_cmpps' needs target feature avx}}
}
#endif
>From d8979e29e1932b42bb54424e1c6aa97ba9ba07e7 Mon Sep 17 00:00:00 2001
From: Freddy Ye <freddy.ye at intel.com>
Date: Wed, 6 Mar 2024 20:12:10 +0800
Subject: [PATCH 3/6] Address comments.
---
clang/include/clang/Basic/BuiltinsX86.def | 8 +-
clang/lib/CodeGen/CodeGenFunction.cpp | 2 +-
clang/test/CodeGen/X86/sse-builtins.c | 108 +++++++++++-----------
clang/test/CodeGen/X86/sse2-builtins.c | 108 +++++++++++-----------
4 files changed, 113 insertions(+), 113 deletions(-)
diff --git a/clang/include/clang/Basic/BuiltinsX86.def b/clang/include/clang/Basic/BuiltinsX86.def
index d141d51a509e8d..eafcc219c10966 100644
--- a/clang/include/clang/Basic/BuiltinsX86.def
+++ b/clang/include/clang/Basic/BuiltinsX86.def
@@ -226,6 +226,8 @@ TARGET_BUILTIN(__builtin_ia32_minps, "V4fV4fV4f", "ncV:128:", "sse")
TARGET_BUILTIN(__builtin_ia32_maxps, "V4fV4fV4f", "ncV:128:", "sse")
TARGET_BUILTIN(__builtin_ia32_minss, "V4fV4fV4f", "ncV:128:", "sse")
TARGET_BUILTIN(__builtin_ia32_maxss, "V4fV4fV4f", "ncV:128:", "sse")
+TARGET_BUILTIN(__builtin_ia32_cmpps, "V4fV4fV4fIc", "ncV:128:", "sse")
+TARGET_BUILTIN(__builtin_ia32_cmpss, "V4fV4fV4fIc", "ncV:128:", "sse")
TARGET_BUILTIN(__builtin_ia32_cmpeqpd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_cmpltpd, "V2dV2dV2d", "ncV:128:", "sse2")
@@ -243,6 +245,8 @@ TARGET_BUILTIN(__builtin_ia32_cmpneqsd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_cmpnltsd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_cmpnlesd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_cmpordsd, "V2dV2dV2d", "ncV:128:", "sse2")
+TARGET_BUILTIN(__builtin_ia32_cmpsd, "V2dV2dV2dIc", "ncV:128:", "sse2")
+TARGET_BUILTIN(__builtin_ia32_cmppd, "V2dV2dV2dIc", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_minpd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_maxpd, "V2dV2dV2d", "ncV:128:", "sse2")
TARGET_BUILTIN(__builtin_ia32_minsd, "V2dV2dV2d", "ncV:128:", "sse2")
@@ -462,12 +466,8 @@ TARGET_BUILTIN(__builtin_ia32_blendvps256, "V8fV8fV8fV8f", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_shufpd256, "V4dV4dV4dIi", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_shufps256, "V8fV8fV8fIi", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_dpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmppd, "V2dV2dV2dIc", "ncV:128:", "avx|sse2")
TARGET_BUILTIN(__builtin_ia32_cmppd256, "V4dV4dV4dIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpps, "V4fV4fV4fIc", "ncV:128:", "avx|sse")
TARGET_BUILTIN(__builtin_ia32_cmpps256, "V8fV8fV8fIc", "ncV:256:", "avx")
-TARGET_BUILTIN(__builtin_ia32_cmpsd, "V2dV2dV2dIc", "ncV:128:", "avx|sse2")
-TARGET_BUILTIN(__builtin_ia32_cmpss, "V4fV4fV4fIc", "ncV:128:", "avx|sse")
TARGET_BUILTIN(__builtin_ia32_vextractf128_pd256, "V2dV4dIi", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_vextractf128_ps256, "V4fV8fIi", "ncV:256:", "avx")
TARGET_BUILTIN(__builtin_ia32_vextractf128_si256, "V4iV8iIi", "ncV:256:", "avx")
diff --git a/clang/lib/CodeGen/CodeGenFunction.cpp b/clang/lib/CodeGen/CodeGenFunction.cpp
index f6544452ffb7c1..4a3ff49b0007a3 100644
--- a/clang/lib/CodeGen/CodeGenFunction.cpp
+++ b/clang/lib/CodeGen/CodeGenFunction.cpp
@@ -2614,7 +2614,7 @@ void CGBuilderInserter::InsertHelper(
// called function.
void CodeGenFunction::checkTargetFeatures(const CallExpr *E,
const FunctionDecl *TargetDecl) {
- // SemaCheking cannot handle below x86 builtins because they have different
+ // SemaChecking cannot handle below x86 builtins because they have different
// parameter ranges with different TargetAttribute of caller.
if (CGM.getContext().getTargetInfo().getTriple().isX86()) {
unsigned BuiltinID = TargetDecl->getBuiltinID();
diff --git a/clang/test/CodeGen/X86/sse-builtins.c b/clang/test/CodeGen/X86/sse-builtins.c
index 9494fa46d54aed..aae5cfb8bb1d99 100644
--- a/clang/test/CodeGen/X86/sse-builtins.c
+++ b/clang/test/CodeGen/X86/sse-builtins.c
@@ -34,6 +34,60 @@ __m128 test_mm_andnot_ps(__m128 A, __m128 B) {
return _mm_andnot_ps(A, B);
}
+__m128 test_mm_cmp_ps_eq_oq(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_eq_oq
+ // CHECK: fcmp oeq <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_EQ_OQ);
+}
+
+__m128 test_mm_cmp_ps_lt_os(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_lt_os
+ // CHECK: fcmp olt <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_LT_OS);
+}
+
+__m128 test_mm_cmp_ps_le_os(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_le_os
+ // CHECK: fcmp ole <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_LE_OS);
+}
+
+__m128 test_mm_cmp_ps_unord_q(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_unord_q
+ // CHECK: fcmp uno <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_UNORD_Q);
+}
+
+__m128 test_mm_cmp_ps_neq_uq(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_neq_uq
+ // CHECK: fcmp une <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_NEQ_UQ);
+}
+
+__m128 test_mm_cmp_ps_nlt_us(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_nlt_us
+ // CHECK: fcmp uge <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_NLT_US);
+}
+
+__m128 test_mm_cmp_ps_nle_us(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_nle_us
+ // CHECK: fcmp ugt <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_NLE_US);
+}
+
+__m128 test_mm_cmp_ps_ord_q(__m128 a, __m128 b) {
+ // CHECK-LABEL: test_mm_cmp_ps_ord_q
+ // CHECK: fcmp ord <4 x float> %{{.*}}, %{{.*}}
+ return _mm_cmp_ps(a, b, _CMP_ORD_Q);
+}
+
+__m128 test_mm_cmp_ss(__m128 A, __m128 B) {
+ // CHECK-LABEL: test_mm_cmp_ss
+ // CHECK: call <4 x float> @llvm.x86.sse.cmp.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 7)
+ return _mm_cmp_ss(A, B, _CMP_ORD_Q);
+}
+
__m128 test_mm_cmpeq_ps(__m128 __a, __m128 __b) {
// CHECK-LABEL: test_mm_cmpeq_ps
// CHECK: [[CMP:%.*]] = fcmp oeq <4 x float>
@@ -813,57 +867,3 @@ __m128 test_mm_xor_ps(__m128 A, __m128 B) {
// CHECK: xor <4 x i32>
return _mm_xor_ps(A, B);
}
-
-__m128 test_mm_cmp_ps_eq_oq(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_eq_oq
- // CHECK: fcmp oeq <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_EQ_OQ);
-}
-
-__m128 test_mm_cmp_ps_lt_os(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_lt_os
- // CHECK: fcmp olt <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_LT_OS);
-}
-
-__m128 test_mm_cmp_ps_le_os(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_le_os
- // CHECK: fcmp ole <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_LE_OS);
-}
-
-__m128 test_mm_cmp_ps_unord_q(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_unord_q
- // CHECK: fcmp uno <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_UNORD_Q);
-}
-
-__m128 test_mm_cmp_ps_neq_uq(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_neq_uq
- // CHECK: fcmp une <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_NEQ_UQ);
-}
-
-__m128 test_mm_cmp_ps_nlt_us(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_nlt_us
- // CHECK: fcmp uge <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_NLT_US);
-}
-
-__m128 test_mm_cmp_ps_nle_us(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_nle_us
- // CHECK: fcmp ugt <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_NLE_US);
-}
-
-__m128 test_mm_cmp_ps_ord_q(__m128 a, __m128 b) {
- // CHECK-LABEL: test_mm_cmp_ps_ord_q
- // CHECK: fcmp ord <4 x float> %{{.*}}, %{{.*}}
- return _mm_cmp_ps(a, b, _CMP_ORD_Q);
-}
-
-__m128 test_mm_cmp_ss(__m128 A, __m128 B) {
- // CHECK-LABEL: test_mm_cmp_ss
- // CHECK: call <4 x float> @llvm.x86.sse.cmp.ss(<4 x float> %{{.*}}, <4 x float> %{{.*}}, i8 7)
- return _mm_cmp_ss(A, B, _CMP_ORD_Q);
-}
diff --git a/clang/test/CodeGen/X86/sse2-builtins.c b/clang/test/CodeGen/X86/sse2-builtins.c
index 1189fb964dff07..3ca7777e763359 100644
--- a/clang/test/CodeGen/X86/sse2-builtins.c
+++ b/clang/test/CodeGen/X86/sse2-builtins.c
@@ -160,6 +160,60 @@ void test_mm_clflush(void* A) {
_mm_clflush(A);
}
+__m128d test_mm_cmp_pd_eq_oq(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_eq_oq
+ // CHECK: fcmp oeq <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_EQ_OQ);
+}
+
+__m128d test_mm_cmp_pd_lt_os(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_lt_os
+ // CHECK: fcmp olt <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_LT_OS);
+}
+
+__m128d test_mm_cmp_pd_le_os(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_le_os
+ // CHECK: fcmp ole <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_LE_OS);
+}
+
+__m128d test_mm_cmp_pd_unord_q(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_unord_q
+ // CHECK: fcmp uno <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_UNORD_Q);
+}
+
+__m128d test_mm_cmp_pd_neq_uq(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_neq_uq
+ // CHECK: fcmp une <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_NEQ_UQ);
+}
+
+__m128d test_mm_cmp_pd_nlt_us(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_nlt_us
+ // CHECK: fcmp uge <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_NLT_US);
+}
+
+__m128d test_mm_cmp_pd_nle_us(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_nle_us
+ // CHECK: fcmp ugt <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_NLE_US);
+}
+
+__m128d test_mm_cmp_pd_ord_q(__m128d a, __m128d b) {
+ // CHECK-LABEL: test_mm_cmp_pd_ord_q
+ // CHECK: fcmp ord <2 x double> %{{.*}}, %{{.*}}
+ return _mm_cmp_pd(a, b, _CMP_ORD_Q);
+}
+
+__m128d test_mm_cmp_sd(__m128d A, __m128d B) {
+ // CHECK-LABEL: test_mm_cmp_sd
+ // CHECK: call <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 7)
+ return _mm_cmp_sd(A, B, _CMP_ORD_Q);
+}
+
__m128i test_mm_cmpeq_epi8(__m128i A, __m128i B) {
// CHECK-LABEL: test_mm_cmpeq_epi8
// CHECK: icmp eq <16 x i8>
@@ -1719,57 +1773,3 @@ __m128i test_mm_xor_si128(__m128i A, __m128i B) {
// CHECK: xor <2 x i64> %{{.*}}, %{{.*}}
return _mm_xor_si128(A, B);
}
-
-__m128d test_mm_cmp_pd_eq_oq(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_eq_oq
- // CHECK: fcmp oeq <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_EQ_OQ);
-}
-
-__m128d test_mm_cmp_pd_lt_os(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_lt_os
- // CHECK: fcmp olt <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_LT_OS);
-}
-
-__m128d test_mm_cmp_pd_le_os(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_le_os
- // CHECK: fcmp ole <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_LE_OS);
-}
-
-__m128d test_mm_cmp_pd_unord_q(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_unord_q
- // CHECK: fcmp uno <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_UNORD_Q);
-}
-
-__m128d test_mm_cmp_pd_neq_uq(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_neq_uq
- // CHECK: fcmp une <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_NEQ_UQ);
-}
-
-__m128d test_mm_cmp_pd_nlt_us(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_nlt_us
- // CHECK: fcmp uge <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_NLT_US);
-}
-
-__m128d test_mm_cmp_pd_nle_us(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_nle_us
- // CHECK: fcmp ugt <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_NLE_US);
-}
-
-__m128d test_mm_cmp_pd_ord_q(__m128d a, __m128d b) {
- // CHECK-LABEL: test_mm_cmp_pd_ord_q
- // CHECK: fcmp ord <2 x double> %{{.*}}, %{{.*}}
- return _mm_cmp_pd(a, b, _CMP_ORD_Q);
-}
-
-__m128d test_mm_cmp_sd(__m128d A, __m128d B) {
- // CHECK-LABEL: test_mm_cmp_sd
- // CHECK: call <2 x double> @llvm.x86.sse2.cmp.sd(<2 x double> %{{.*}}, <2 x double> %{{.*}}, i8 7)
- return _mm_cmp_sd(A, B, _CMP_ORD_Q);
-}
>From c9a3d8905c02f50a0938a88cc391ded1b623884f Mon Sep 17 00:00:00 2001
From: Freddy Ye <freddy.ye at intel.com>
Date: Wed, 6 Mar 2024 20:16:42 +0800
Subject: [PATCH 4/6] typo
---
clang/test/CodeGen/X86/cmp-avx-builtins-error.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/clang/test/CodeGen/X86/cmp-avx-builtins-error.c b/clang/test/CodeGen/X86/cmp-avx-builtins-error.c
index 2b35aa84492b64..bad2606021b69e 100644
--- a/clang/test/CodeGen/X86/cmp-avx-builtins-error.c
+++ b/clang/test/CodeGen/X86/cmp-avx-builtins-error.c
@@ -14,9 +14,9 @@ __m128d test_mm_cmp_sd(__m128d a, __m128d b) {
}
__m128 test_mm_cmp_ps(__m128 a, __m128 b) {
- return _mm_cmp_pd(a, b, 32); // expected-error {{argument value 32 is outside the valid range [0, 31]}}
+ return _mm_cmp_ps(a, b, 32); // expected-error {{argument value 32 is outside the valid range [0, 31]}}
}
__m128 test_mm_cmp_ss(__m128 a, __m128 b) {
- return _mm_cmp_sd(a, b, 32); // expected-error {{argument value 32 is outside the valid range [0, 31]}}
+ return _mm_cmp_ss(a, b, 32); // expected-error {{argument value 32 is outside the valid range [0, 31]}}
}
>From 51eda069c797c91fd3072f347e1dd4a20ab53702 Mon Sep 17 00:00:00 2001
From: Freddy Ye <freddy.ye at intel.com>
Date: Thu, 7 Mar 2024 09:44:52 +0800
Subject: [PATCH 5/6] Address comments.
---
clang/lib/Headers/avxintrin.h | 235 ++++++++++++++++++++++++++++++++++
clang/lib/Headers/emmintrin.h | 50 --------
clang/lib/Headers/xmmintrin.h | 50 --------
3 files changed, 235 insertions(+), 100 deletions(-)
diff --git a/clang/lib/Headers/avxintrin.h b/clang/lib/Headers/avxintrin.h
index d6192518ea24ba..ecd9bf18a41ca0 100644
--- a/clang/lib/Headers/avxintrin.h
+++ b/clang/lib/Headers/avxintrin.h
@@ -1573,6 +1573,7 @@ _mm256_blendv_ps(__m256 __a, __m256 __b, __m256 __c)
((__m256d)__builtin_ia32_shufpd256((__v4df)(__m256d)(a), \
(__v4df)(__m256d)(b), (int)(mask)))
+/* Compare */
#define _CMP_EQ_UQ 0x08 /* Equal (unordered, non-signaling) */
#define _CMP_NGE_US 0x09 /* Not-greater-than-or-equal (unordered, signaling) */
#define _CMP_NGT_US 0x0a /* Not-greater-than (unordered, signaling) */
@@ -1598,6 +1599,124 @@ _mm256_blendv_ps(__m256 __a, __m256 __b, __m256 __c)
#define _CMP_GT_OQ 0x1e /* Greater-than (ordered, non-signaling) */
#define _CMP_TRUE_US 0x1f /* True (unordered, signaling) */
+/* Below intrinsic defined in emmintrin.h can be used for AVX */
+/// Compares each of the corresponding double-precision values of two
+/// 128-bit vectors of [2 x double], using the operation specified by the
+/// immediate integer operand.
+///
+/// Returns a [2 x double] vector consisting of two doubles corresponding to
+/// the two comparison results: zero if the comparison is false, and all 1's
+/// if the comparison is true.
+///
+/// \headerfile <x86intrin.h>
+///
+/// \code
+/// __m128d _mm_cmp_pd(__m128d a, __m128d b, const int c);
+/// \endcode
+///
+/// This intrinsic corresponds to the <c> VCMPPD </c> instruction.
+///
+/// \param a
+/// A 128-bit vector of [2 x double].
+/// \param b
+/// A 128-bit vector of [2 x double].
+/// \param c
+/// An immediate integer operand, with bits [4:0] specifying which comparison
+/// operation to use: \n
+/// 0x00: Equal (ordered, non-signaling) \n
+/// 0x01: Less-than (ordered, signaling) \n
+/// 0x02: Less-than-or-equal (ordered, signaling) \n
+/// 0x03: Unordered (non-signaling) \n
+/// 0x04: Not-equal (unordered, non-signaling) \n
+/// 0x05: Not-less-than (unordered, signaling) \n
+/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
+/// 0x07: Ordered (non-signaling) \n
+/// 0x08: Equal (unordered, non-signaling) \n
+/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
+/// 0x0A: Not-greater-than (unordered, signaling) \n
+/// 0x0B: False (ordered, non-signaling) \n
+/// 0x0C: Not-equal (ordered, non-signaling) \n
+/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
+/// 0x0E: Greater-than (ordered, signaling) \n
+/// 0x0F: True (unordered, non-signaling) \n
+/// 0x10: Equal (ordered, signaling) \n
+/// 0x11: Less-than (ordered, non-signaling) \n
+/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
+/// 0x13: Unordered (signaling) \n
+/// 0x14: Not-equal (unordered, signaling) \n
+/// 0x15: Not-less-than (unordered, non-signaling) \n
+/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
+/// 0x17: Ordered (signaling) \n
+/// 0x18: Equal (unordered, signaling) \n
+/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
+/// 0x1A: Not-greater-than (unordered, non-signaling) \n
+/// 0x1B: False (ordered, signaling) \n
+/// 0x1C: Not-equal (ordered, signaling) \n
+/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
+/// 0x1E: Greater-than (ordered, non-signaling) \n
+/// 0x1F: True (unordered, signaling)
+/// \returns A 128-bit vector of [2 x double] containing the comparison results.
+/// \fn __m128d _mm_cmp_pd(__m128d a, __m128d b, const int c)
+
+/* Below intrinsic defined in xmmintrin.h can be used for AVX */
+/// Compares each of the corresponding values of two 128-bit vectors of
+/// [4 x float], using the operation specified by the immediate integer
+/// operand.
+///
+/// Returns a [4 x float] vector consisting of four floats corresponding to
+/// the four comparison results: zero if the comparison is false, and all 1's
+/// if the comparison is true.
+///
+/// \headerfile <x86intrin.h>
+///
+/// \code
+/// __m128 _mm_cmp_ps(__m128 a, __m128 b, const int c);
+/// \endcode
+///
+/// This intrinsic corresponds to the <c> VCMPPS </c> instruction.
+///
+/// \param a
+/// A 128-bit vector of [4 x float].
+/// \param b
+/// A 128-bit vector of [4 x float].
+/// \param c
+/// An immediate integer operand, with bits [4:0] specifying which comparison
+/// operation to use: \n
+/// 0x00: Equal (ordered, non-signaling) \n
+/// 0x01: Less-than (ordered, signaling) \n
+/// 0x02: Less-than-or-equal (ordered, signaling) \n
+/// 0x03: Unordered (non-signaling) \n
+/// 0x04: Not-equal (unordered, non-signaling) \n
+/// 0x05: Not-less-than (unordered, signaling) \n
+/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
+/// 0x07: Ordered (non-signaling) \n
+/// 0x08: Equal (unordered, non-signaling) \n
+/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
+/// 0x0A: Not-greater-than (unordered, signaling) \n
+/// 0x0B: False (ordered, non-signaling) \n
+/// 0x0C: Not-equal (ordered, non-signaling) \n
+/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
+/// 0x0E: Greater-than (ordered, signaling) \n
+/// 0x0F: True (unordered, non-signaling) \n
+/// 0x10: Equal (ordered, signaling) \n
+/// 0x11: Less-than (ordered, non-signaling) \n
+/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
+/// 0x13: Unordered (signaling) \n
+/// 0x14: Not-equal (unordered, signaling) \n
+/// 0x15: Not-less-than (unordered, non-signaling) \n
+/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
+/// 0x17: Ordered (signaling) \n
+/// 0x18: Equal (unordered, signaling) \n
+/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
+/// 0x1A: Not-greater-than (unordered, non-signaling) \n
+/// 0x1B: False (ordered, signaling) \n
+/// 0x1C: Not-equal (ordered, signaling) \n
+/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
+/// 0x1E: Greater-than (ordered, non-signaling) \n
+/// 0x1F: True (unordered, signaling)
+/// \returns A 128-bit vector of [4 x float] containing the comparison results.
+/// \fn __m128 _mm_cmp_ps(__m128 a, __m128 b, const int c)
+
/// Compares each of the corresponding double-precision values of two
/// 256-bit vectors of [4 x double], using the operation specified by the
/// immediate integer operand.
@@ -1718,6 +1837,122 @@ _mm256_blendv_ps(__m256 __a, __m256 __b, __m256 __c)
((__m256)__builtin_ia32_cmpps256((__v8sf)(__m256)(a), \
(__v8sf)(__m256)(b), (c)))
+/* Below intrinsic defined in emmintrin.h can be used for AVX */
+/// Compares each of the corresponding scalar double-precision values of
+/// two 128-bit vectors of [2 x double], using the operation specified by the
+/// immediate integer operand.
+///
+/// If the result is true, all 64 bits of the destination vector are set;
+/// otherwise they are cleared.
+///
+/// \headerfile <x86intrin.h>
+///
+/// \code
+/// __m128d _mm_cmp_sd(__m128d a, __m128d b, const int c);
+/// \endcode
+///
+/// This intrinsic corresponds to the <c> VCMPSD </c> instruction.
+///
+/// \param a
+/// A 128-bit vector of [2 x double].
+/// \param b
+/// A 128-bit vector of [2 x double].
+/// \param c
+/// An immediate integer operand, with bits [4:0] specifying which comparison
+/// operation to use: \n
+/// 0x00: Equal (ordered, non-signaling) \n
+/// 0x01: Less-than (ordered, signaling) \n
+/// 0x02: Less-than-or-equal (ordered, signaling) \n
+/// 0x03: Unordered (non-signaling) \n
+/// 0x04: Not-equal (unordered, non-signaling) \n
+/// 0x05: Not-less-than (unordered, signaling) \n
+/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
+/// 0x07: Ordered (non-signaling) \n
+/// 0x08: Equal (unordered, non-signaling) \n
+/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
+/// 0x0A: Not-greater-than (unordered, signaling) \n
+/// 0x0B: False (ordered, non-signaling) \n
+/// 0x0C: Not-equal (ordered, non-signaling) \n
+/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
+/// 0x0E: Greater-than (ordered, signaling) \n
+/// 0x0F: True (unordered, non-signaling) \n
+/// 0x10: Equal (ordered, signaling) \n
+/// 0x11: Less-than (ordered, non-signaling) \n
+/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
+/// 0x13: Unordered (signaling) \n
+/// 0x14: Not-equal (unordered, signaling) \n
+/// 0x15: Not-less-than (unordered, non-signaling) \n
+/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
+/// 0x17: Ordered (signaling) \n
+/// 0x18: Equal (unordered, signaling) \n
+/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
+/// 0x1A: Not-greater-than (unordered, non-signaling) \n
+/// 0x1B: False (ordered, signaling) \n
+/// 0x1C: Not-equal (ordered, signaling) \n
+/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
+/// 0x1E: Greater-than (ordered, non-signaling) \n
+/// 0x1F: True (unordered, signaling)
+/// \returns A 128-bit vector of [2 x double] containing the comparison results.
+/// \fn __m128d _mm_cmp_sd(__m128d a, __m128d b, const int c)
+
+/* Below intrinsic defined in xmmintrin.h can be used for AVX */
+/// Compares each of the corresponding scalar values of two 128-bit
+/// vectors of [4 x float], using the operation specified by the immediate
+/// integer operand.
+///
+/// If the result is true, all 32 bits of the destination vector are set;
+/// otherwise they are cleared.
+///
+/// \headerfile <x86intrin.h>
+///
+/// \code
+/// __m128 _mm_cmp_ss(__m128 a, __m128 b, const int c);
+/// \endcode
+///
+/// This intrinsic corresponds to the <c> VCMPSS </c> instruction.
+///
+/// \param a
+/// A 128-bit vector of [4 x float].
+/// \param b
+/// A 128-bit vector of [4 x float].
+/// \param c
+/// An immediate integer operand, with bits [4:0] specifying which comparison
+/// operation to use: \n
+/// 0x00: Equal (ordered, non-signaling) \n
+/// 0x01: Less-than (ordered, signaling) \n
+/// 0x02: Less-than-or-equal (ordered, signaling) \n
+/// 0x03: Unordered (non-signaling) \n
+/// 0x04: Not-equal (unordered, non-signaling) \n
+/// 0x05: Not-less-than (unordered, signaling) \n
+/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
+/// 0x07: Ordered (non-signaling) \n
+/// 0x08: Equal (unordered, non-signaling) \n
+/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
+/// 0x0A: Not-greater-than (unordered, signaling) \n
+/// 0x0B: False (ordered, non-signaling) \n
+/// 0x0C: Not-equal (ordered, non-signaling) \n
+/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
+/// 0x0E: Greater-than (ordered, signaling) \n
+/// 0x0F: True (unordered, non-signaling) \n
+/// 0x10: Equal (ordered, signaling) \n
+/// 0x11: Less-than (ordered, non-signaling) \n
+/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
+/// 0x13: Unordered (signaling) \n
+/// 0x14: Not-equal (unordered, signaling) \n
+/// 0x15: Not-less-than (unordered, non-signaling) \n
+/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
+/// 0x17: Ordered (signaling) \n
+/// 0x18: Equal (unordered, signaling) \n
+/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
+/// 0x1A: Not-greater-than (unordered, non-signaling) \n
+/// 0x1B: False (ordered, signaling) \n
+/// 0x1C: Not-equal (ordered, signaling) \n
+/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
+/// 0x1E: Greater-than (ordered, non-signaling) \n
+/// 0x1F: True (unordered, signaling)
+/// \returns A 128-bit vector of [4 x float] containing the comparison results.
+/// \fn __m128 _mm_cmp_ss(__m128 a, __m128 b, const int c)
+
/// Takes a [8 x i32] vector and returns the vector element value
/// indexed by the immediate constant operand.
///
diff --git a/clang/lib/Headers/emmintrin.h b/clang/lib/Headers/emmintrin.h
index db6d43ea4ad6fe..cfd2277bbd88cc 100644
--- a/clang/lib/Headers/emmintrin.h
+++ b/clang/lib/Headers/emmintrin.h
@@ -4768,7 +4768,6 @@ static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_castsi128_pd(__m128i __a) {
/// \param c
/// An immediate integer operand, with bits [4:0] specifying which comparison
/// operation to use: \n
-/// (Note that without avx enabled, only bits [2:0] are supported) \n
/// 0x00: Equal (ordered, non-signaling) \n
/// 0x01: Less-than (ordered, signaling) \n
/// 0x02: Less-than-or-equal (ordered, signaling) \n
@@ -4777,30 +4776,6 @@ static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_castsi128_pd(__m128i __a) {
/// 0x05: Not-less-than (unordered, signaling) \n
/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
/// 0x07: Ordered (non-signaling) \n
-/// 0x08: Equal (unordered, non-signaling) \n
-/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
-/// 0x0A: Not-greater-than (unordered, signaling) \n
-/// 0x0B: False (ordered, non-signaling) \n
-/// 0x0C: Not-equal (ordered, non-signaling) \n
-/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
-/// 0x0E: Greater-than (ordered, signaling) \n
-/// 0x0F: True (unordered, non-signaling) \n
-/// 0x10: Equal (ordered, signaling) \n
-/// 0x11: Less-than (ordered, non-signaling) \n
-/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
-/// 0x13: Unordered (signaling) \n
-/// 0x14: Not-equal (unordered, signaling) \n
-/// 0x15: Not-less-than (unordered, non-signaling) \n
-/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
-/// 0x17: Ordered (signaling) \n
-/// 0x18: Equal (unordered, signaling) \n
-/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
-/// 0x1A: Not-greater-than (unordered, non-signaling) \n
-/// 0x1B: False (ordered, signaling) \n
-/// 0x1C: Not-equal (ordered, signaling) \n
-/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
-/// 0x1E: Greater-than (ordered, non-signaling) \n
-/// 0x1F: True (unordered, signaling)
/// \returns A 128-bit vector of [2 x double] containing the comparison results.
#define _mm_cmp_pd(a, b, c) \
((__m128d)__builtin_ia32_cmppd((__v2df)(__m128d)(a), (__v2df)(__m128d)(b), \
@@ -4828,7 +4803,6 @@ static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_castsi128_pd(__m128i __a) {
/// \param c
/// An immediate integer operand, with bits [4:0] specifying which comparison
/// operation to use: \n
-/// (Note that without avx enabled, only bits [2:0] are supported) \n
/// 0x00: Equal (ordered, non-signaling) \n
/// 0x01: Less-than (ordered, signaling) \n
/// 0x02: Less-than-or-equal (ordered, signaling) \n
@@ -4837,30 +4811,6 @@ static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_castsi128_pd(__m128i __a) {
/// 0x05: Not-less-than (unordered, signaling) \n
/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
/// 0x07: Ordered (non-signaling) \n
-/// 0x08: Equal (unordered, non-signaling) \n
-/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
-/// 0x0A: Not-greater-than (unordered, signaling) \n
-/// 0x0B: False (ordered, non-signaling) \n
-/// 0x0C: Not-equal (ordered, non-signaling) \n
-/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
-/// 0x0E: Greater-than (ordered, signaling) \n
-/// 0x0F: True (unordered, non-signaling) \n
-/// 0x10: Equal (ordered, signaling) \n
-/// 0x11: Less-than (ordered, non-signaling) \n
-/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
-/// 0x13: Unordered (signaling) \n
-/// 0x14: Not-equal (unordered, signaling) \n
-/// 0x15: Not-less-than (unordered, non-signaling) \n
-/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
-/// 0x17: Ordered (signaling) \n
-/// 0x18: Equal (unordered, signaling) \n
-/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
-/// 0x1A: Not-greater-than (unordered, non-signaling) \n
-/// 0x1B: False (ordered, signaling) \n
-/// 0x1C: Not-equal (ordered, signaling) \n
-/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
-/// 0x1E: Greater-than (ordered, non-signaling) \n
-/// 0x1F: True (unordered, signaling)
/// \returns A 128-bit vector of [2 x double] containing the comparison results.
#define _mm_cmp_sd(a, b, c) \
((__m128d)__builtin_ia32_cmpsd((__v2df)(__m128d)(a), (__v2df)(__m128d)(b), \
diff --git a/clang/lib/Headers/xmmintrin.h b/clang/lib/Headers/xmmintrin.h
index f9894c23ce25bf..a99381ac26207e 100644
--- a/clang/lib/Headers/xmmintrin.h
+++ b/clang/lib/Headers/xmmintrin.h
@@ -2973,7 +2973,6 @@ _mm_movemask_ps(__m128 __a)
/// \param c
/// An immediate integer operand, with bits [4:0] specifying which comparison
/// operation to use: \n
-/// (Note that without avx enabled, only bits [2:0] are supported) \n
/// 0x00: Equal (ordered, non-signaling) \n
/// 0x01: Less-than (ordered, signaling) \n
/// 0x02: Less-than-or-equal (ordered, signaling) \n
@@ -2982,30 +2981,6 @@ _mm_movemask_ps(__m128 __a)
/// 0x05: Not-less-than (unordered, signaling) \n
/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
/// 0x07: Ordered (non-signaling) \n
-/// 0x08: Equal (unordered, non-signaling) \n
-/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
-/// 0x0A: Not-greater-than (unordered, signaling) \n
-/// 0x0B: False (ordered, non-signaling) \n
-/// 0x0C: Not-equal (ordered, non-signaling) \n
-/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
-/// 0x0E: Greater-than (ordered, signaling) \n
-/// 0x0F: True (unordered, non-signaling) \n
-/// 0x10: Equal (ordered, signaling) \n
-/// 0x11: Less-than (ordered, non-signaling) \n
-/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
-/// 0x13: Unordered (signaling) \n
-/// 0x14: Not-equal (unordered, signaling) \n
-/// 0x15: Not-less-than (unordered, non-signaling) \n
-/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
-/// 0x17: Ordered (signaling) \n
-/// 0x18: Equal (unordered, signaling) \n
-/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
-/// 0x1A: Not-greater-than (unordered, non-signaling) \n
-/// 0x1B: False (ordered, signaling) \n
-/// 0x1C: Not-equal (ordered, signaling) \n
-/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
-/// 0x1E: Greater-than (ordered, non-signaling) \n
-/// 0x1F: True (unordered, signaling)
/// \returns A 128-bit vector of [4 x float] containing the comparison results.
#define _mm_cmp_ps(a, b, c) \
((__m128)__builtin_ia32_cmpps((__v4sf)(__m128)(a), (__v4sf)(__m128)(b), (c)))
@@ -3032,7 +3007,6 @@ _mm_movemask_ps(__m128 __a)
/// \param c
/// An immediate integer operand, with bits [4:0] specifying which comparison
/// operation to use: \n
-/// (Note that without avx enabled, only bits [2:0] are supported) \n
/// 0x00: Equal (ordered, non-signaling) \n
/// 0x01: Less-than (ordered, signaling) \n
/// 0x02: Less-than-or-equal (ordered, signaling) \n
@@ -3041,30 +3015,6 @@ _mm_movemask_ps(__m128 __a)
/// 0x05: Not-less-than (unordered, signaling) \n
/// 0x06: Not-less-than-or-equal (unordered, signaling) \n
/// 0x07: Ordered (non-signaling) \n
-/// 0x08: Equal (unordered, non-signaling) \n
-/// 0x09: Not-greater-than-or-equal (unordered, signaling) \n
-/// 0x0A: Not-greater-than (unordered, signaling) \n
-/// 0x0B: False (ordered, non-signaling) \n
-/// 0x0C: Not-equal (ordered, non-signaling) \n
-/// 0x0D: Greater-than-or-equal (ordered, signaling) \n
-/// 0x0E: Greater-than (ordered, signaling) \n
-/// 0x0F: True (unordered, non-signaling) \n
-/// 0x10: Equal (ordered, signaling) \n
-/// 0x11: Less-than (ordered, non-signaling) \n
-/// 0x12: Less-than-or-equal (ordered, non-signaling) \n
-/// 0x13: Unordered (signaling) \n
-/// 0x14: Not-equal (unordered, signaling) \n
-/// 0x15: Not-less-than (unordered, non-signaling) \n
-/// 0x16: Not-less-than-or-equal (unordered, non-signaling) \n
-/// 0x17: Ordered (signaling) \n
-/// 0x18: Equal (unordered, signaling) \n
-/// 0x19: Not-greater-than-or-equal (unordered, non-signaling) \n
-/// 0x1A: Not-greater-than (unordered, non-signaling) \n
-/// 0x1B: False (ordered, signaling) \n
-/// 0x1C: Not-equal (ordered, signaling) \n
-/// 0x1D: Greater-than-or-equal (ordered, non-signaling) \n
-/// 0x1E: Greater-than (ordered, non-signaling) \n
-/// 0x1F: True (unordered, signaling)
/// \returns A 128-bit vector of [4 x float] containing the comparison results.
#define _mm_cmp_ss(a, b, c) \
((__m128)__builtin_ia32_cmpss((__v4sf)(__m128)(a), (__v4sf)(__m128)(b), (c)))
>From 71641346ebadf8d21a109813fcf66f78e0932955 Mon Sep 17 00:00:00 2001
From: Freddy Ye <freddy.ye at intel.com>
Date: Sat, 9 Mar 2024 13:46:29 +0800
Subject: [PATCH 6/6] Address comments.
---
clang/lib/Headers/emmintrin.h | 7 ++-----
clang/lib/Headers/xmmintrin.h | 7 ++-----
2 files changed, 4 insertions(+), 10 deletions(-)
diff --git a/clang/lib/Headers/emmintrin.h b/clang/lib/Headers/emmintrin.h
index cfd2277bbd88cc..ebe295f160b2ae 100644
--- a/clang/lib/Headers/emmintrin.h
+++ b/clang/lib/Headers/emmintrin.h
@@ -4749,9 +4749,7 @@ static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_castsi128_pd(__m128i __a) {
/// 128-bit vectors of [2 x double], using the operation specified by the
/// immediate integer operand.
///
-/// Returns a [2 x double] vector consisting of two doubles corresponding to
-/// the two comparison results: zero if the comparison is false, and all 1's
-/// if the comparison is true.
+/// Each comparison yields 0x0 for false, 0xFFFFFFFFFFFFFFFF for true.
///
/// \headerfile <x86intrin.h>
///
@@ -4785,8 +4783,7 @@ static __inline__ __m128d __DEFAULT_FN_ATTRS _mm_castsi128_pd(__m128i __a) {
/// two 128-bit vectors of [2 x double], using the operation specified by the
/// immediate integer operand.
///
-/// If the result is true, all 64 bits of the destination vector are set;
-/// otherwise they are cleared.
+/// Each comparison yields 0x0 for false, 0xFFFFFFFFFFFFFFFF for true.
///
/// \headerfile <x86intrin.h>
///
diff --git a/clang/lib/Headers/xmmintrin.h b/clang/lib/Headers/xmmintrin.h
index a99381ac26207e..1f5993e0c368d9 100644
--- a/clang/lib/Headers/xmmintrin.h
+++ b/clang/lib/Headers/xmmintrin.h
@@ -2954,9 +2954,7 @@ _mm_movemask_ps(__m128 __a)
/// [4 x float], using the operation specified by the immediate integer
/// operand.
///
-/// Returns a [4 x float] vector consisting of four floats corresponding to
-/// the four comparison results: zero if the comparison is false, and all 1's
-/// if the comparison is true.
+/// Each comparison yields 0x0 for false, 0xFFFFFFFF for true.
///
/// \headerfile <x86intrin.h>
///
@@ -2989,8 +2987,7 @@ _mm_movemask_ps(__m128 __a)
/// vectors of [4 x float], using the operation specified by the immediate
/// integer operand.
///
-/// If the result is true, all 32 bits of the destination vector are set;
-/// otherwise they are cleared.
+/// Each comparison yields 0x0 for false, 0xFFFFFFFF for true.
///
/// \headerfile <x86intrin.h>
///
More information about the cfe-commits
mailing list