Author: David Green Date: 2022-11-03T16:06:35Z New Revision: e7deca525058778df15e7888ed24974a32c8686c
URL: https://github.com/llvm/llvm-project/commit/e7deca525058778df15e7888ed24974a32c8686c DIFF: https://github.com/llvm/llvm-project/commit/e7deca525058778df15e7888ed24974a32c8686c.diff LOG: [AArch64] Alter arm_fp16.h to be target-based, not preprocessor based. As the other recent patches, this alters the arm_fp16 intrinsics to be target based, not preprocessor based. Apparently arm_fp16.h is AArch64 only under clang, making this mostly trivial with the TargetGuard infrastructure. Differential Revision: https://reviews.llvm.org/D137256 Added: Modified: clang/include/clang/Basic/arm_fp16.td clang/test/Sema/aarch64-fp16-target.c Removed: ################################################################################ diff --git a/clang/include/clang/Basic/arm_fp16.td b/clang/include/clang/Basic/arm_fp16.td index 79cd16233c10..cb2a09303e8e 100644 --- a/clang/include/clang/Basic/arm_fp16.td +++ b/clang/include/clang/Basic/arm_fp16.td @@ -14,7 +14,7 @@ include "arm_neon_incl.td" // ARMv8.2-A FP16 intrinsics. -let ArchGuard = "defined(__ARM_FEATURE_FP16_SCALAR_ARITHMETIC) && defined(__aarch64__)" in { +let ArchGuard = "defined(__aarch64__)", TargetGuard = "fullfp16" in { // Negate def VNEGSH : SInst<"vneg", "11", "Sh">; diff --git a/clang/test/Sema/aarch64-fp16-target.c b/clang/test/Sema/aarch64-fp16-target.c index 13cee64c52f5..9a921e96e88e 100644 --- a/clang/test/Sema/aarch64-fp16-target.c +++ b/clang/test/Sema/aarch64-fp16-target.c @@ -7,19 +7,19 @@ __attribute__((target("fullfp16"))) void test_fullfp16(float16_t f16) { - vabdh_f16(f16, f16); // expected-error {{call to undeclared function 'vabdh_f16'}} + vabdh_f16(f16, f16); } __attribute__((target("arch=armv8-a+fp16"))) void test_fp16_arch(float16_t f16) { - vabdh_f16(f16, f16); // expected-error {{call to undeclared function 'vabdh_f16'}} + vabdh_f16(f16, f16); } __attribute__((target("+fp16"))) void test_fp16(float16_t f16) { - vabdh_f16(f16, f16); // expected-error {{call to undeclared function 'vabdh_f16'}} + vabdh_f16(f16, f16); } void undefined(float16_t f16) { - vabdh_f16(f16, f16); // expected-error {{call to undeclared function 'vabdh_f16'}} + vabdh_f16(f16, f16); // expected-error {{'__builtin_neon_vabdh_f16' needs target feature fullfp16}} } _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
