Skip to content

[AArch64][Clang][NEON] Remove undefined vcmla intrinsics #112575

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Oct 17, 2024
Merged

[AArch64][Clang][NEON] Remove undefined vcmla intrinsics #112575

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 5 additions & 2 deletions clang/include/clang/Basic/arm_neon.td
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1968,13 +1968,16 @@ let TargetGuard = "v8.3a,neon" in {
def VCADDQ_ROT90 : SInst<"vcaddq_rot90", "QQQ", "f">;
def VCADDQ_ROT270 : SInst<"vcaddq_rot270", "QQQ", "f">;

defm VCMLA_F32 : VCMLA_ROTS<"f", "uint64x1_t", "uint64x2_t">;
defm VCMLA_F32 : VCMLA_ROTS<"f", "uint64x1_t", "uint64x2_t">;
}
let ArchGuard = "defined(__aarch64__) || defined(__arm64ec__)", TargetGuard = "v8.3a,neon" in {
def VCADDQ_ROT90_FP64 : SInst<"vcaddq_rot90", "QQQ", "d">;
def VCADDQ_ROT270_FP64 : SInst<"vcaddq_rot270", "QQQ", "d">;

defm VCMLA_FP64 : VCMLA_ROTS<"d", "uint64x2_t", "uint64x2_t">;
def VCMLAQ_FP64 : SInst<"vcmlaq", "QQQQ", "d">;
def VCMLAQ_ROT90_FP64 : SInst<"vcmlaq_rot90", "QQQQ", "d">;
def VCMLAQ_ROT180_FP64 : SInst<"vcmlaq_rot180", "QQQQ", "d">;
def VCMLAQ_ROT270_FP64 : SInst<"vcmlaq_rot270", "QQQQ", "d">;
}

// V8.2-A BFloat intrinsics
Expand Down
4 changes: 2 additions & 2 deletions clang/test/Sema/aarch64-neon-target.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -58,7 +58,7 @@ __attribute__((target("arch=armv8.3-a+fp16")))
void test_v83(float32x4_t v4f32, float16x4_t v4f16, float64x2_t v2f64) {
vcaddq_rot90_f32(v4f32, v4f32);
vcmla_rot90_f16(v4f16, v4f16, v4f16);
vcmlaq_rot270_laneq_f64(v2f64, v2f64, v2f64, 1);
vcmlaq_rot270_f64(v2f64, v2f64, v2f64);
}

__attribute__((target("arch=armv8.5-a")))
Expand Down Expand Up @@ -95,7 +95,7 @@ void undefined(uint32x2_t v2i32, uint32x4_t v4i32, uint16x8_t v8i16, uint8x16_t
// 8.3 - complex
vcaddq_rot90_f32(v4f32, v4f32); // expected-error {{always_inline function 'vcaddq_rot90_f32' requires target feature 'v8.3a'}}
vcmla_rot90_f16(v4f16, v4f16, v4f16); // expected-error {{always_inline function 'vcmla_rot90_f16' requires target feature 'v8.3a'}}
vcmlaq_rot270_laneq_f64(v2f64, v2f64, v2f64, 1); // expected-error {{always_inline function 'vcmlaq_rot270_f64' requires target feature 'v8.3a'}}
vcmlaq_rot270_f64(v2f64, v2f64, v2f64); // expected-error {{always_inline function 'vcmlaq_rot270_f64' requires target feature 'v8.3a'}}
// 8.5 - frint
vrnd32xq_f32(v4f32); // expected-error {{always_inline function 'vrnd32xq_f32' requires target feature 'v8.5a'}}

Expand Down
31 changes: 31 additions & 0 deletions clang/test/Sema/aarch64-vcmla-undef.c
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,31 @@
// RUN: %clang_cc1 -triple aarch64-linux-gnu -target-feature +neon -target-feature +v8.3a -ffreestanding -fsyntax-only -verify -verify-ignore-unexpected=note %s

// REQUIRES: aarch64-registered-target

#include <arm_neon.h>

void test(float64x1_t v1f64, float64x2_t v2f64) {
vcmla_f64(v1f64, v1f64, v1f64); // expected-error {{call to undeclared function 'vcmla_f64'}}
vcmla_lane_f64(v1f64, v1f64, v1f64, 0); // expected-error {{call to undeclared function 'vcmla_lane_f64'}}
vcmla_laneq_f64(v1f64, v1f64, v2f64, 0); // expected-error {{call to undeclared function 'vcmla_laneq_f64'}}
vcmlaq_lane_f64(v2f64, v2f64, v1f64, 0); // expected-error {{call to undeclared function 'vcmlaq_lane_f64'}}
vcmlaq_laneq_f64(v2f64, v2f64, v2f64, 0); // expected-error {{call to undeclared function 'vcmlaq_laneq_f64'}}

vcmla_rot90_f64(v1f64, v1f64, v1f64); // expected-error {{call to undeclared function 'vcmla_rot90_f64'}}
vcmla_rot90_lane_f64(v1f64, v1f64, v1f64, 0); // expected-error {{call to undeclared function 'vcmla_rot90_lane_f64'}}
vcmla_rot90_laneq_f64(v1f64, v1f64, v2f64, 0); // expected-error {{call to undeclared function 'vcmla_rot90_laneq_f64'}}
vcmlaq_rot90_lane_f64(v2f64, v2f64, v1f64, 0); // expected-error {{call to undeclared function 'vcmlaq_rot90_lane_f64'}}
vcmlaq_rot90_laneq_f64(v2f64, v2f64, v2f64, 0); // expected-error {{call to undeclared function 'vcmlaq_rot90_laneq_f64'}}

vcmla_rot180_f64(v1f64, v1f64, v1f64); // expected-error {{call to undeclared function 'vcmla_rot180_f64'}}
vcmla_rot180_lane_f64(v1f64, v1f64, v1f64, 0); // expected-error {{call to undeclared function 'vcmla_rot180_lane_f64'}}
vcmla_rot180_laneq_f64(v1f64, v1f64, v2f64, 0); // expected-error {{call to undeclared function 'vcmla_rot180_laneq_f64'}}
vcmlaq_rot180_lane_f64(v2f64, v2f64, v1f64, 0); // expected-error {{call to undeclared function 'vcmlaq_rot180_lane_f64'}}
vcmlaq_rot180_laneq_f64(v2f64, v2f64, v2f64, 0); // expected-error {{call to undeclared function 'vcmlaq_rot180_laneq_f64'}}

vcmla_rot270_f64(v1f64, v1f64, v1f64); // expected-error {{call to undeclared function 'vcmla_rot270_f64'}}
vcmla_rot270_lane_f64(v1f64, v1f64, v1f64, 0); // expected-error {{call to undeclared function 'vcmla_rot270_lane_f64'}}
vcmla_rot270_laneq_f64(v1f64, v1f64, v2f64, 0); // expected-error {{call to undeclared function 'vcmla_rot270_laneq_f64'}}
vcmlaq_rot270_lane_f64(v2f64, v2f64, v1f64, 0); // expected-error {{call to undeclared function 'vcmlaq_rot270_lane_f64'}}
vcmlaq_rot270_laneq_f64(v1f64, v1f64, v2f64, 0); // expected-error {{call to undeclared function 'vcmlaq_rot270_laneq_f64'}}
}
Loading