(__mmask8) __U,
__R);
}
-
-extern __inline __m256d
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_reduce_round_pd (__m256d __A, const int __C, const int __R)
-{
- return (__m256d) __builtin_ia32_reducepd256_mask_round ((__v4df) __A,
- __C,
- (__v4df)
- _mm256_setzero_pd (),
- (__mmask8) -1,
- __R);
-}
-
-extern __inline __m256d
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_mask_reduce_round_pd (__m256d __W, __mmask8 __U, __m256d __A,
- const int __C, const int __R)
-{
- return (__m256d) __builtin_ia32_reducepd256_mask_round ((__v4df) __A,
- __C,
- (__v4df) __W,
- (__mmask8) __U,
- __R);
-}
-
-extern __inline __m256d
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_maskz_reduce_round_pd (__mmask8 __U, __m256d __A, const int __C,
- const int __R)
-{
- return (__m256d) __builtin_ia32_reducepd256_mask_round ((__v4df) __A,
- __C,
- (__v4df)
- _mm256_setzero_pd (),
- (__mmask8) __U,
- __R);
-}
-
-extern __inline __m256h
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_reduce_round_ph (__m256h __A, const int __C, const int __R)
-{
- return (__m256h) __builtin_ia32_reduceph256_mask_round ((__v16hf) __A,
- __C,
- (__v16hf)
- _mm256_setzero_ph (),
- (__mmask16) -1,
- __R);
-}
-
-extern __inline __m256h
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_mask_reduce_round_ph (__m256h __W, __mmask16 __U, __m256h __A,
- const int __C, const int __R)
-{
- return (__m256h) __builtin_ia32_reduceph256_mask_round ((__v16hf) __A,
- __C,
- (__v16hf) __W,
- (__mmask16) __U,
- __R);
-}
-
-extern __inline __m256h
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_maskz_reduce_round_ph (__mmask16 __U, __m256h __A, const int __C,
- const int __R)
-{
- return (__m256h) __builtin_ia32_reduceph256_mask_round ((__v16hf) __A,
- __C,
- (__v16hf)
- _mm256_setzero_ph (),
- (__mmask16) __U,
- __R);
-}
-
-extern __inline __m256
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_reduce_round_ps (__m256 __A, const int __C, const int __R)
-{
- return (__m256) __builtin_ia32_reduceps256_mask_round ((__v8sf) __A,
- __C,
- (__v8sf)
- _mm256_setzero_ps (),
- (__mmask8) -1,
- __R);
-}
-
-extern __inline __m256
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_mask_reduce_round_ps (__m256 __W, __mmask8 __U, __m256 __A,
- const int __C, const int __R)
-{
- return (__m256) __builtin_ia32_reduceps256_mask_round ((__v8sf) __A,
- __C,
- (__v8sf) __W,
- (__mmask8) __U,
- __R);
-}
-
-extern __inline __m256
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_maskz_reduce_round_ps (__mmask8 __U, __m256 __A, const int __C,
- const int __R)
-{
- return (__m256) __builtin_ia32_reduceps256_mask_round ((__v8sf) __A,
- __C,
- (__v8sf)
- _mm256_setzero_ps (),
- (__mmask8) __U,
- __R);
-}
-
-extern __inline __m256d
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_roundscale_round_pd (__m256d __A, const int __C, const int __R)
-{
- return
- (__m256d) __builtin_ia32_rndscalepd256_mask_round ((__v4df) __A,
- __C,
- (__v4df)
- _mm256_undefined_pd (),
- (__mmask8) -1,
- __R);
-}
-
-extern __inline __m256d
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_mask_roundscale_round_pd (__m256d __W, __mmask8 __U, __m256d __A,
- const int __C, const int __R)
-{
- return (__m256d) __builtin_ia32_rndscalepd256_mask_round ((__v4df) __A,
- __C,
- (__v4df) __W,
- (__mmask8) __U,
- __R);
-}
-
-extern __inline __m256d
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_maskz_roundscale_round_pd (__mmask8 __U, __m256d __A, const int __C,
- const int __R)
-{
- return
- (__m256d) __builtin_ia32_rndscalepd256_mask_round ((__v4df) __A,
- __C,
- (__v4df)
- _mm256_setzero_pd (),
- (__mmask8) __U,
- __R);
-}
-
-extern __inline __m256h
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_roundscale_round_ph (__m256h __A, const int __C, const int __R)
-{
- return
- (__m256h) __builtin_ia32_rndscaleph256_mask_round ((__v16hf) __A,
- __C,
- (__v16hf)
- _mm256_undefined_ph (),
- (__mmask16) -1,
- __R);
-}
-
-extern __inline __m256h
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_mask_roundscale_round_ph (__m256h __W, __mmask16 __U, __m256h __A,
- const int __C, const int __R)
-{
- return (__m256h) __builtin_ia32_rndscaleph256_mask_round ((__v16hf) __A,
- __C,
- (__v16hf) __W,
- (__mmask16) __U,
- __R);
-}
-
-extern __inline __m256h
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_maskz_roundscale_round_ph (__mmask16 __U, __m256h __A, const int __C,
- const int __R)
-{
- return
- (__m256h) __builtin_ia32_rndscaleph256_mask_round ((__v16hf) __A,
- __C,
- (__v16hf)
- _mm256_setzero_ph (),
- (__mmask16) __U,
- __R);
-}
-
-extern __inline __m256
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_roundscale_round_ps (__m256 __A, const int __C, const int __R)
-{
- return
- (__m256) __builtin_ia32_rndscaleps256_mask_round ((__v8sf) __A,
- __C,
- (__v8sf)
- _mm256_undefined_ps (),
- (__mmask8) -1,
- __R);
-}
-
-extern __inline __m256
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_mask_roundscale_round_ps (__m256 __W, __mmask8 __U, __m256 __A,
- const int __C, const int __R)
-{
- return (__m256) __builtin_ia32_rndscaleps256_mask_round ((__v8sf) __A,
- __C,
- (__v8sf) __W,
- (__mmask8) __U,
- __R);
-}
-
-extern __inline __m256
-__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
-_mm256_maskz_roundscale_round_ps (__mmask8 __U, __m256 __A, const int __C,
- const int __R)
-{
- return (__m256) __builtin_ia32_rndscaleps256_mask_round ((__v8sf) __A,
- __C,
- (__v8sf)
- _mm256_setzero_ps (),
- (__mmask8) __U,
- __R);
-}
#else
#define _mm256_add_round_pd(A, B, R) \
((__m256d) __builtin_ia32_addpd256_mask_round ((__v4df) (A), \
(_mm256_setzero_ps ()), \
(__mmask8) (U), \
(R)))
-
-#define _mm256_reduce_round_pd(A, C, R) \
- ((__m256d) __builtin_ia32_reducepd256_mask_round ((__v4df) (A), \
- (C), \
- (__v4df) \
- (_mm256_setzero_pd ()), \
- (__mmask8) (-1), \
- (R)))
-
-#define _mm256_mask_reduce_round_pd(W, U, A, C, R) \
- ((__m256d) __builtin_ia32_reducepd256_mask_round ((__v4df) (A), \
- (C), \
- (__v4df) (W), \
- (__mmask8) (U), \
- (R)))
-
-#define _mm256_maskz_reduce_round_pd(U, A, C, R) \
- ((__m256d) __builtin_ia32_reducepd256_mask_round ((__v4df) (A), \
- (C), \
- (__v4df) \
- (_mm256_setzero_pd ()), \
- (__mmask8) (U), \
- (R)))
-
-#define _mm256_reduce_round_ph(A, C, R) \
- ((__m256h) __builtin_ia32_reduceph256_mask_round ((__v16hf) (A), \
- (C), \
- (__v16hf) \
- (_mm256_setzero_ph ()), \
- (__mmask16) (-1), \
- (R)))
-
-#define _mm256_mask_reduce_round_ph(W, U, A, C, R) \
- ((__m256h) __builtin_ia32_reduceph256_mask_round ((__v16hf) (A), \
- (C), \
- (__v16hf) (W), \
- (__mmask16) (U), \
- (R)))
-
-#define _mm256_maskz_reduce_round_ph(U, A, C, R) \
- ((__m256h) __builtin_ia32_reduceph256_mask_round ((__v16hf) (A), \
- (C), \
- (__v16hf) \
- (_mm256_setzero_ph ()), \
- (__mmask16) (U), \
- (R)))
-
-#define _mm256_reduce_round_ps(A, C, R) \
- ((__m256) __builtin_ia32_reduceps256_mask_round ((__v8sf) (A), \
- (C), \
- (__v8sf) \
- (_mm256_setzero_ps ()), \
- (__mmask8) (-1), \
- (R)))
-
-#define _mm256_mask_reduce_round_ps(W, U, A, C, R) \
- ((__m256) __builtin_ia32_reduceps256_mask_round ((__v8sf) (A), \
- (C), \
- (__v8sf) (W), \
- (__mmask8) (U), \
- (R)))
-
-#define _mm256_maskz_reduce_round_ps(U, A, C, R) \
- ((__m256) __builtin_ia32_reduceps256_mask_round ((__v8sf) (A), \
- (C), \
- (__v8sf) \
- (_mm256_setzero_ps ()), \
- (__mmask8) (U), \
- (R)))
-
-#define _mm256_roundscale_round_pd(A, C, R) \
- ((__m256d) \
- __builtin_ia32_rndscalepd256_mask_round ((__v4df) (A), \
- (C), \
- (__v4df) \
- (_mm256_undefined_pd ()), \
- (__mmask8) (-1), \
- (R)))
-
-#define _mm256_mask_roundscale_round_pd(W, U, A, C, R) \
- ((__m256d) __builtin_ia32_rndscalepd256_mask_round ((__v4df) (A), \
- (C), \
- (__v4df) (W), \
- (__mmask8) (U), \
- (R)))
-
-#define _mm256_maskz_roundscale_round_pd(U, A, C, R) \
- ((__m256d) __builtin_ia32_rndscalepd256_mask_round ((__v4df) (A), \
- (C), \
- (__v4df) \
- (_mm256_setzero_pd ()), \
- (__mmask8) (U), \
- (R)))
-
-#define _mm256_roundscale_round_ph(A, C, R) \
- ((__m256h) \
- __builtin_ia32_rndscaleph256_mask_round ((__v16hf) (A), \
- (C), \
- (__v16hf) \
- (_mm256_undefined_ph ()), \
- (__mmask16) (-1), \
- (R)))
-
-#define _mm256_mask_roundscale_round_ph(W, U, A, C, R) \
- ((__m256h) __builtin_ia32_rndscaleph256_mask_round ((__v16hf) (A), \
- (C), \
- (__v16hf) (W), \
- (__mmask16) (U), \
- (R)))
-
-#define _mm256_maskz_roundscale_round_ph(U, A, C, R) \
- ((__m256h) __builtin_ia32_rndscaleph256_mask_round ((__v16hf) (A), \
- (C), \
- (__v16hf) \
- (_mm256_setzero_ph ()), \
- (__mmask16) (U), \
- (R)))
-
-#define _mm256_roundscale_round_ps(A, C, R) \
- ((__m256) __builtin_ia32_rndscaleps256_mask_round ((__v8sf) (A), \
- (C), \
- (__v8sf) \
- (_mm256_undefined_ps ()), \
- (__mmask8) (-1), \
- (R)))
-
-#define _mm256_mask_roundscale_round_ps(W, U, A, C, R) \
- ((__m256) __builtin_ia32_rndscaleps256_mask_round ((__v8sf) (A), \
- (C), \
- (__v8sf) (W), \
- (__mmask8) (U), \
- (R)))
-
-#define _mm256_maskz_roundscale_round_ps(U, A, C, R) \
- ((__m256) __builtin_ia32_rndscaleps256_mask_round ((__v8sf) (A), \
- (C), \
- (__v8sf) \
- (_mm256_setzero_ps ()), \
- (__mmask8) (U), \
- (R)))
#endif
#define _mm256_cmul_round_pch(A, B, R) _mm256_fcmul_round_pch ((A), (B), (R))
BDESC (0, OPTION_MASK_ISA2_AVX10_2_256, CODE_FOR_mulv8sf3_mask_round, "__builtin_ia32_mulps256_mask_round", IX86_BUILTIN_VMULPS256_MASK_ROUND, UNKNOWN, (int) V8SF_FTYPE_V8SF_V8SF_V8SF_UQI_INT)
BDESC (0, OPTION_MASK_ISA2_AVX10_2_256, CODE_FOR_avx512dq_rangepv4df_mask_round, "__builtin_ia32_rangepd256_mask_round", IX86_BUILTIN_VRANGEPD256_MASK_ROUND, UNKNOWN, (int) V4DF_FTYPE_V4DF_V4DF_INT_V4DF_UQI_INT)
BDESC (0, OPTION_MASK_ISA2_AVX10_2_256, CODE_FOR_avx512dq_rangepv8sf_mask_round, "__builtin_ia32_rangeps256_mask_round", IX86_BUILTIN_VRANGEPS256_MASK_ROUND, UNKNOWN, (int) V8SF_FTYPE_V8SF_V8SF_INT_V8SF_UQI_INT)
-BDESC (0, OPTION_MASK_ISA2_AVX10_2_256, CODE_FOR_reducepv4df_mask_round, "__builtin_ia32_reducepd256_mask_round", IX86_BUILTIN_VREDUCEPD256_MASK_ROUND, UNKNOWN, (int) V4DF_FTYPE_V4DF_INT_V4DF_UQI_INT)
-BDESC (0, OPTION_MASK_ISA2_AVX10_2_256, CODE_FOR_reducepv16hf_mask_round, "__builtin_ia32_reduceph256_mask_round", IX86_BUILTIN_VREDUCEPH256_MASK_ROUND, UNKNOWN, (int) V16HF_FTYPE_V16HF_INT_V16HF_UHI_INT)
-BDESC (0, OPTION_MASK_ISA2_AVX10_2_256, CODE_FOR_reducepv8sf_mask_round, "__builtin_ia32_reduceps256_mask_round", IX86_BUILTIN_VREDUCEPS256_MASK_ROUND, UNKNOWN, (int) V8SF_FTYPE_V8SF_INT_V8SF_UQI_INT)
-BDESC (0, OPTION_MASK_ISA2_AVX10_2_256, CODE_FOR_avx512vl_rndscalev4df_mask_round, "__builtin_ia32_rndscalepd256_mask_round", IX86_BUILTIN_VRNDSCALEPD256_MASK_ROUND, UNKNOWN, (int) V4DF_FTYPE_V4DF_INT_V4DF_UQI_INT)
-BDESC (0, OPTION_MASK_ISA2_AVX10_2_256, CODE_FOR_avx512vl_rndscalev16hf_mask_round, "__builtin_ia32_rndscaleph256_mask_round", IX86_BUILTIN_VRNDSCALEPH256_MASK_ROUND, UNKNOWN, (int) V16HF_FTYPE_V16HF_INT_V16HF_UHI_INT)
-BDESC (0, OPTION_MASK_ISA2_AVX10_2_256, CODE_FOR_avx512vl_rndscalev8sf_mask_round, "__builtin_ia32_rndscaleps256_mask_round", IX86_BUILTIN_VRNDSCALEPS256_MASK_ROUND, UNKNOWN, (int) V8SF_FTYPE_V8SF_INT_V8SF_UQI_INT)
BDESC (0, OPTION_MASK_ISA2_AVX10_2_512, CODE_FOR_avx10_2_cvt2ps2phx_v32hf_mask_round, "__builtin_ia32_vcvt2ps2phx512_mask_round", IX86_BUILTIN_VCVT2PS2PHX_V32HF_MASK_ROUND, UNKNOWN, (int) V32HF_FTYPE_V16SF_V16SF_V32HF_USI_INT)
BDESC (0, OPTION_MASK_ISA2_AVX10_2_512, CODE_FOR_avx10_2_cvtph2ibsv32hf_mask_round, "__builtin_ia32_cvtph2ibs512_mask_round", IX86_BUILTIN_CVTPH2IBS512_MASK_ROUND, UNKNOWN, (int) V32HI_FTYPE_V32HF_V32HI_USI_INT)
BDESC (0, OPTION_MASK_ISA2_AVX10_2_512, CODE_FOR_avx10_2_cvtph2iubsv32hf_mask_round, "__builtin_ia32_cvtph2iubs512_mask_round", IX86_BUILTIN_CVTPH2IUBS512_MASK_ROUND, UNKNOWN, (int) V32HI_FTYPE_V32HF_V32HI_USI_INT)
[(match_operand:VFH_AVX512VL 1 "<round_saeonly_nimm_predicate>" "<round_saeonly_constraint>")
(match_operand:SI 2 "const_0_to_255_operand")]
UNSPEC_REDUCE))]
- "(TARGET_AVX512DQ || (VALID_AVX512FP16_REG_MODE (<MODE>mode))) && <round_saeonly_mode_condition>"
+ "TARGET_AVX512DQ || (VALID_AVX512FP16_REG_MODE (<MODE>mode))"
"vreduce<ssemodesuffix>\t{%2, <round_saeonly_mask_op3>%1, %0<mask_operand3>|%0<mask_operand3>, %1<round_saeonly_mask_op3>, %2}"
[(set_attr "type" "sse")
(set_attr "prefix" "evex")
[(match_operand:VFH_AVX512VL 1 "nonimmediate_operand" "<round_saeonly_constraint>")
(match_operand:SI 2 "const_0_to_255_operand")]
UNSPEC_ROUND))]
- "TARGET_AVX512F && <round_saeonly_mode_condition>"
+ "TARGET_AVX512F"
"vrndscale<ssemodesuffix>\t{%2, <round_saeonly_mask_op3>%1, %0<mask_operand3>|%0<mask_operand3>, %1<round_saeonly_mask_op3>, %2}"
[(set_attr "length_immediate" "1")
(set_attr "prefix" "evex")
#define __builtin_ia32_mulps256_mask_round(A, B, C, D, E) __builtin_ia32_mulps256_mask_round(A, B, C, D, 8)
#define __builtin_ia32_rangeps256_mask_round(A, B, I, D, E, F) __builtin_ia32_rangeps256_mask_round(A, B, 1, D, E, 8)
#define __builtin_ia32_rangepd256_mask_round(A, B, I, D, E, F) __builtin_ia32_rangepd256_mask_round(A, B, 1, D, E, 8)
-#define __builtin_ia32_reducepd256_mask_round(A, B, C, D, E) __builtin_ia32_reducepd256_mask_round(A, 8, C, D, 8)
-#define __builtin_ia32_reduceph256_mask_round(A, B, C, D, E) __builtin_ia32_reduceph256_mask_round(A, 8, C, D, 8)
-#define __builtin_ia32_reduceps256_mask_round(A, B, C, D, E) __builtin_ia32_reduceps256_mask_round(A, 8, C, D, 8)
-#define __builtin_ia32_rndscalepd256_mask_round(A, B, C, D, E) __builtin_ia32_rndscalepd256_mask_round(A, 1, C, D, 8)
-#define __builtin_ia32_rndscaleph256_mask_round(A, B, C, D, E) __builtin_ia32_rndscaleph256_mask_round(A, 1, C, D, 8)
-#define __builtin_ia32_rndscaleps256_mask_round(A, B, C, D, E) __builtin_ia32_rndscaleps256_mask_round(A, 1, C, D, 8)
/* avx10_2-512mediaintrin.h */
#define __builtin_ia32_mpsadbw512(A, B, C) __builtin_ia32_mpsadbw512 (A, B, 1)
/* { dg-final { scan-assembler-times "vrangeps\[ \\t\]+\[^\$\n\]*\\$\[^\{\n\]*\{sae\}\[^\n\]*%ymm\[0-9\]+(?:\n|\[ \\t\]+#)" 1 } } */
/* { dg-final { scan-assembler-times "vrangeps\[ \\t\]+\[^\$\n\]*\\$\[^\{\n\]*\{sae\}\[^\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}(?:\n|\[ \\t\]+#)" 1 } } */
/* { dg-final { scan-assembler-times "vrangeps\[ \\t\]+\[^\$\n\]*\\$\[^\{\n\]*\{sae\}\[^\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}\{z\}(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vreducepd\[ \\t\]+\[^\{\n\]*\{sae\}\[^\n\]*%ymm\[0-9\]+(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vreducepd\[ \\t\]+\[^\{\n\]*\{sae\}\[^\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vreducepd\[ \\t\]+\[^\{\n\]*\{sae\}\[^\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}\{z\}(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vreduceph\[ \\t\]+\[^\{\n\]*\{sae\}\[^\{\n\]*%ymm\[0-9\]+(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vreduceph\[ \\t\]+\[^\{\n\]*\{sae\}\[^\{\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vreduceph\[ \\t\]+\[^\{\n\]*\{sae\}\[^\{\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}\{z\}(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vreduceps\[ \\t\]+\[^\{\n\]*\{sae\}\[^\n\]*%ymm\[0-9\]+(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vreduceps\[ \\t\]+\[^\{\n\]*\{sae\}\[^\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vreduceps\[ \\t\]+\[^\{\n\]*\{sae\}\[^\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}\{z\}(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vrndscalepd\[ \\t\]+\\S*,\[ \\t\]+\{sae\}\[^\n\]*%ymm\[0-9\]+\[^\n\]*%ymm\[0-9\]+(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vrndscalepd\[ \\t\]+\\S*,\[ \\t\]+\{sae\}\[^\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vrndscalepd\[ \\t\]+\\S*,\[ \\t\]+\{sae\}\[^\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}\{z\}(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vrndscaleph\[ \\t\]+\[^\n\]*\{sae\}\[^\{\n\]*%ymm\[0-9\]+(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vrndscaleph\[ \\t\]+\[^\n\]*\{sae\}\[^\{\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vrndscaleph\[ \\t\]+\[^\n\]*\{sae\}\[^\{\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}\{z\}(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vrndscaleps\[ \\t\]+\\S*,\[ \\t\]+\{sae\}\[^\n\]*%ymm\[0-9\]+\[^\n\]*%ymm\[0-9\]+(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vrndscaleps\[ \\t\]+\\S*,\[ \\t\]+\{sae\}\[^\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}(?:\n|\[ \\t\]+#)" 1 } } */
-/* { dg-final { scan-assembler-times "vrndscaleps\[ \\t\]+\\S*,\[ \\t\]+\{sae\}\[^\n\]*%ymm\[0-9\]+\{%k\[1-7\]\}\{z\}(?:\n|\[ \\t\]+#)" 1 } } */
#include <immintrin.h>
x = _mm256_mask_range_round_ps (x, m16, x, x, 15, _MM_FROUND_NO_EXC);
x = _mm256_maskz_range_round_ps (m16, x, x, 15, _MM_FROUND_NO_EXC);
}
-
-void extern
-avx10_2_test_22 (void)
-{
- xd = _mm256_reduce_round_pd (xd, 123, _MM_FROUND_NO_EXC);
- xd = _mm256_mask_reduce_round_pd (xd, m8, xd, 123, _MM_FROUND_NO_EXC);
- xd = _mm256_maskz_reduce_round_pd (m8, xd, 123, _MM_FROUND_NO_EXC);
-
- xh = _mm256_reduce_round_ph (xh, 123, _MM_FROUND_NO_EXC);
- xh = _mm256_mask_reduce_round_ph (xh, m16, xh, 123, _MM_FROUND_NO_EXC);
- xh = _mm256_maskz_reduce_round_ph (m16, xh, 123, _MM_FROUND_NO_EXC);
-
- x = _mm256_reduce_round_ps (x, 123, _MM_FROUND_NO_EXC);
- x = _mm256_mask_reduce_round_ps (x, m8, x, 123, _MM_FROUND_NO_EXC);
- x = _mm256_maskz_reduce_round_ps (m8, x, 123, _MM_FROUND_NO_EXC);
-}
-
-void extern
-avx10_2_test_23 (void)
-{
- xd = _mm256_roundscale_round_pd (xd, 0x42, _MM_FROUND_NO_EXC);
- xd = _mm256_mask_roundscale_round_pd (xd, 2, xd, 0x42, _MM_FROUND_NO_EXC);
- xd = _mm256_maskz_roundscale_round_pd (2, xd, 0x42, _MM_FROUND_NO_EXC);
-
- xh = _mm256_roundscale_round_ph (xh, 123, 8);
- xh = _mm256_mask_roundscale_round_ph (xh, m16, xh, 123, 8);
- xh = _mm256_maskz_roundscale_round_ph (m16, xh, 123, 8);
-
- x = _mm256_roundscale_round_ps (x, 0x42, _MM_FROUND_NO_EXC);
- x = _mm256_mask_roundscale_round_ps (x, 2, x, 0x42, _MM_FROUND_NO_EXC);
- x = _mm256_maskz_roundscale_round_ps (2, x, 0x42, _MM_FROUND_NO_EXC);
-}
#define __builtin_ia32_mulps256_mask_round(A, B, C, D, E) __builtin_ia32_mulps256_mask_round(A, B, C, D, 8)
#define __builtin_ia32_rangeps256_mask_round(A, B, C, D, E, F) __builtin_ia32_rangeps256_mask_round(A, B, 1, D, E, 8)
#define __builtin_ia32_rangepd256_mask_round(A, B, C, D, E, F) __builtin_ia32_rangepd256_mask_round(A, B, 1, D, E, 8)
-#define __builtin_ia32_reducepd256_mask_round(A, B, C, D, E) __builtin_ia32_reducepd256_mask_round(A, 8, C, D, 8)
-#define __builtin_ia32_reduceph256_mask_round(A, B, C, D, E) __builtin_ia32_reduceph256_mask_round(A, 8, C, D, 8)
-#define __builtin_ia32_reduceps256_mask_round(A, B, C, D, E) __builtin_ia32_reduceps256_mask_round(A, 8, C, D, 8)
-#define __builtin_ia32_rndscalepd256_mask_round(A, B, C, D, E) __builtin_ia32_rndscalepd256_mask_round(A, 1, C, D, 8)
-#define __builtin_ia32_rndscaleph256_mask_round(A, B, C, D, E) __builtin_ia32_rndscaleph256_mask_round(A, 1, C, D, 8)
-#define __builtin_ia32_rndscaleps256_mask_round(A, B, C, D, E) __builtin_ia32_rndscaleps256_mask_round(A, 1, C, D, 8)
/* avx10_2-512mediaintrin.h */
#define __builtin_ia32_mpsadbw512(A, B, C) __builtin_ia32_mpsadbw512 (A, B, 1)
test_1 (_mm256_getexp_round_pd, __m256d, __m256d, 8)
test_1 (_mm256_getexp_round_ph, __m256h, __m256h, 8)
test_1 (_mm256_getexp_round_ps, __m256, __m256, 8)
-test_1x (_mm256_reduce_round_ph, __m256h, __m256h, 123, 8)
-test_1x (_mm256_reduce_round_ps, __m256, __m256, 123, 8)
-test_1x (_mm256_reduce_round_pd, __m256d, __m256d, 123, 8)
-test_1x (_mm256_roundscale_round_ph, __m256h, __m256h, 123, 8)
-test_1x (_mm256_roundscale_round_ps, __m256, __m256, 123, 8)
-test_1x (_mm256_roundscale_round_pd, __m256d, __m256d, 123, 8)
test_1y (_mm256_getmant_round_ph, __m256h, __m256h, 1, 1, 8)
test_1y (_mm256_getmant_round_ps, __m256, __m256, 1, 1, 8)
test_1y (_mm256_getmant_round_pd, __m256d, __m256d, 1, 1, 8)
test_2x (_mm256_cmp_round_ps_mask, __mmask8, __m256, __m256, 1, 8)
test_2x (_mm256_range_round_pd, __m256d, __m256d, __m256d, 15, 8)
test_2x (_mm256_range_round_ps, __m256, __m256, __m256, 15, 8)
-test_2x (_mm256_maskz_reduce_round_pd, __m256d, __mmask8, __m256d, 123, 8)
-test_2x (_mm256_maskz_reduce_round_ph, __m256h, __mmask8, __m256h, 123, 8)
-test_2x (_mm256_maskz_reduce_round_ps, __m256, __mmask16, __m256, 123, 8)
-test_2x (_mm256_maskz_roundscale_round_pd, __m256d, __mmask8, __m256d, 123, 8)
-test_2x (_mm256_maskz_roundscale_round_ph, __m256h, __mmask8, __m256h, 123, 8)
-test_2x (_mm256_maskz_roundscale_round_ps, __m256, __mmask16, __m256, 123, 8)
test_2y (_mm256_maskz_getmant_round_pd, __m256d, __mmask8, __m256d, 1, 1, 8)
test_2y (_mm256_maskz_getmant_round_ph, __m256h, __mmask16, __m256h, 1, 1, 8)
test_2y (_mm256_maskz_getmant_round_ps, __m256, __mmask8, __m256, 1, 1, 8)
test_3x (_mm256_fixupimm_round_ps, __m256, __m256, __m256, __m256i, 3, 8)
test_3x (_mm256_maskz_range_round_pd, __m256d, __mmask8, __m256d, __m256d, 15, 8)
test_3x (_mm256_maskz_range_round_ps, __m256, __mmask8, __m256, __m256, 15, 8)
-test_3x (_mm256_mask_reduce_round_ph, __m256h, __m256h, __mmask8, __m256h, 123, 8)
-test_3x (_mm256_mask_reduce_round_ps, __m256, __m256, __mmask16, __m256, 123, 8)
-test_3x (_mm256_mask_reduce_round_pd, __m256d, __m256d, __mmask8, __m256d, 123, 8)
-test_3x (_mm256_mask_roundscale_round_ph, __m256h, __m256h, __mmask8, __m256h, 123, 8)
-test_3x (_mm256_mask_roundscale_round_ps, __m256, __m256, __mmask16, __m256, 123, 8)
-test_3x (_mm256_mask_roundscale_round_pd, __m256d, __m256d, __mmask8, __m256d, 123, 8)
test_3y (_mm256_mask_getmant_round_pd, __m256d, __m256d, __mmask8, __m256d, 1, 1, 8)
test_3y (_mm256_mask_getmant_round_ph, __m256h, __m256h, __mmask16, __m256h, 1, 1, 8)
test_3y (_mm256_mask_getmant_round_ps, __m256, __m256, __mmask8, __m256, 1, 1, 8)
test_1 (_mm256_getexp_round_pd, __m256d, __m256d, 8)
test_1 (_mm256_getexp_round_ph, __m256h, __m256h, 8)
test_1 (_mm256_getexp_round_ps, __m256, __m256, 8)
-test_1x (_mm256_reduce_round_ph, __m256h, __m256h, 123, 8)
-test_1x (_mm256_reduce_round_ps, __m256, __m256, 123, 8)
-test_1x (_mm256_reduce_round_pd, __m256d, __m256d, 123, 8)
-test_1x (_mm256_roundscale_round_ph, __m256h, __m256h, 123, 8)
-test_1x (_mm256_roundscale_round_ps, __m256, __m256, 123, 8)
-test_1x (_mm256_roundscale_round_pd, __m256d, __m256d, 123, 8)
test_1y (_mm256_getmant_round_ph, __m256h, __m256h, 1, 1, 8)
test_1y (_mm256_getmant_round_ps, __m256, __m256, 1, 1, 8)
test_1y (_mm256_getmant_round_pd, __m256d, __m256d, 1, 1, 8)
test_2x (_mm256_cmp_round_ps_mask, __mmask8, __m256, __m256, 1, 8)
test_2x (_mm256_range_round_pd, __m256d, __m256d, __m256d, 15, 8)
test_2x (_mm256_range_round_ps, __m256, __m256, __m256, 15, 8)
-test_2x (_mm256_maskz_reduce_round_pd, __m256d, __mmask8, __m256d, 123, 8)
-test_2x (_mm256_maskz_reduce_round_ph, __m256h, __mmask8, __m256h, 123, 8)
-test_2x (_mm256_maskz_reduce_round_ps, __m256, __mmask16, __m256, 123, 8)
-test_2x (_mm256_maskz_roundscale_round_pd, __m256d, __mmask8, __m256d, 123, 8)
-test_2x (_mm256_maskz_roundscale_round_ph, __m256h, __mmask8, __m256h, 123, 8)
-test_2x (_mm256_maskz_roundscale_round_ps, __m256, __mmask16, __m256, 123, 8)
test_2y (_mm256_maskz_getmant_round_pd, __m256d, __mmask8, __m256d, 1, 1, 8)
test_2y (_mm256_maskz_getmant_round_ph, __m256h, __mmask16, __m256h, 1, 1, 8)
test_2y (_mm256_maskz_getmant_round_ps, __m256, __mmask8, __m256, 1, 1, 8)
test_3x (_mm256_fixupimm_round_ps, __m256, __m256, __m256, __m256i, 3, 8)
test_3x (_mm256_maskz_range_round_pd, __m256d, __mmask8, __m256d, __m256d, 15, 8)
test_3x (_mm256_maskz_range_round_ps, __m256, __mmask16, __m256, __m256, 15, 8)
-test_3x (_mm256_mask_reduce_round_ph, __m256h, __m256h, __mmask8, __m256h, 123, 8)
-test_3x (_mm256_mask_reduce_round_ps, __m256, __m256, __mmask16, __m256, 123, 8)
-test_3x (_mm256_mask_reduce_round_pd, __m256d, __m256d, __mmask8, __m256d, 123, 8)
-test_3x (_mm256_mask_roundscale_round_ph, __m256h, __m256h, __mmask8, __m256h, 123, 8)
-test_3x (_mm256_mask_roundscale_round_ps, __m256, __m256, __mmask16, __m256, 123, 8)
-test_3x (_mm256_mask_roundscale_round_pd, __m256d, __m256d, __mmask8, __m256d, 123, 8)
test_3y (_mm256_mask_getmant_round_pd, __m256d, __m256d, __mmask8, __m256d, 1, 1, 8)
test_3y (_mm256_mask_getmant_round_ph, __m256h, __m256h, __mmask16, __m256h, 1, 1, 8)
test_3y (_mm256_mask_getmant_round_ps, __m256, __m256, __mmask8, __m256, 1, 1, 8)
#define __builtin_ia32_mulps256_mask_round(A, B, C, D, E) __builtin_ia32_mulps256_mask_round(A, B, C, D, 8)
#define __builtin_ia32_rangeps256_mask_round(A, B, C, D, E, F) __builtin_ia32_rangeps256_mask_round(A, B, 1, D, E, 8)
#define __builtin_ia32_rangepd256_mask_round(A, B, C, D, E, F) __builtin_ia32_rangepd256_mask_round(A, B, 1, D, E, 8)
-#define __builtin_ia32_reducepd256_mask_round(A, B, C, D, E) __builtin_ia32_reducepd256_mask_round(A, 8, C, D, 8)
-#define __builtin_ia32_reduceph256_mask_round(A, B, C, D, E) __builtin_ia32_reduceph256_mask_round(A, 8, C, D, 8)
-#define __builtin_ia32_reduceps256_mask_round(A, B, C, D, E) __builtin_ia32_reduceps256_mask_round(A, 8, C, D, 8)
-#define __builtin_ia32_rndscalepd256_mask_round(A, B, C, D, E) __builtin_ia32_rndscalepd256_mask_round(A, 1, C, D, 8)
-#define __builtin_ia32_rndscaleph256_mask_round(A, B, C, D, E) __builtin_ia32_rndscaleph256_mask_round(A, 1, C, D, 8)
-#define __builtin_ia32_rndscaleps256_mask_round(A, B, C, D, E) __builtin_ia32_rndscaleps256_mask_round(A, 1, C, D, 8)
/* avx10_2-512mediaintrin.h */
#define __builtin_ia32_mpsadbw512(A, B, C) __builtin_ia32_mpsadbw512 (A, B, 1)
projects / thirdparty / gcc.git / commitdiff
