+2015-09-08 Alan Lawrence <alan.lawrence@arm.com>
+
+ * gcc.target/aarch64/advsimd-intrinsics/advsimd-intrinsics.exp:
+ Set additional_flags for neon-fp16 if supported, else fallback to neon.
+
+ * gcc.target/aarch64/advsimd-intrinsics/arm-neon-ref.h
+ (hfloat16_t): New.
+ (result, expected, clean_results, DECL_VARIABLE_64BITS_VARIANTS,
+ DECL_VARIABLE_128BITS_VARIANTS): Add float16x4_t and float16x8_t cases
+ if supported.
+ (CHECK_RESULTS): Redefine using CHECK_RESULTS_NAMED.
+ (CHECK_RESULTS_NAMED): Move body to CHECK_RESULTS_NAMED_NO_FP16;
+ redefine in terms of CHECK_RESULTS_NAMED_NO_FP16 with float16 variants
+ when those are supported.
+ (CHECK_RESULTS_NAMED_NO_FP16, CHECK_RESULTS_NO_FP16): New.
+ (vdup_n_f16): New.
+
+ * gcc.target/aarch64/advsimd-intrinsics/compute-ref-data.h (buffer,
+ buffer_pad, buffer_dup, buffer_dup_pad): Add float16x4 and float16x8_t
+ cases if supported.
+
+ * gcc.target/aarch64/advsimd-intrinsics/vbsl.c (exec_vbsl):
+ Use CHECK_RESULTS_NO_FP16 in place of CHECK_RESULTS.
+ * gcc.target/aarch64/advsimd-intrinsics/vdup-vmov.c (exec_vdup_vmov):
+ Likewise.
+ * gcc.target/aarch64/advsimd-intrinsics/vdup_lane.c (exec_vdup_lane):
+ Likewise.
+ * gcc.target/aarch64/advsimd-intrinsics/vext.c (exec_vext): Likewise.
+
+ * gcc.target/aarch64/advsimd-intrinsics/vcombine.c (expected):
+ Add float16x8_t case.
+ (main, exec_vcombine): test float16x4_t -> float16x8_t, if supported.
+ * gcc.target/aarch64/advsimd-intrinsics/vcreate.c (expected,
+ main, exec_vcreate): Likewise.
+ * gcc.target/aarch64/advsimd-intrinsics/vget_high (expected,
+ exec_vget_high): Likewise.
+ * gcc.target/aarch64/advsimd-intrinsics/vget_low.c (expected,
+ exec_vget_low): Likewise.
+ * gcc.target/aarch64/advsimd-intrinsics/vld1.c (expected, exec_vld1):
+ Likewise.
+ * gcc.target/aarch64/advsimd-intrinsics/vld1_dup.c (expected,
+ exec_vld1_dup): Likewise.
+ * gcc.target/aarch64/advsimd-intrinsics/vld1_lane.c (expected,
+ exec_vld1_lane): Likewise.
+ * gcc.target/aarch64/advsimd-intrinsics/vldX.c (expected, exec_vldX):
+ Likewise.
+ * gcc.target/aarch64/advsimd-intrinsics/vldX_dup.c (expected,
+ exec_vldX_dup): Likewise.
+ * gcc.target/aarch64/advsimd-intrinsics/vldX_lane.c (expected,
+ exec_vldX_lane): Likewise.
+ * gcc.target/aarch64/advsimd-intrinsics/vset_lane.c (expected,
+ exec_vset_lane): Likewise.
+ * gcc.target/aarch64/advsimd-intrinsics/vst1_lane.c (expected,
+ exec_vst1_lane): Likewise.
+
2015-09-08 Alan Lawrence <alan.lawrence@arm.com>
* gcc.target/aarch64/vget_high_1.c: Add float16x8->float16x4 case.
torture-init
set-torture-options $C_TORTURE_OPTIONS {{}} $LTO_TORTURE_OPTIONS
-# Make sure Neon flags are provided, if necessary.
-set additional_flags [add_options_for_arm_neon ""]
+# Make sure Neon flags are provided, if necessary. Use fp16 if we can.
+if {[check_effective_target_arm_neon_fp16_ok]} then {
+ set additional_flags [add_options_for_arm_neon_fp16 ""]
+} else {
+ set additional_flags [add_options_for_arm_neon ""]
+}
# Main loop.
gcc-dg-runtest [lsort [glob -nocomplain $srcdir/$subdir/*.c]] \
#include <inttypes.h>
/* helper type, to help write floating point results in integer form. */
+typedef uint16_t hfloat16_t;
typedef uint32_t hfloat32_t;
typedef uint64_t hfloat64_t;
static ARRAY(result, uint, 64, 1);
static ARRAY(result, poly, 8, 8);
static ARRAY(result, poly, 16, 4);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+static ARRAY(result, float, 16, 4);
+#endif
static ARRAY(result, float, 32, 2);
static ARRAY(result, int, 8, 16);
static ARRAY(result, int, 16, 8);
static ARRAY(result, uint, 64, 2);
static ARRAY(result, poly, 8, 16);
static ARRAY(result, poly, 16, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+static ARRAY(result, float, 16, 8);
+#endif
static ARRAY(result, float, 32, 4);
#ifdef __aarch64__
static ARRAY(result, float, 64, 2);
extern ARRAY(expected, uint, 64, 1);
extern ARRAY(expected, poly, 8, 8);
extern ARRAY(expected, poly, 16, 4);
+extern ARRAY(expected, hfloat, 16, 4);
extern ARRAY(expected, hfloat, 32, 2);
extern ARRAY(expected, int, 8, 16);
extern ARRAY(expected, int, 16, 8);
extern ARRAY(expected, uint, 64, 2);
extern ARRAY(expected, poly, 8, 16);
extern ARRAY(expected, poly, 16, 8);
+extern ARRAY(expected, hfloat, 16, 8);
extern ARRAY(expected, hfloat, 32, 4);
extern ARRAY(expected, hfloat, 64, 2);
-/* Check results. Operates on all possible vector types. */
-#define CHECK_RESULTS(test_name,comment) \
- { \
- CHECK(test_name, int, 8, 8, PRIx8, expected, comment); \
- CHECK(test_name, int, 16, 4, PRIx16, expected, comment); \
- CHECK(test_name, int, 32, 2, PRIx32, expected, comment); \
- CHECK(test_name, int, 64, 1, PRIx64, expected, comment); \
- CHECK(test_name, uint, 8, 8, PRIx8, expected, comment); \
- CHECK(test_name, uint, 16, 4, PRIx16, expected, comment); \
- CHECK(test_name, uint, 32, 2, PRIx32, expected, comment); \
- CHECK(test_name, uint, 64, 1, PRIx64, expected, comment); \
- CHECK(test_name, poly, 8, 8, PRIx8, expected, comment); \
- CHECK(test_name, poly, 16, 4, PRIx16, expected, comment); \
- CHECK_FP(test_name, float, 32, 2, PRIx32, expected, comment); \
- \
- CHECK(test_name, int, 8, 16, PRIx8, expected, comment); \
- CHECK(test_name, int, 16, 8, PRIx16, expected, comment); \
- CHECK(test_name, int, 32, 4, PRIx32, expected, comment); \
- CHECK(test_name, int, 64, 2, PRIx64, expected, comment); \
- CHECK(test_name, uint, 8, 16, PRIx8, expected, comment); \
- CHECK(test_name, uint, 16, 8, PRIx16, expected, comment); \
- CHECK(test_name, uint, 32, 4, PRIx32, expected, comment); \
- CHECK(test_name, uint, 64, 2, PRIx64, expected, comment); \
- CHECK(test_name, poly, 8, 16, PRIx8, expected, comment); \
- CHECK(test_name, poly, 16, 8, PRIx16, expected, comment); \
- CHECK_FP(test_name, float, 32, 4, PRIx32, expected, comment); \
- } \
-
-#define CHECK_RESULTS_NAMED(test_name,EXPECTED,comment) \
+#define CHECK_RESULTS_NAMED_NO_FP16(test_name,EXPECTED,comment) \
{ \
CHECK(test_name, int, 8, 8, PRIx8, EXPECTED, comment); \
CHECK(test_name, int, 16, 4, PRIx16, EXPECTED, comment); \
CHECK_FP(test_name, float, 32, 4, PRIx32, EXPECTED, comment); \
} \
+/* Check results against EXPECTED. Operates on all possible vector types. */
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define CHECK_RESULTS_NAMED(test_name,EXPECTED,comment) \
+ { \
+ CHECK_RESULTS_NAMED_NO_FP16(test_name, EXPECTED, comment) \
+ CHECK_FP(test_name, float, 16, 4, PRIx16, EXPECTED, comment); \
+ CHECK_FP(test_name, float, 16, 8, PRIx16, EXPECTED, comment); \
+ }
+#else
+#define CHECK_RESULTS_NAMED(test_name,EXPECTED,comment) \
+ CHECK_RESULTS_NAMED_NO_FP16(test_name, EXPECTED, comment)
+#endif
+
+#define CHECK_RESULTS_NO_FP16(test_name,comment) \
+ CHECK_RESULTS_NAMED_NO_FP16(test_name, expected, comment)
+
+#define CHECK_RESULTS(test_name,comment) \
+ CHECK_RESULTS_NAMED(test_name, expected, comment)
#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
CLEAN(result, uint, 64, 1);
CLEAN(result, poly, 8, 8);
CLEAN(result, poly, 16, 4);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ CLEAN(result, float, 16, 4);
+#endif
CLEAN(result, float, 32, 2);
CLEAN(result, int, 8, 16);
CLEAN(result, uint, 64, 2);
CLEAN(result, poly, 8, 16);
CLEAN(result, poly, 16, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ CLEAN(result, float, 16, 8);
+#endif
CLEAN(result, float, 32, 4);
#if defined(__aarch64__)
DECL_VARIABLE(VAR, uint, 64, 2)
/* Declare all 64 bits variants. */
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define DECL_VARIABLE_64BITS_VARIANTS(VAR) \
+ DECL_VARIABLE_64BITS_SIGNED_VARIANTS(VAR); \
+ DECL_VARIABLE_64BITS_UNSIGNED_VARIANTS(VAR); \
+ DECL_VARIABLE(VAR, poly, 8, 8); \
+ DECL_VARIABLE(VAR, poly, 16, 4); \
+ DECL_VARIABLE(VAR, float, 16, 4); \
+ DECL_VARIABLE(VAR, float, 32, 2)
+#else
#define DECL_VARIABLE_64BITS_VARIANTS(VAR) \
DECL_VARIABLE_64BITS_SIGNED_VARIANTS(VAR); \
DECL_VARIABLE_64BITS_UNSIGNED_VARIANTS(VAR); \
DECL_VARIABLE(VAR, poly, 8, 8); \
DECL_VARIABLE(VAR, poly, 16, 4); \
DECL_VARIABLE(VAR, float, 32, 2)
+#endif
/* Declare all 128 bits variants. */
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
#define DECL_VARIABLE_128BITS_VARIANTS(VAR) \
DECL_VARIABLE_128BITS_SIGNED_VARIANTS(VAR); \
DECL_VARIABLE_128BITS_UNSIGNED_VARIANTS(VAR); \
DECL_VARIABLE(VAR, poly, 8, 16); \
DECL_VARIABLE(VAR, poly, 16, 8); \
+ DECL_VARIABLE(VAR, float, 16, 8); \
DECL_VARIABLE(VAR, float, 32, 4)
-
+#else
+#define DECL_VARIABLE_128BITS_VARIANTS(VAR) \
+ DECL_VARIABLE_128BITS_SIGNED_VARIANTS(VAR); \
+ DECL_VARIABLE_128BITS_UNSIGNED_VARIANTS(VAR); \
+ DECL_VARIABLE(VAR, poly, 8, 16); \
+ DECL_VARIABLE(VAR, poly, 16, 8); \
+ DECL_VARIABLE(VAR, float, 32, 4)
+#endif
/* Declare all variants. */
#define DECL_VARIABLE_ALL_VARIANTS(VAR) \
DECL_VARIABLE_64BITS_VARIANTS(VAR); \
/* Helpers to initialize vectors. */
#define VDUP(VAR, Q, T1, T2, W, N, V) \
VECT_VAR(VAR, T1, W, N) = vdup##Q##_n_##T2##W(V)
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+/* Work around that there is no vdup_n_f16 intrinsic. */
+#define vdup_n_f16(VAL) \
+ __extension__ \
+ ({ \
+ float16_t f = VAL; \
+ vld1_dup_f16(&f); \
+ })
+#endif
#define VSET_LANE(VAR, Q, T1, T2, W, N, L, V) \
VECT_VAR(VAR, T1, W, N) = vset##Q##_lane_##T2##W(V, \
PAD(buffer_pad, uint, 32, 2);
VECT_VAR_DECL_INIT(buffer, uint, 64, 1);
PAD(buffer_pad, uint, 64, 1);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+VECT_VAR_DECL_INIT(buffer, float, 16, 4);
+PAD(buffer_pad, float, 16, 4);
+#endif
VECT_VAR_DECL_INIT(buffer, float, 32, 2);
PAD(buffer_pad, float, 32, 2);
VECT_VAR_DECL_INIT(buffer, int, 8, 16);
PAD(buffer_pad, poly, 8, 16);
VECT_VAR_DECL_INIT(buffer, poly, 16, 8);
PAD(buffer_pad, poly, 16, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+VECT_VAR_DECL_INIT(buffer, float, 16, 8);
+PAD(buffer_pad, float, 16, 8);
+#endif
VECT_VAR_DECL_INIT(buffer, float, 32, 4);
PAD(buffer_pad, float, 32, 4);
#ifdef __aarch64__
VECT_VAR_DECL(buffer_dup_pad, poly, 8, 8);
VECT_VAR_DECL_INIT(buffer_dup, poly, 16, 4);
VECT_VAR_DECL(buffer_dup_pad, poly, 16, 4);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+VECT_VAR_DECL_INIT4(buffer_dup, float, 16, 4);
+VECT_VAR_DECL(buffer_dup_pad, float, 16, 4);
+#endif
VECT_VAR_DECL_INIT4(buffer_dup, float, 32, 2);
VECT_VAR_DECL(buffer_dup_pad, float, 32, 2);
VECT_VAR_DECL(buffer_dup_pad, poly, 8, 16);
VECT_VAR_DECL_INIT(buffer_dup, poly, 16, 8);
VECT_VAR_DECL(buffer_dup_pad, poly, 16, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+VECT_VAR_DECL_INIT(buffer_dup, float, 16, 8);
+VECT_VAR_DECL(buffer_dup_pad, float, 16, 8);
+#endif
VECT_VAR_DECL_INIT(buffer_dup, float, 32, 4);
VECT_VAR_DECL(buffer_dup_pad, float, 32, 4);
TEST_VBSL(uint, , float, f, 32, 2);
TEST_VBSL(uint, q, float, f, 32, 4);
- CHECK_RESULTS (TEST_MSG, "");
+ CHECK_RESULTS_NO_FP16 (TEST_MSG, "");
}
int main (void)
0x66, 0x66, 0x66, 0x66 };
VECT_VAR_DECL(expected,hfloat,32,4) [] = { 0xc1800000, 0xc1700000,
0x40533333, 0x40533333 };
+VECT_VAR_DECL(expected,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80,
+ 0x4080, 0x4080, 0x4080, 0x4080 };
#define TEST_MSG "VCOMBINE"
void exec_vcombine (void)
/* Initialize input "vector64_a" from "buffer". */
TEST_MACRO_64BITS_VARIANTS_2_5(VLOAD, vector64_a, buffer);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VLOAD(vector64_a, buffer, , float, f, 16, 4);
+#endif
VLOAD(vector64_a, buffer, , float, f, 32, 2);
/* Choose init value arbitrarily. */
VDUP(vector64_b, , uint, u, 64, 1, 0x88);
VDUP(vector64_b, , poly, p, 8, 8, 0x55);
VDUP(vector64_b, , poly, p, 16, 4, 0x66);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VDUP(vector64_b, , float, f, 16, 4, 2.25);
+#endif
VDUP(vector64_b, , float, f, 32, 2, 3.3f);
clean_results ();
TEST_VCOMBINE(uint, u, 64, 1, 2);
TEST_VCOMBINE(poly, p, 8, 8, 16);
TEST_VCOMBINE(poly, p, 16, 4, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ TEST_VCOMBINE(float, f, 16, 4, 8);
+#endif
TEST_VCOMBINE(float, f, 32, 2, 4);
CHECK(TEST_MSG, int, 8, 16, PRIx8, expected, "");
CHECK(TEST_MSG, uint, 64, 2, PRIx64, expected, "");
CHECK(TEST_MSG, poly, 8, 16, PRIx8, expected, "");
CHECK(TEST_MSG, poly, 16, 8, PRIx16, expected, "");
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ CHECK_FP(TEST_MSG, float, 16, 8, PRIx16, expected, "");
+#endif
CHECK_FP(TEST_MSG, float, 32, 4, PRIx32, expected, "");
}
VECT_VAR_DECL(expected,poly,8,8) [] = { 0xf0, 0xde, 0xbc, 0x9a,
0x78, 0x56, 0x34, 0x12 };
VECT_VAR_DECL(expected,poly,16,4) [] = { 0xdef0, 0x9abc, 0x5678, 0x1234 };
+VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xdef0, 0x9abc, 0x5678, 0x1234 };
VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0x9abcdef0, 0x12345678 };
#define INSN_NAME vcreate
DECL_VAL(val, int, 16, 4);
DECL_VAL(val, int, 32, 2);
DECL_VAL(val, int, 64, 1);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ DECL_VAL(val, float, 16, 4);
+#endif
DECL_VAL(val, float, 32, 2);
DECL_VAL(val, uint, 8, 8);
DECL_VAL(val, uint, 16, 4);
DECL_VARIABLE(vector_res, int, 16, 4);
DECL_VARIABLE(vector_res, int, 32, 2);
DECL_VARIABLE(vector_res, int, 64, 1);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ DECL_VARIABLE(vector_res, float, 16, 4);
+#endif
DECL_VARIABLE(vector_res, float, 32, 2);
DECL_VARIABLE(vector_res, uint, 8, 8);
DECL_VARIABLE(vector_res, uint, 16, 4);
VECT_VAR(val, int, 16, 4) = 0x123456789abcdef0LL;
VECT_VAR(val, int, 32, 2) = 0x123456789abcdef0LL;
VECT_VAR(val, int, 64, 1) = 0x123456789abcdef0LL;
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VECT_VAR(val, float, 16, 4) = 0x123456789abcdef0LL;
+#endif
VECT_VAR(val, float, 32, 2) = 0x123456789abcdef0LL;
VECT_VAR(val, uint, 8, 8) = 0x123456789abcdef0ULL;
VECT_VAR(val, uint, 16, 4) = 0x123456789abcdef0ULL;
TEST_VCREATE(int, s, 8, 8);
TEST_VCREATE(int, s, 16, 4);
TEST_VCREATE(int, s, 32, 2);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ TEST_VCREATE(float, f, 16, 4);
+#endif
TEST_VCREATE(float, f, 32, 2);
TEST_VCREATE(int, s, 64, 1);
TEST_VCREATE(uint, u, 8, 8);
CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected, "");
CHECK(TEST_MSG, poly, 8, 8, PRIx8, expected, "");
CHECK(TEST_MSG, poly, 16, 4, PRIx16, expected, "");
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ CHECK_FP(TEST_MSG, float, 16, 4, PRIx16, expected, "");
+#endif
CHECK_FP(TEST_MSG, float, 32, 2, PRIx32, expected, "");
}
switch (i) {
case 0:
- CHECK_RESULTS_NAMED (TEST_MSG, expected0, "");
+ CHECK_RESULTS_NAMED_NO_FP16 (TEST_MSG, expected0, "");
break;
case 1:
- CHECK_RESULTS_NAMED (TEST_MSG, expected1, "");
+ CHECK_RESULTS_NAMED_NO_FP16 (TEST_MSG, expected1, "");
break;
case 2:
- CHECK_RESULTS_NAMED (TEST_MSG, expected2, "");
+ CHECK_RESULTS_NAMED_NO_FP16 (TEST_MSG, expected2, "");
break;
default:
abort();
switch (i) {
case 0:
- CHECK_RESULTS_NAMED (TEST_MSG, expected0, "");
+ CHECK_RESULTS_NAMED_NO_FP16 (TEST_MSG, expected0, "");
break;
case 1:
- CHECK_RESULTS_NAMED (TEST_MSG, expected1, "");
+ CHECK_RESULTS_NAMED_NO_FP16 (TEST_MSG, expected1, "");
break;
case 2:
- CHECK_RESULTS_NAMED (TEST_MSG, expected2, "");
+ CHECK_RESULTS_NAMED_NO_FP16 (TEST_MSG, expected2, "");
break;
default:
abort();
TEST_VDUP_LANE(q, poly, p, 16, 8, 4, 1);
TEST_VDUP_LANE(q, float, f, 32, 4, 2, 1);
- CHECK_RESULTS (TEST_MSG, "");
+ CHECK_RESULTS_NO_FP16 (TEST_MSG, "");
}
int main (void)
TEST_VEXT(q, poly, p, 16, 8, 6);
TEST_VEXT(q, float, f, 32, 4, 3);
- CHECK_RESULTS (TEST_MSG, "");
+ CHECK_RESULTS_NO_FP16 (TEST_MSG, "");
}
int main (void)
VECT_VAR_DECL(expected,poly,8,8) [] = { 0xf8, 0xf9, 0xfa, 0xfb,
0xfc, 0xfd, 0xfe, 0xff };
VECT_VAR_DECL(expected,poly,16,4) [] = { 0xfff4, 0xfff5, 0xfff6, 0xfff7 };
+VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xca00, 0xc980, 0xc900, 0xc880 };
VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 };
#define TEST_MSG "VGET_HIGH"
DECL_VARIABLE_128BITS_VARIANTS(vector128);
TEST_MACRO_128BITS_VARIANTS_2_5(VLOAD, vector128, buffer);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VLOAD(vector128, buffer, q, float, f, 16, 8);
+#endif
VLOAD(vector128, buffer, q, float, f, 32, 4);
clean_results ();
TEST_VGET_HIGH(uint, u, 64, 1, 2);
TEST_VGET_HIGH(poly, p, 8, 8, 16);
TEST_VGET_HIGH(poly, p, 16, 4, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ TEST_VGET_HIGH(float, f, 16, 4, 8);
+#endif
TEST_VGET_HIGH(float, f, 32, 2, 4);
CHECK(TEST_MSG, int, 8, 8, PRIx8, expected, "");
VECT_VAR_DECL(expected,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf4, 0xf5, 0xf6, 0xf7 };
VECT_VAR_DECL(expected,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 };
+VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 };
VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 };
#define TEST_MSG "VGET_LOW"
DECL_VARIABLE_128BITS_VARIANTS(vector128);
TEST_MACRO_128BITS_VARIANTS_2_5(VLOAD, vector128, buffer);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VLOAD(vector128, buffer, q, float, f, 16, 8);
+#endif
VLOAD(vector128, buffer, q, float, f, 32, 4);
clean_results ();
TEST_VGET_LOW(uint, u, 64, 1, 2);
TEST_VGET_LOW(poly, p, 8, 8, 16);
TEST_VGET_LOW(poly, p, 16, 4, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ TEST_VGET_LOW(float, f, 16, 4, 8);
+#endif
TEST_VGET_LOW(float, f, 32, 2, 4);
CHECK(TEST_MSG, int, 8, 8, PRIx8, expected, "");
CHECK(TEST_MSG, uint, 64, 1, PRIx64, expected, "");
CHECK(TEST_MSG, poly, 8, 8, PRIx8, expected, "");
CHECK(TEST_MSG, poly, 16, 4, PRIx16, expected, "");
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ CHECK_FP(TEST_MSG, float, 16, 4, PRIx16, expected, "");
+#endif
CHECK_FP(TEST_MSG, float, 32, 2, PRIx32, expected, "");
}
VECT_VAR_DECL(expected,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf4, 0xf5, 0xf6, 0xf7 };
VECT_VAR_DECL(expected,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 };
+VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 };
VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 };
VECT_VAR_DECL(expected,int,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf4, 0xf5, 0xf6, 0xf7,
0xfc, 0xfd, 0xfe, 0xff };
VECT_VAR_DECL(expected,poly,16,8) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3,
0xfff4, 0xfff5, 0xfff6, 0xfff7 };
+VECT_VAR_DECL(expected,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80,
+ 0xca00, 0xc980, 0xc900, 0xc880 };
VECT_VAR_DECL(expected,hfloat,32,4) [] = { 0xc1800000, 0xc1700000,
0xc1600000, 0xc1500000 };
TEST_MACRO_ALL_VARIANTS_2_5(TEST_VLD1, vector, buffer);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ TEST_VLD1(vector, buffer, , float, f, 16, 4);
+ TEST_VLD1(vector, buffer, q, float, f, 16, 8);
+#endif
TEST_VLD1(vector, buffer, , float, f, 32, 2);
TEST_VLD1(vector, buffer, q, float, f, 32, 4);
VECT_VAR_DECL(expected0,poly,8,8) [] = { 0xf0, 0xf0, 0xf0, 0xf0,
0xf0, 0xf0, 0xf0, 0xf0 };
VECT_VAR_DECL(expected0,poly,16,4) [] = { 0xfff0, 0xfff0, 0xfff0, 0xfff0 };
+VECT_VAR_DECL(expected0,hfloat,16,4) [] = { 0xcc00, 0xcc00, 0xcc00, 0xcc00 };
VECT_VAR_DECL(expected0,hfloat,32,2) [] = { 0xc1800000, 0xc1800000 };
VECT_VAR_DECL(expected0,int,8,16) [] = { 0xf0, 0xf0, 0xf0, 0xf0,
0xf0, 0xf0, 0xf0, 0xf0,
0xf0, 0xf0, 0xf0, 0xf0 };
VECT_VAR_DECL(expected0,poly,16,8) [] = { 0xfff0, 0xfff0, 0xfff0, 0xfff0,
0xfff0, 0xfff0, 0xfff0, 0xfff0 };
+VECT_VAR_DECL(expected0,hfloat,16,8) [] = { 0xcc00, 0xcc00, 0xcc00, 0xcc00,
+ 0xcc00, 0xcc00, 0xcc00, 0xcc00 };
VECT_VAR_DECL(expected0,hfloat,32,4) [] = { 0xc1800000, 0xc1800000,
0xc1800000, 0xc1800000 };
VECT_VAR_DECL(expected1,poly,8,8) [] = { 0xf1, 0xf1, 0xf1, 0xf1,
0xf1, 0xf1, 0xf1, 0xf1 };
VECT_VAR_DECL(expected1,poly,16,4) [] = { 0xfff1, 0xfff1, 0xfff1, 0xfff1 };
+VECT_VAR_DECL(expected1,hfloat,16,4) [] = { 0xcb80, 0xcb80, 0xcb80, 0xcb80 };
VECT_VAR_DECL(expected1,hfloat,32,2) [] = { 0xc1700000, 0xc1700000 };
VECT_VAR_DECL(expected1,int,8,16) [] = { 0xf1, 0xf1, 0xf1, 0xf1,
0xf1, 0xf1, 0xf1, 0xf1,
0xf1, 0xf1, 0xf1, 0xf1 };
VECT_VAR_DECL(expected1,poly,16,8) [] = { 0xfff1, 0xfff1, 0xfff1, 0xfff1,
0xfff1, 0xfff1, 0xfff1, 0xfff1 };
+VECT_VAR_DECL(expected1,hfloat,16,8) [] = { 0xcb80, 0xcb80, 0xcb80, 0xcb80,
+ 0xcb80, 0xcb80, 0xcb80, 0xcb80 };
VECT_VAR_DECL(expected1,hfloat,32,4) [] = { 0xc1700000, 0xc1700000,
0xc1700000, 0xc1700000 };
VECT_VAR_DECL(expected2,poly,8,8) [] = { 0xf2, 0xf2, 0xf2, 0xf2,
0xf2, 0xf2, 0xf2, 0xf2 };
VECT_VAR_DECL(expected2,poly,16,4) [] = { 0xfff2, 0xfff2, 0xfff2, 0xfff2 };
+VECT_VAR_DECL(expected2,hfloat,16,4) [] = { 0xcb00, 0xcb00, 0xcb00, 0xcb00 };
VECT_VAR_DECL(expected2,hfloat,32,2) [] = { 0xc1600000, 0xc1600000 };
VECT_VAR_DECL(expected2,int,8,16) [] = { 0xf2, 0xf2, 0xf2, 0xf2,
0xf2, 0xf2, 0xf2, 0xf2,
0xf2, 0xf2, 0xf2, 0xf2 };
VECT_VAR_DECL(expected2,poly,16,8) [] = { 0xfff2, 0xfff2, 0xfff2, 0xfff2,
0xfff2, 0xfff2, 0xfff2, 0xfff2 };
+VECT_VAR_DECL(expected2,hfloat,16,8) [] = { 0xcb00, 0xcb00, 0xcb00, 0xcb00,
+ 0xcb00, 0xcb00, 0xcb00, 0xcb00 };
VECT_VAR_DECL(expected2,hfloat,32,4) [] = { 0xc1600000, 0xc1600000,
0xc1600000, 0xc1600000 };
TEST_MACRO_ALL_VARIANTS_2_5(TEST_VLD1_DUP, vector, buffer_dup);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ TEST_VLD1_DUP(vector, buffer_dup, , float, f, 16, 4);
+ TEST_VLD1_DUP(vector, buffer_dup, q, float, f, 16, 8);
+#endif
TEST_VLD1_DUP(vector, buffer_dup, , float, f, 32, 2);
TEST_VLD1_DUP(vector, buffer_dup, q, float, f, 32, 4);
VECT_VAR_DECL(expected,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa,
0xaa, 0xaa, 0xaa, 0xf0 };
VECT_VAR_DECL(expected,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xfff0 };
+VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xcc00, 0xaaaa };
VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0xaaaaaaaa, 0xc1800000 };
VECT_VAR_DECL(expected,int,8,16) [] = { 0xaa, 0xaa, 0xaa, 0xaa,
0xaa, 0xaa, 0xaa, 0xaa,
0xf0, 0xaa, 0xaa, 0xaa };
VECT_VAR_DECL(expected,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xfff0, 0xaaaa };
+VECT_VAR_DECL(expected,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
+ 0xaaaa, 0xcc00, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa,
0xc1800000, 0xaaaaaaaa };
ARRAY(buffer_src, uint, 64, 1);
ARRAY(buffer_src, poly, 8, 8);
ARRAY(buffer_src, poly, 16, 4);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ ARRAY(buffer_src, float, 16, 4);
+#endif
ARRAY(buffer_src, float, 32, 2);
ARRAY(buffer_src, int, 8, 16);
ARRAY(buffer_src, uint, 64, 2);
ARRAY(buffer_src, poly, 8, 16);
ARRAY(buffer_src, poly, 16, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ ARRAY(buffer_src, float, 16, 8);
+#endif
ARRAY(buffer_src, float, 32, 4);
clean_results ();
TEST_VLD1_LANE(, uint, u, 64, 1, 0);
TEST_VLD1_LANE(, poly, p, 8, 8, 7);
TEST_VLD1_LANE(, poly, p, 16, 4, 3);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ TEST_VLD1_LANE(, float, f, 16, 4, 2);
+#endif
TEST_VLD1_LANE(, float, f, 32, 2, 1);
TEST_VLD1_LANE(q, int, s, 8, 16, 15);
TEST_VLD1_LANE(q, uint, u, 64, 2, 0);
TEST_VLD1_LANE(q, poly, p, 8, 16, 12);
TEST_VLD1_LANE(q, poly, p, 16, 8, 6);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ TEST_VLD1_LANE(q, float, f, 16, 8, 5);
+#endif
TEST_VLD1_LANE(q, float, f, 32, 4, 2);
CHECK_RESULTS (TEST_MSG, "");
VECT_VAR_DECL(expected_vld2_0,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf4, 0xf5, 0xf6, 0xf7 };
VECT_VAR_DECL(expected_vld2_0,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 };
+VECT_VAR_DECL(expected_vld2_0,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 };
VECT_VAR_DECL(expected_vld2_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 };
VECT_VAR_DECL(expected_vld2_0,int,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf4, 0xf5, 0xf6, 0xf7,
0xfc, 0xfd, 0xfe, 0xff };
VECT_VAR_DECL(expected_vld2_0,poly,16,8) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3,
0xfff4, 0xfff5, 0xfff6, 0xfff7 };
+VECT_VAR_DECL(expected_vld2_0,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80,
+ 0xca00, 0xc980, 0xc900, 0xc880 };
VECT_VAR_DECL(expected_vld2_0,hfloat,32,4) [] = { 0xc1800000, 0xc1700000,
0xc1600000, 0xc1500000 };
VECT_VAR_DECL(expected_vld2_1,poly,8,8) [] = { 0xf8, 0xf9, 0xfa, 0xfb,
0xfc, 0xfd, 0xfe, 0xff };
VECT_VAR_DECL(expected_vld2_1,poly,16,4) [] = { 0xfff4, 0xfff5, 0xfff6, 0xfff7 };
+VECT_VAR_DECL(expected_vld2_1,hfloat,16,4) [] = { 0xca00, 0xc980, 0xc900, 0xc880 };
VECT_VAR_DECL(expected_vld2_1,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 };
VECT_VAR_DECL(expected_vld2_1,int,8,16) [] = { 0x0, 0x1, 0x2, 0x3,
0x4, 0x5, 0x6, 0x7,
0xc, 0xd, 0xe, 0xf };
VECT_VAR_DECL(expected_vld2_1,poly,16,8) [] = { 0xfff8, 0xfff9, 0xfffa, 0xfffb,
0xfffc, 0xfffd, 0xfffe, 0xffff };
+VECT_VAR_DECL(expected_vld2_1,hfloat,16,8) [] = { 0xc800, 0xc700, 0xc600, 0xc500,
+ 0xc400, 0xc200, 0xc000, 0xbc00 };
VECT_VAR_DECL(expected_vld2_1,hfloat,32,4) [] = { 0xc1400000, 0xc1300000,
0xc1200000, 0xc1100000 };
VECT_VAR_DECL(expected_vld3_0,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf4, 0xf5, 0xf6, 0xf7 };
VECT_VAR_DECL(expected_vld3_0,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 };
+VECT_VAR_DECL(expected_vld3_0,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 };
VECT_VAR_DECL(expected_vld3_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 };
VECT_VAR_DECL(expected_vld3_0,int,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf4, 0xf5, 0xf6, 0xf7,
0xfc, 0xfd, 0xfe, 0xff };
VECT_VAR_DECL(expected_vld3_0,poly,16,8) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3,
0xfff4, 0xfff5, 0xfff6, 0xfff7 };
+VECT_VAR_DECL(expected_vld3_0,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80,
+ 0xca00, 0xc980, 0xc900, 0xc880 };
VECT_VAR_DECL(expected_vld3_0,hfloat,32,4) [] = { 0xc1800000, 0xc1700000,
0xc1600000, 0xc1500000 };
VECT_VAR_DECL(expected_vld3_1,poly,8,8) [] = { 0xf8, 0xf9, 0xfa, 0xfb,
0xfc, 0xfd, 0xfe, 0xff };
VECT_VAR_DECL(expected_vld3_1,poly,16,4) [] = { 0xfff4, 0xfff5, 0xfff6, 0xfff7 };
+VECT_VAR_DECL(expected_vld3_1,hfloat,16,4) [] = { 0xca00, 0xc980, 0xc900, 0xc880 };
VECT_VAR_DECL(expected_vld3_1,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 };
VECT_VAR_DECL(expected_vld3_1,int,8,16) [] = { 0x0, 0x1, 0x2, 0x3,
0x4, 0x5, 0x6, 0x7,
0xc, 0xd, 0xe, 0xf };
VECT_VAR_DECL(expected_vld3_1,poly,16,8) [] = { 0xfff8, 0xfff9, 0xfffa, 0xfffb,
0xfffc, 0xfffd, 0xfffe, 0xffff };
+VECT_VAR_DECL(expected_vld3_1,hfloat,16,8) [] = { 0xc800, 0xc700, 0xc600, 0xc500,
+ 0xc400, 0xc200, 0xc000, 0xbc00 };
VECT_VAR_DECL(expected_vld3_1,hfloat,32,4) [] = { 0xc1400000, 0xc1300000,
0xc1200000, 0xc1100000 };
0x4, 0x5, 0x6, 0x7 };
VECT_VAR_DECL(expected_vld3_2,poly,16,4) [] = { 0xfff8, 0xfff9,
0xfffa, 0xfffb };
+VECT_VAR_DECL(expected_vld3_2,hfloat,16,4) [] = { 0xc800, 0xc700, 0xc600, 0xc500 };
VECT_VAR_DECL(expected_vld3_2,hfloat,32,2) [] = { 0xc1400000, 0xc1300000 };
VECT_VAR_DECL(expected_vld3_2,int,8,16) [] = { 0x10, 0x11, 0x12, 0x13,
0x14, 0x15, 0x16, 0x17,
0x1c, 0x1d, 0x1e, 0x1f };
VECT_VAR_DECL(expected_vld3_2,poly,16,8) [] = { 0x0, 0x1, 0x2, 0x3,
0x4, 0x5, 0x6, 0x7 };
+VECT_VAR_DECL(expected_vld3_2,hfloat,16,8) [] = { 0x0000, 0x3c00, 0x4000, 0x4200,
+ 0x4400, 0x4500, 0x4600, 0x4700 };
VECT_VAR_DECL(expected_vld3_2,hfloat,32,4) [] = { 0xc1000000, 0xc0e00000,
0xc0c00000, 0xc0a00000 };
VECT_VAR_DECL(expected_vld4_0,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf4, 0xf5, 0xf6, 0xf7 };
VECT_VAR_DECL(expected_vld4_0,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 };
+VECT_VAR_DECL(expected_vld4_0,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 };
VECT_VAR_DECL(expected_vld4_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 };
VECT_VAR_DECL(expected_vld4_0,int,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf4, 0xf5, 0xf6, 0xf7,
0xfc, 0xfd, 0xfe, 0xff };
VECT_VAR_DECL(expected_vld4_0,poly,16,8) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3,
0xfff4, 0xfff5, 0xfff6, 0xfff7 };
+VECT_VAR_DECL(expected_vld4_0,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80,
+ 0xca00, 0xc980, 0xc900, 0xc880 };
VECT_VAR_DECL(expected_vld4_0,hfloat,32,4) [] = { 0xc1800000, 0xc1700000,
0xc1600000, 0xc1500000 };
VECT_VAR_DECL(expected_vld4_1,poly,8,8) [] = { 0xf8, 0xf9, 0xfa, 0xfb,
0xfc, 0xfd, 0xfe, 0xff };
VECT_VAR_DECL(expected_vld4_1,poly,16,4) [] = { 0xfff4, 0xfff5, 0xfff6, 0xfff7 };
+VECT_VAR_DECL(expected_vld4_1,hfloat,16,4) [] = { 0xca00, 0xc980, 0xc900, 0xc880 };
VECT_VAR_DECL(expected_vld4_1,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 };
VECT_VAR_DECL(expected_vld4_1,int,8,16) [] = { 0x0, 0x1, 0x2, 0x3,
0x4, 0x5, 0x6, 0x7,
0xc, 0xd, 0xe, 0xf };
VECT_VAR_DECL(expected_vld4_1,poly,16,8) [] = { 0xfff8, 0xfff9, 0xfffa, 0xfffb,
0xfffc, 0xfffd, 0xfffe, 0xffff };
+VECT_VAR_DECL(expected_vld4_1,hfloat,16,8) [] = { 0xc800, 0xc700, 0xc600, 0xc500,
+ 0xc400, 0xc200, 0xc000, 0xbc00 };
VECT_VAR_DECL(expected_vld4_1,hfloat,32,4) [] = { 0xc1400000, 0xc1300000,
0xc1200000, 0xc1100000 };
VECT_VAR_DECL(expected_vld4_2,poly,8,8) [] = { 0x0, 0x1, 0x2, 0x3,
0x4, 0x5, 0x6, 0x7 };
VECT_VAR_DECL(expected_vld4_2,poly,16,4) [] = { 0xfff8, 0xfff9, 0xfffa, 0xfffb };
+VECT_VAR_DECL(expected_vld4_2,hfloat,16,4) [] = { 0xc800, 0xc700, 0xc600, 0xc500 };
VECT_VAR_DECL(expected_vld4_2,hfloat,32,2) [] = { 0xc1400000, 0xc1300000 };
VECT_VAR_DECL(expected_vld4_2,int,8,16) [] = { 0x10, 0x11, 0x12, 0x13,
0x14, 0x15, 0x16, 0x17,
0x1c, 0x1d, 0x1e, 0x1f };
VECT_VAR_DECL(expected_vld4_2,poly,16,8) [] = { 0x0, 0x1, 0x2, 0x3,
0x4, 0x5, 0x6, 0x7 };
+VECT_VAR_DECL(expected_vld4_2,hfloat,16,8) [] = { 0x0000, 0x3c00, 0x4000, 0x4200,
+ 0x4400, 0x4500, 0x4600, 0x4700 };
VECT_VAR_DECL(expected_vld4_2,hfloat,32,4) [] = { 0xc1000000, 0xc0e00000,
0xc0c00000, 0xc0a00000 };
VECT_VAR_DECL(expected_vld4_3,poly,8,8) [] = { 0x8, 0x9, 0xa, 0xb,
0xc, 0xd, 0xe, 0xf };
VECT_VAR_DECL(expected_vld4_3,poly,16,4) [] = { 0xfffc, 0xfffd, 0xfffe, 0xffff };
+VECT_VAR_DECL(expected_vld4_3,hfloat,16,4) [] = { 0xc400, 0xc200, 0xc000, 0xbc00 };
VECT_VAR_DECL(expected_vld4_3,hfloat,32,2) [] = { 0xc1200000, 0xc1100000 };
VECT_VAR_DECL(expected_vld4_3,int,8,16) [] = { 0x20, 0x21, 0x22, 0x23,
0x24, 0x25, 0x26, 0x27,
0x2c, 0x2d, 0x2e, 0x2f };
VECT_VAR_DECL(expected_vld4_3,poly,16,8) [] = { 0x8, 0x9, 0xa, 0xb,
0xc, 0xd, 0xe, 0xf };
+VECT_VAR_DECL(expected_vld4_3,hfloat,16,8) [] = { 0x4800, 0x4880, 0x4900, 0x4980,
+ 0x4a00, 0x4a80, 0x4b00, 0x4b80 };
VECT_VAR_DECL(expected_vld4_3,hfloat,32,4) [] = { 0xc0800000, 0xc0400000,
0xc0000000, 0xbf800000 };
sizeof(VECT_VAR(result, T1, W, N)));
/* We need all variants in 64 bits, but there is no 64x2 variant. */
-#define DECL_ALL_VLDX(X) \
+#define DECL_ALL_VLDX_NO_FP16(X) \
DECL_VLDX(int, 8, 8, X); \
DECL_VLDX(int, 16, 4, X); \
DECL_VLDX(int, 32, 2, X); \
DECL_VLDX(poly, 16, 8, X); \
DECL_VLDX(float, 32, 4, X)
-#define TEST_ALL_VLDX(X) \
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define DECL_ALL_VLDX(X) \
+ DECL_ALL_VLDX_NO_FP16(X); \
+ DECL_VLDX(float, 16, 4, X); \
+ DECL_VLDX(float, 16, 8, X)
+#else
+#define DECL_ALL_VLDX(X) DECL_ALL_VLDX_NO_FP16(X)
+#endif
+
+#define TEST_ALL_VLDX_NO_FP16(X) \
TEST_VLDX(, int, s, 8, 8, X); \
TEST_VLDX(, int, s, 16, 4, X); \
TEST_VLDX(, int, s, 32, 2, X); \
TEST_VLDX(q, poly, p, 16, 8, X); \
TEST_VLDX(q, float, f, 32, 4, X)
-#define TEST_ALL_EXTRA_CHUNKS(X, Y) \
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define TEST_ALL_VLDX(X) \
+ TEST_ALL_VLDX_NO_FP16(X); \
+ TEST_VLDX(, float, f, 16, 4, X); \
+ TEST_VLDX(q, float, f, 16, 8, X)
+#else
+#define TEST_ALL_VLDX(X) TEST_ALL_VLDX_NO_FP16(X)
+#endif
+
+#define TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y) \
TEST_EXTRA_CHUNK(int, 8, 8, X, Y); \
TEST_EXTRA_CHUNK(int, 16, 4, X, Y); \
TEST_EXTRA_CHUNK(int, 32, 2, X, Y); \
TEST_EXTRA_CHUNK(poly, 16, 8, X, Y); \
TEST_EXTRA_CHUNK(float, 32, 4, X, Y)
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define TEST_ALL_EXTRA_CHUNKS(X, Y) \
+ TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y) \
+ TEST_EXTRA_CHUNK(float, 16, 4, X, Y); \
+ TEST_EXTRA_CHUNK(float, 16, 8, X, Y);
+#else
+#define TEST_ALL_EXTRA_CHUNKS(X, Y) TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y)
+#endif
+
/* vldX supports all vector types except [u]int64x2. */
-#define CHECK_RESULTS_VLDX(test_name,EXPECTED,comment) \
- { \
+#define CHECK_RESULTS_VLDX_NO_FP16(test_name,EXPECTED,comment) \
CHECK(test_name, int, 8, 8, PRIx8, EXPECTED, comment); \
CHECK(test_name, int, 16, 4, PRIx16, EXPECTED, comment); \
CHECK(test_name, int, 32, 2, PRIx32, EXPECTED, comment); \
CHECK(test_name, uint, 32, 4, PRIx32, EXPECTED, comment); \
CHECK(test_name, poly, 8, 16, PRIx8, EXPECTED, comment); \
CHECK(test_name, poly, 16, 8, PRIx16, EXPECTED, comment); \
- CHECK_FP(test_name, float, 32, 4, PRIx32, EXPECTED, comment); \
- } \
+ CHECK_FP(test_name, float, 32, 4, PRIx32, EXPECTED, comment)
+
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define CHECK_RESULTS_VLDX(test_name,EXPECTED,comment) \
+ { \
+ CHECK_RESULTS_VLDX_NO_FP16(test_name, EXPECTED, comment); \
+ CHECK_FP(test_name, float, 16, 4, PRIx16, EXPECTED, comment); \
+ CHECK_FP(test_name, float, 16, 8, PRIx16, EXPECTED, comment); \
+ }
+#else
+#define CHECK_RESULTS_VLDX(test_name, EXPECTED, comment) \
+ { CHECK_RESULTS_VLDX_NO_FP16(test_name, EXPECTED, comment); }
+#endif
DECL_ALL_VLDX(2);
DECL_ALL_VLDX(3);
PAD(buffer_vld2_pad, poly, 8, 8);
VECT_ARRAY_INIT2(buffer_vld2, poly, 16, 4);
PAD(buffer_vld2_pad, poly, 16, 4);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VECT_ARRAY_INIT2(buffer_vld2, float, 16, 4);
+ PAD(buffer_vld2_pad, float, 16, 4);
+#endif
VECT_ARRAY_INIT2(buffer_vld2, float, 32, 2);
PAD(buffer_vld2_pad, float, 32, 2);
PAD(buffer_vld2_pad, poly, 8, 16);
VECT_ARRAY_INIT2(buffer_vld2, poly, 16, 8);
PAD(buffer_vld2_pad, poly, 16, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VECT_ARRAY_INIT2(buffer_vld2, float, 16, 8);
+ PAD(buffer_vld2_pad, float, 16, 8);
+#endif
VECT_ARRAY_INIT2(buffer_vld2, float, 32, 4);
PAD(buffer_vld2_pad, float, 32, 4);
PAD(buffer_vld3_pad, poly, 8, 8);
VECT_ARRAY_INIT3(buffer_vld3, poly, 16, 4);
PAD(buffer_vld3_pad, poly, 16, 4);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VECT_ARRAY_INIT3(buffer_vld3, float, 16, 4);
+ PAD(buffer_vld3_pad, float, 16, 4);
+#endif
VECT_ARRAY_INIT3(buffer_vld3, float, 32, 2);
PAD(buffer_vld3_pad, float, 32, 2);
PAD(buffer_vld3_pad, poly, 8, 16);
VECT_ARRAY_INIT3(buffer_vld3, poly, 16, 8);
PAD(buffer_vld3_pad, poly, 16, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VECT_ARRAY_INIT3(buffer_vld3, float, 16, 8);
+ PAD(buffer_vld3_pad, float, 16, 8);
+#endif
VECT_ARRAY_INIT3(buffer_vld3, float, 32, 4);
PAD(buffer_vld3_pad, float, 32, 4);
PAD(buffer_vld4_pad, poly, 8, 8);
VECT_ARRAY_INIT4(buffer_vld4, poly, 16, 4);
PAD(buffer_vld4_pad, poly, 16, 4);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VECT_ARRAY_INIT4(buffer_vld4, float, 16, 4);
+ PAD(buffer_vld4_pad, float, 16, 4);
+#endif
VECT_ARRAY_INIT4(buffer_vld4, float, 32, 2);
PAD(buffer_vld4_pad, float, 32, 2);
PAD(buffer_vld4_pad, poly, 8, 16);
VECT_ARRAY_INIT4(buffer_vld4, poly, 16, 8);
PAD(buffer_vld4_pad, poly, 16, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VECT_ARRAY_INIT4(buffer_vld4, float, 16, 8);
+ PAD(buffer_vld4_pad, float, 16, 8);
+#endif
VECT_ARRAY_INIT4(buffer_vld4, float, 32, 4);
PAD(buffer_vld4_pad, float, 32, 4);
VECT_VAR_DECL(expected_vld2_0,poly,8,8) [] = { 0xf0, 0xf1, 0xf0, 0xf1,
0xf0, 0xf1, 0xf0, 0xf1 };
VECT_VAR_DECL(expected_vld2_0,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff0, 0xfff1 };
+VECT_VAR_DECL(expected_vld2_0,hfloat,16,4) [] = {0xcc00, 0xcb80, 0xcc00, 0xcb80 };
VECT_VAR_DECL(expected_vld2_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 };
/* vld2_dup/chunk 1. */
0xf0, 0xf1, 0xf0, 0xf1 };
VECT_VAR_DECL(expected_vld2_1,poly,16,4) [] = { 0xfff0, 0xfff1,
0xfff0, 0xfff1 };
+VECT_VAR_DECL(expected_vld2_1,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcc00, 0xcb80 };
VECT_VAR_DECL(expected_vld2_1,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 };
/* vld3_dup/chunk 0. */
0xf1, 0xf2, 0xf0, 0xf1 };
VECT_VAR_DECL(expected_vld3_0,poly,16,4) [] = { 0xfff0, 0xfff1,
0xfff2, 0xfff0 };
+VECT_VAR_DECL(expected_vld3_0,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xcc00 };
VECT_VAR_DECL(expected_vld3_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 };
/* vld3_dup/chunk 1. */
0xf0, 0xf1, 0xf2, 0xf0 };
VECT_VAR_DECL(expected_vld3_1,poly,16,4) [] = { 0xfff1, 0xfff2,
0xfff0, 0xfff1 };
+VECT_VAR_DECL(expected_vld3_1,hfloat,16,4) [] = { 0xcb80, 0xcb00, 0xcc00, 0xcb80 };
VECT_VAR_DECL(expected_vld3_1,hfloat,32,2) [] = { 0xc1600000, 0xc1800000 };
/* vld3_dup/chunk 2. */
0xf2, 0xf0, 0xf1, 0xf2 };
VECT_VAR_DECL(expected_vld3_2,poly,16,4) [] = { 0xfff2, 0xfff0,
0xfff1, 0xfff2 };
+VECT_VAR_DECL(expected_vld3_2,hfloat,16,4) [] = { 0xcb00, 0xcc00, 0xcb80, 0xcb00 };
VECT_VAR_DECL(expected_vld3_2,hfloat,32,2) [] = { 0xc1700000, 0xc1600000 };
/* vld4_dup/chunk 0. */
VECT_VAR_DECL(expected_vld4_0,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf0, 0xf1, 0xf2, 0xf3 };
VECT_VAR_DECL(expected_vld4_0,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 };
+VECT_VAR_DECL(expected_vld4_0,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 };
VECT_VAR_DECL(expected_vld4_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 };
/* vld4_dup/chunk 1. */
VECT_VAR_DECL(expected_vld4_1,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf0, 0xf1, 0xf2, 0xf3 };
VECT_VAR_DECL(expected_vld4_1,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 };
+VECT_VAR_DECL(expected_vld4_1,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 };
VECT_VAR_DECL(expected_vld4_1,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 };
/* vld4_dup/chunk 2. */
VECT_VAR_DECL(expected_vld4_2,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf0, 0xf1, 0xf2, 0xf3 };
VECT_VAR_DECL(expected_vld4_2,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 };
+VECT_VAR_DECL(expected_vld4_2,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 };
VECT_VAR_DECL(expected_vld4_2,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 };
/* vld4_dup/chunk3. */
VECT_VAR_DECL(expected_vld4_3,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf0, 0xf1, 0xf2, 0xf3 };
VECT_VAR_DECL(expected_vld4_3,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 };
+VECT_VAR_DECL(expected_vld4_3,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 };
VECT_VAR_DECL(expected_vld4_3,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 };
void exec_vldX_dup (void)
&(VECT_VAR(result_bis_##X, T1, W, N)[Y*N]), \
sizeof(VECT_VAR(result, T1, W, N)));
-#define DECL_ALL_VLDX_DUP(X) \
+#define DECL_ALL_VLDX_DUP_NO_FP16(X) \
DECL_VLDX_DUP(int, 8, 8, X); \
DECL_VLDX_DUP(int, 16, 4, X); \
DECL_VLDX_DUP(int, 32, 2, X); \
DECL_VLDX_DUP(poly, 16, 4, X); \
DECL_VLDX_DUP(float, 32, 2, X)
-#define TEST_ALL_VLDX_DUP(X) \
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define DECL_ALL_VLDX_DUP(X) \
+ DECL_ALL_VLDX_DUP_NO_FP16(X); \
+ DECL_VLDX_DUP(float, 16, 4, X)
+#else
+#define DECL_ALL_VLDX_DUP(X) DECL_ALL_VLDX_DUP_NO_FP16(X)
+#endif
+
+#define TEST_ALL_VLDX_DUP_NO_FP16(X) \
TEST_VLDX_DUP(, int, s, 8, 8, X); \
TEST_VLDX_DUP(, int, s, 16, 4, X); \
TEST_VLDX_DUP(, int, s, 32, 2, X); \
TEST_VLDX_DUP(, poly, p, 16, 4, X); \
TEST_VLDX_DUP(, float, f, 32, 2, X)
-#define TEST_ALL_EXTRA_CHUNKS(X, Y) \
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define TEST_ALL_VLDX_DUP(X) \
+ TEST_ALL_VLDX_DUP_NO_FP16(X); \
+ TEST_VLDX_DUP(, float, f, 16, 4, X)
+#else
+#define TEST_ALL_VLDX_DUP(X) TEST_ALL_VLDX_DUP_NO_FP16(X)
+#endif
+
+#define TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y) \
TEST_EXTRA_CHUNK(int, 8, 8, X, Y); \
TEST_EXTRA_CHUNK(int, 16, 4, X, Y); \
TEST_EXTRA_CHUNK(int, 32, 2, X, Y); \
TEST_EXTRA_CHUNK(poly, 16, 4, X, Y); \
TEST_EXTRA_CHUNK(float, 32, 2, X, Y)
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define TEST_ALL_EXTRA_CHUNKS(X, Y) \
+ TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y); \
+ TEST_EXTRA_CHUNK(float, 16, 4, X, Y)
+#else
+#define TEST_ALL_EXTRA_CHUNKS(X, Y) TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y)
+#endif
+
/* vldX_dup supports only 64-bit inputs. */
-#define CHECK_RESULTS_VLDX_DUP(test_name,EXPECTED,comment) \
- { \
+#define CHECK_RESULTS_VLDX_DUP_NO_FP16(test_name,EXPECTED,comment) \
CHECK(test_name, int, 8, 8, PRIx8, EXPECTED, comment); \
CHECK(test_name, int, 16, 4, PRIx16, EXPECTED, comment); \
CHECK(test_name, int, 32, 2, PRIx32, EXPECTED, comment); \
CHECK(test_name, uint, 64, 1, PRIx64, EXPECTED, comment); \
CHECK(test_name, poly, 8, 8, PRIx8, EXPECTED, comment); \
CHECK(test_name, poly, 16, 4, PRIx16, EXPECTED, comment); \
- CHECK_FP(test_name, float, 32, 2, PRIx32, EXPECTED, comment); \
- } \
+ CHECK_FP(test_name, float, 32, 2, PRIx32, EXPECTED, comment)
+
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define CHECK_RESULTS_VLDX_DUP(test_name,EXPECTED,comment) \
+ { \
+ CHECK_RESULTS_VLDX_DUP_NO_FP16(test_name,EXPECTED,comment); \
+ CHECK_FP(test_name, float, 16, 4, PRIx16, EXPECTED, comment); \
+ }
+#else
+#define CHECK_RESULTS_VLDX_DUP(test_name,EXPECTED,comment) \
+ { \
+ CHECK_RESULTS_VLDX_DUP_NO_FP16(test_name,EXPECTED,comment); \
+ }
+#endif
DECL_ALL_VLDX_DUP(2);
DECL_ALL_VLDX_DUP(3);
PAD(buffer_vld2_pad, poly, 8, 8);
VECT_ARRAY_INIT2(buffer_vld2, poly, 16, 4);
PAD(buffer_vld2_pad, poly, 16, 4);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VECT_ARRAY_INIT2(buffer_vld2, float, 16, 4);
+ PAD(buffer_vld2_pad, float, 16, 4);
+#endif
VECT_ARRAY_INIT2(buffer_vld2, float, 32, 2);
PAD(buffer_vld2_pad, float, 32, 2);
PAD(buffer_vld2_pad, poly, 8, 16);
VECT_ARRAY_INIT2(buffer_vld2, poly, 16, 8);
PAD(buffer_vld2_pad, poly, 16, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VECT_ARRAY_INIT2(buffer_vld2, float, 16, 8);
+ PAD(buffer_vld2_pad, float, 16, 8);
+#endif
VECT_ARRAY_INIT2(buffer_vld2, float, 32, 4);
PAD(buffer_vld2_pad, float, 32, 4);
PAD(buffer_vld3_pad, poly, 8, 8);
VECT_ARRAY_INIT3(buffer_vld3, poly, 16, 4);
PAD(buffer_vld3_pad, poly, 16, 4);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VECT_ARRAY_INIT3(buffer_vld3, float, 16, 4);
+ PAD(buffer_vld3_pad, float, 16, 4);
+#endif
VECT_ARRAY_INIT3(buffer_vld3, float, 32, 2);
PAD(buffer_vld3_pad, float, 32, 2);
PAD(buffer_vld3_pad, poly, 8, 16);
VECT_ARRAY_INIT3(buffer_vld3, poly, 16, 8);
PAD(buffer_vld3_pad, poly, 16, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VECT_ARRAY_INIT3(buffer_vld3, float, 16, 8);
+ PAD(buffer_vld3_pad, float, 16, 8);
+#endif
VECT_ARRAY_INIT3(buffer_vld3, float, 32, 4);
PAD(buffer_vld3_pad, float, 32, 4);
PAD(buffer_vld4_pad, poly, 8, 8);
VECT_ARRAY_INIT4(buffer_vld4, poly, 16, 4);
PAD(buffer_vld4_pad, poly, 16, 4);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VECT_ARRAY_INIT4(buffer_vld4, float, 16, 4);
+ PAD(buffer_vld4_pad, float, 16, 4);
+#endif
VECT_ARRAY_INIT4(buffer_vld4, float, 32, 2);
PAD(buffer_vld4_pad, float, 32, 2);
PAD(buffer_vld4_pad, poly, 8, 16);
VECT_ARRAY_INIT4(buffer_vld4, poly, 16, 8);
PAD(buffer_vld4_pad, poly, 16, 8);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VECT_ARRAY_INIT4(buffer_vld4, float, 16, 8);
+ PAD(buffer_vld4_pad, float, 16, 8);
+#endif
VECT_ARRAY_INIT4(buffer_vld4, float, 32, 4);
PAD(buffer_vld4_pad, float, 32, 4);
0xaa, 0xaa, 0xaa, 0xaa };
VECT_VAR_DECL(expected_vld2_0,poly,16,4) [] = { 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa };
+VECT_VAR_DECL(expected_vld2_0,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld2_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 };
VECT_VAR_DECL(expected_vld2_0,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld2_0,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
+VECT_VAR_DECL(expected_vld2_0,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
+ 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa } ;
VECT_VAR_DECL(expected_vld2_0,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa,
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld2_1,poly,8,8) [] = { 0xf0, 0xf1, 0xaa, 0xaa,
0xaa, 0xaa, 0xaa, 0xaa };
VECT_VAR_DECL(expected_vld2_1,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xfff0, 0xfff1 };
+VECT_VAR_DECL(expected_vld2_1,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld2_1,hfloat,32,2) [] = { 0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld2_1,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xfff0, 0xfff1, 0xaaaa, 0xaaaa };
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld2_1,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xfff0, 0xfff1,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
+VECT_VAR_DECL(expected_vld2_1,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
+ 0xcc00, 0xcb80, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld2_1,hfloat,32,4) [] = { 0xc1800000, 0xc1700000,
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld3_0,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa,
0xaa, 0xaa, 0xaa, 0xaa };
VECT_VAR_DECL(expected_vld3_0,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
+VECT_VAR_DECL(expected_vld3_0,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld3_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 };
VECT_VAR_DECL(expected_vld3_0,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
0xfffffff2, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld3_0,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
+VECT_VAR_DECL(expected_vld3_0,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
+ 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld3_0,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa,
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld3_1,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa,
0xf0, 0xf1, 0xf2, 0xaa };
VECT_VAR_DECL(expected_vld3_1,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
+VECT_VAR_DECL(expected_vld3_1,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xcc00, 0xcb80 };
VECT_VAR_DECL(expected_vld3_1,hfloat,32,2) [] = { 0xc1600000, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld3_1,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld3_1,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xaaaa, 0xfff0 };
+VECT_VAR_DECL(expected_vld3_1,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
+ 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld3_1,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa,
0xc1800000, 0xc1700000 };
VECT_VAR_DECL(expected_vld3_2,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa,
0xaa, 0xaa, 0xaa, 0xaa };
VECT_VAR_DECL(expected_vld3_2,poly,16,4) [] = { 0xaaaa, 0xfff0, 0xfff1, 0xfff2 };
+VECT_VAR_DECL(expected_vld3_2,hfloat,16,4) [] = { 0xcb00, 0xaaaa, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld3_2,hfloat,32,2) [] = { 0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld3_2,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xfff0, 0xfff1,
0xfff2, 0xaaaa, 0xaaaa, 0xaaaa };
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld3_2,poly,16,8) [] = { 0xfff1, 0xfff2, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
+VECT_VAR_DECL(expected_vld3_2,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xcc00, 0xcb80,
+ 0xcb00, 0xaaaa, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld3_2,hfloat,32,4) [] = { 0xc1600000, 0xaaaaaaaa,
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld4_0,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa,
0xaa, 0xaa, 0xaa, 0xaa };
VECT_VAR_DECL(expected_vld4_0,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
+VECT_VAR_DECL(expected_vld4_0,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld4_0,hfloat,32,2) [] = { 0xc1800000, 0xc1700000 };
VECT_VAR_DECL(expected_vld4_0,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
0xfffffff2, 0xfffffff3 };
VECT_VAR_DECL(expected_vld4_0,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
+VECT_VAR_DECL(expected_vld4_0,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
+ 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld4_0,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa,
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld4_1,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa,
0xaa, 0xaa, 0xaa, 0xaa };
VECT_VAR_DECL(expected_vld4_1,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
+VECT_VAR_DECL(expected_vld4_1,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld4_1,hfloat,32,2) [] = { 0xc1600000, 0xc1500000 };
VECT_VAR_DECL(expected_vld4_1,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld4_1,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
+VECT_VAR_DECL(expected_vld4_1,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
+ 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld4_1,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa,
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld4_2,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xaa, 0xaa, 0xaa, 0xaa };
VECT_VAR_DECL(expected_vld4_2,poly,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
+VECT_VAR_DECL(expected_vld4_2,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80 };
VECT_VAR_DECL(expected_vld4_2,hfloat,32,2) [] = { 0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld4_2,int,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld4_2,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xfff0, 0xfff1, 0xfff2, 0xfff3 };
+VECT_VAR_DECL(expected_vld4_2,hfloat,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
+ 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld4_2,hfloat,32,4) [] = { 0xc1800000, 0xc1700000,
0xc1600000, 0xc1500000 };
VECT_VAR_DECL(expected_vld4_3,poly,8,8) [] = { 0xaa, 0xaa, 0xaa, 0xaa,
0xaa, 0xaa, 0xaa, 0xaa };
VECT_VAR_DECL(expected_vld4_3,poly,16,4) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3 };
+VECT_VAR_DECL(expected_vld4_3,hfloat,16,4) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld4_3,hfloat,32,2) [] = { 0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld4_3,int,16,8) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL(expected_vld4_3,poly,16,8) [] = { 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa,
0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
+VECT_VAR_DECL(expected_vld4_3,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80,
+ 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa };
VECT_VAR_DECL(expected_vld4_3,hfloat,32,4) [] = { 0xaaaaaaaa, 0xaaaaaaaa,
0xaaaaaaaa, 0xaaaaaaaa };
VECT_VAR_DECL_INIT(buffer_vld2_lane, uint, 64, 2);
VECT_VAR_DECL_INIT(buffer_vld2_lane, poly, 8, 2);
VECT_VAR_DECL_INIT(buffer_vld2_lane, poly, 16, 2);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+VECT_VAR_DECL_INIT(buffer_vld2_lane, float, 16, 2);
+#endif
VECT_VAR_DECL_INIT(buffer_vld2_lane, float, 32, 2);
/* Input buffers for vld3_lane */
VECT_VAR_DECL_INIT(buffer_vld3_lane, uint, 64, 3);
VECT_VAR_DECL_INIT(buffer_vld3_lane, poly, 8, 3);
VECT_VAR_DECL_INIT(buffer_vld3_lane, poly, 16, 3);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+VECT_VAR_DECL_INIT(buffer_vld3_lane, float, 16, 3);
+#endif
VECT_VAR_DECL_INIT(buffer_vld3_lane, float, 32, 3);
/* Input buffers for vld4_lane */
VECT_VAR_DECL_INIT(buffer_vld4_lane, uint, 64, 4);
VECT_VAR_DECL_INIT(buffer_vld4_lane, poly, 8, 4);
VECT_VAR_DECL_INIT(buffer_vld4_lane, poly, 16, 4);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+VECT_VAR_DECL_INIT(buffer_vld4_lane, float, 16, 4);
+#endif
VECT_VAR_DECL_INIT(buffer_vld4_lane, float, 32, 4);
void exec_vldX_lane (void)
sizeof(VECT_VAR(result, T1, W, N)));
/* We need all variants in 64 bits, but there is no 64x2 variant. */
-#define DECL_ALL_VLDX_LANE(X) \
+#define DECL_ALL_VLDX_LANE_NO_FP16(X) \
DECL_VLDX_LANE(int, 8, 8, X); \
DECL_VLDX_LANE(int, 16, 4, X); \
DECL_VLDX_LANE(int, 32, 2, X); \
DECL_VLDX_LANE(float, 32, 2, X); \
DECL_VLDX_LANE(float, 32, 4, X)
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define DECL_ALL_VLDX_LANE(X) \
+ DECL_ALL_VLDX_LANE_NO_FP16(X); \
+ DECL_VLDX_LANE(float, 16, 4, X); \
+ DECL_VLDX_LANE(float, 16, 8, X)
+#else
+#define DECL_ALL_VLDX_LANE(X) DECL_ALL_VLDX_LANE_NO_FP16(X)
+#endif
+
/* Add some padding to try to catch out of bound accesses. */
#define ARRAY1(V, T, W, N) VECT_VAR_DECL(V,T,W,N)[1]={42}
#define DUMMY_ARRAY(V, T, W, N, L) \
/* Use the same lanes regardless of the size of the array (X), for
simplicity. */
-#define TEST_ALL_VLDX_LANE(X) \
+#define TEST_ALL_VLDX_LANE_NO_FP16(X) \
TEST_VLDX_LANE(, int, s, 8, 8, X, 7); \
TEST_VLDX_LANE(, int, s, 16, 4, X, 2); \
TEST_VLDX_LANE(, int, s, 32, 2, X, 0); \
TEST_VLDX_LANE(, float, f, 32, 2, X, 0); \
TEST_VLDX_LANE(q, float, f, 32, 4, X, 2)
-#define TEST_ALL_EXTRA_CHUNKS(X, Y) \
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define TEST_ALL_VLDX_LANE(X) \
+ TEST_ALL_VLDX_LANE_NO_FP16(X); \
+ TEST_VLDX_LANE(, float, f, 16, 4, X, 2); \
+ TEST_VLDX_LANE(q, float, f, 16, 8, X, 6)
+#else
+#define TEST_ALL_VLDX_LANE(X) TEST_ALL_VLDX_LANE_NO_FP16(X)
+#endif
+
+#define TEST_ALL_EXTRA_CHUNKS_NO_FP16(X,Y) \
TEST_EXTRA_CHUNK(int, 8, 8, X, Y); \
TEST_EXTRA_CHUNK(int, 16, 4, X, Y); \
TEST_EXTRA_CHUNK(int, 32, 2, X, Y); \
TEST_EXTRA_CHUNK(float, 32, 2, X, Y); \
TEST_EXTRA_CHUNK(float, 32, 4, X, Y)
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define TEST_ALL_EXTRA_CHUNKS(X,Y) \
+ TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y); \
+ TEST_EXTRA_CHUNK(float, 16, 4, X, Y); \
+ TEST_EXTRA_CHUNK(float, 16, 8, X, Y)
+#else
+#define TEST_ALL_EXTRA_CHUNKS(X,Y) TEST_ALL_EXTRA_CHUNKS_NO_FP16(X, Y)
+#endif
+
/* vldX_lane supports only a subset of all variants. */
-#define CHECK_RESULTS_VLDX_LANE(test_name,EXPECTED,comment) \
- { \
+#define CHECK_RESULTS_VLDX_LANE_NO_FP16(test_name,EXPECTED,comment) \
CHECK(test_name, int, 8, 8, PRIx8, EXPECTED, comment); \
CHECK(test_name, int, 16, 4, PRIx16, EXPECTED, comment); \
CHECK(test_name, int, 32, 2, PRIx32, EXPECTED, comment); \
CHECK(test_name, uint, 16, 8, PRIx16, EXPECTED, comment); \
CHECK(test_name, uint, 32, 4, PRIx32, EXPECTED, comment); \
CHECK(test_name, poly, 16, 8, PRIx16, EXPECTED, comment); \
- CHECK_FP(test_name, float, 32, 4, PRIx32, EXPECTED, comment); \
- } \
+ CHECK_FP(test_name, float, 32, 4, PRIx32, EXPECTED, comment)
+
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+#define CHECK_RESULTS_VLDX_LANE(test_name,EXPECTED,comment) \
+ { \
+ CHECK_RESULTS_VLDX_LANE_NO_FP16(test_name,EXPECTED,comment); \
+ CHECK_FP(test_name, float, 16, 4, PRIx16, EXPECTED, comment); \
+ CHECK_FP(test_name, float, 16, 8, PRIx16, EXPECTED, comment); \
+ }
+#else
+#define CHECK_RESULTS_VLDX_LANE(test_name,EXPECTED,comment) \
+ { \
+ CHECK_RESULTS_VLDX_LANE_NO_FP16(test_name,EXPECTED,comment); \
+ }
+#endif
/* Declare the temporary buffers / variables. */
DECL_ALL_VLDX_LANE(2);
DUMMY_ARRAY(buffer_src, uint, 16, 8, 4);
DUMMY_ARRAY(buffer_src, uint, 32, 4, 4);
DUMMY_ARRAY(buffer_src, poly, 16, 8, 4);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ DUMMY_ARRAY(buffer_src, float, 16, 4, 4);
+ DUMMY_ARRAY(buffer_src, float, 16, 8, 4);
+#endif
DUMMY_ARRAY(buffer_src, float, 32, 2, 4);
DUMMY_ARRAY(buffer_src, float, 32, 4, 4);
VECT_VAR_DECL(expected,poly,8,8) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf4, 0xf5, 0x55, 0xf7 };
VECT_VAR_DECL(expected,poly,16,4) [] = { 0xfff0, 0xfff1, 0x66, 0xfff3 };
+VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xcc00, 0xcb80, 0x4840, 0xca80 };
VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0xc1800000, 0x4204cccd };
VECT_VAR_DECL(expected,int,8,16) [] = { 0xf0, 0xf1, 0xf2, 0xf3,
0xf4, 0xf5, 0xf6, 0xf7,
0xfc, 0xfd, 0xdd, 0xff };
VECT_VAR_DECL(expected,poly,16,8) [] = { 0xfff0, 0xfff1, 0xfff2, 0xfff3,
0xfff4, 0xfff5, 0xee, 0xfff7 };
+VECT_VAR_DECL(expected,hfloat,16,8) [] = { 0xcc00, 0xcb80, 0xcb00, 0xca80,
+ 0xca00, 0x4480, 0xc900, 0xc880 };
VECT_VAR_DECL(expected,hfloat,32,4) [] = { 0xc1800000, 0xc1700000,
0xc1600000, 0x41333333 };
/* Initialize input "vector" from "buffer". */
TEST_MACRO_ALL_VARIANTS_2_5(VLOAD, vector, buffer);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ VLOAD(vector, buffer, , float, f, 16, 4);
+ VLOAD(vector, buffer, q, float, f, 16, 8);
+#endif
VLOAD(vector, buffer, , float, f, 32, 2);
VLOAD(vector, buffer, q, float, f, 32, 4);
TEST_VSET_LANE(, uint, u, 64, 1, 0x88, 0);
TEST_VSET_LANE(, poly, p, 8, 8, 0x55, 6);
TEST_VSET_LANE(, poly, p, 16, 4, 0x66, 2);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ TEST_VSET_LANE(, float, f, 16, 4, 8.5f, 2);
+#endif
TEST_VSET_LANE(, float, f, 32, 2, 33.2f, 1);
TEST_VSET_LANE(q, int, s, 8, 16, 0x99, 15);
TEST_VSET_LANE(q, uint, u, 64, 2, 0x11, 1);
TEST_VSET_LANE(q, poly, p, 8, 16, 0xDD, 14);
TEST_VSET_LANE(q, poly, p, 16, 8, 0xEE, 6);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ TEST_VSET_LANE(q, float, f, 16, 8, 4.5f, 5);
+#endif
TEST_VSET_LANE(q, float, f, 32, 4, 11.2f, 3);
CHECK_RESULTS(TEST_MSG, "");
VECT_VAR_DECL(expected,poly,8,8) [] = { 0xf6, 0x33, 0x33, 0x33,
0x33, 0x33, 0x33, 0x33 };
VECT_VAR_DECL(expected,poly,16,4) [] = { 0xfff2, 0x3333, 0x3333, 0x3333 };
+VECT_VAR_DECL(expected,hfloat,16,4) [] = { 0xcb80, 0x3333, 0x3333, 0x3333 };
VECT_VAR_DECL(expected,hfloat,32,2) [] = { 0xc1700000, 0x33333333 };
VECT_VAR_DECL(expected,int,8,16) [] = { 0xff, 0x33, 0x33, 0x33,
0x33, 0x33, 0x33, 0x33,
0x33, 0x33, 0x33, 0x33 };
VECT_VAR_DECL(expected,poly,16,8) [] = { 0xfff4, 0x3333, 0x3333, 0x3333,
0x3333, 0x3333, 0x3333, 0x3333 };
+VECT_VAR_DECL(expected,hfloat,16,8) [] = { 0xc900, 0x3333, 0x3333, 0x3333,
+ 0x3333, 0x3333, 0x3333, 0x3333 };
VECT_VAR_DECL(expected,hfloat,32,4) [] = { 0xc1700000, 0x33333333,
0x33333333, 0x33333333 };
TEST_VST1_LANE(, uint, u, 64, 1, 0);
TEST_VST1_LANE(, poly, p, 8, 8, 6);
TEST_VST1_LANE(, poly, p, 16, 4, 2);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ TEST_VST1_LANE(, float, f, 16, 4, 1);
+#endif
TEST_VST1_LANE(, float, f, 32, 2, 1);
TEST_VST1_LANE(q, int, s, 8, 16, 15);
TEST_VST1_LANE(q, uint, u, 64, 2, 0);
TEST_VST1_LANE(q, poly, p, 8, 16, 10);
TEST_VST1_LANE(q, poly, p, 16, 8, 4);
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+ TEST_VST1_LANE(q, float, f, 16, 8, 6);
+#endif
TEST_VST1_LANE(q, float, f, 32, 4, 1);
CHECK_RESULTS(TEST_MSG, "");
projects / gcc.git / commitdiff