From: "Liu, Hongtao" <hongtao.liu@intel.com>
To: "Hu, Lin1" <lin1.hu@intel.com>, Hongtao Liu <crazylht@gmail.com>
Cc: "gcc-patches@gcc.gnu.org" <gcc-patches@gcc.gnu.org>,
"ubizjak@gmail.com" <ubizjak@gmail.com>
Subject: RE: [PATCH] i386: Fix incorrect intrinsic signature for AVX512 s{lli|rai|rli}
Date: Thu, 25 May 2023 08:33:18 +0000 [thread overview]
Message-ID: <SA1PR11MB675758BDFE74E22E82BD17F3E5469@SA1PR11MB6757.namprd11.prod.outlook.com> (raw)
In-Reply-To: <SJ0PR11MB59401FA16B5B598C535BF1A1A6469@SJ0PR11MB5940.namprd11.prod.outlook.com>
> -----Original Message-----
> From: Hu, Lin1 <lin1.hu@intel.com>
> Sent: Thursday, May 25, 2023 3:52 PM
> To: Hongtao Liu <crazylht@gmail.com>
> Cc: gcc-patches@gcc.gnu.org; Liu, Hongtao <hongtao.liu@intel.com>;
> ubizjak@gmail.com
> Subject: RE: [PATCH] i386: Fix incorrect intrinsic signature for AVX512
> s{lli|rai|rli}
>
> OK, I update the change log and modify a part of format. The attached file is
> the new version.
LGTM.
>
> -----Original Message-----
> From: Hongtao Liu <crazylht@gmail.com>
> Sent: Thursday, May 25, 2023 11:40 AM
> To: Hu, Lin1 <lin1.hu@intel.com>
> Cc: gcc-patches@gcc.gnu.org; Liu, Hongtao <hongtao.liu@intel.com>;
> ubizjak@gmail.com
> Subject: Re: [PATCH] i386: Fix incorrect intrinsic signature for AVX512
> s{lli|rai|rli}
>
> On Thu, May 25, 2023 at 10:55 AM Hu, Lin1 via Gcc-patches
> <gcc-patches@gcc.gnu.org> wrote:
> >
> > Hi all,
> >
> > This patch aims to fix incorrect intrinsic signature for
> _mm{512|256|}_s{lli|rai|rli}_epi*. And it has been tested on x86_64-pc-
> linux-gnu. OK for trunk?
> >
> > BRs,
> > Lin
> >
> > gcc/ChangeLog:
> >
> > PR target/109173
> > PR target/109174
> > * config/i386/avx512bwintrin.h (_mm512_srli_epi16): Change type
> from
> > int to const int.
> int to unsigned int or const int to const unsigned int.
> Others LGTM.
> > (_mm512_mask_srli_epi16): Ditto.
> > (_mm512_slli_epi16): Ditto.
> > (_mm512_mask_slli_epi16): Ditto.
> > (_mm512_maskz_slli_epi16): Ditto.
> > (_mm512_srai_epi16): Ditto.
> > (_mm512_mask_srai_epi16): Ditto.
> > (_mm512_maskz_srai_epi16): Ditto.
> > * config/i386/avx512vlintrin.h (_mm256_mask_srli_epi32): Ditto.
> > (_mm256_maskz_srli_epi32): Ditto.
> > (_mm_mask_srli_epi32): Ditto.
> > (_mm_maskz_srli_epi32): Ditto.
> > (_mm256_mask_srli_epi64): Ditto.
> > (_mm256_maskz_srli_epi64): Ditto.
> > (_mm_mask_srli_epi64): Ditto.
> > (_mm_maskz_srli_epi64): Ditto.
> > (_mm256_mask_srai_epi32): Ditto.
> > (_mm256_maskz_srai_epi32): Ditto.
> > (_mm_mask_srai_epi32): Ditto.
> > (_mm_maskz_srai_epi32): Ditto.
> > (_mm256_srai_epi64): Ditto.
> > (_mm256_mask_srai_epi64): Ditto.
> > (_mm256_maskz_srai_epi64): Ditto.
> > (_mm_srai_epi64): Ditto.
> > (_mm_mask_srai_epi64): Ditto.
> > (_mm_maskz_srai_epi64): Ditto.
> > (_mm_mask_slli_epi32): Ditto.
> > (_mm_maskz_slli_epi32): Ditto.
> > (_mm_mask_slli_epi64): Ditto.
> > (_mm_maskz_slli_epi64): Ditto.
> > (_mm256_mask_slli_epi32): Ditto.
> > (_mm256_maskz_slli_epi32): Ditto.
> > (_mm256_mask_slli_epi64): Ditto.
> > (_mm256_maskz_slli_epi64): Ditto.
> > (_mm_mask_srai_epi16): Ditto.
> > (_mm_maskz_srai_epi16): Ditto.
> > (_mm256_srai_epi16): Ditto.
> > (_mm256_mask_srai_epi16): Ditto.
> > (_mm_mask_slli_epi16): Ditto.
> > (_mm_maskz_slli_epi16): Ditto.
> > (_mm256_mask_slli_epi16): Ditto.
> > (_mm256_maskz_slli_epi16): Ditto.
> >
> > gcc/testsuite/ChangeLog:
> >
> > PR target/109173
> > PR target/109174
> > * gcc.target/i386/pr109173-1.c: New test.
> > * gcc.target/i386/pr109174-1.c: Ditto.
> > ---
> > gcc/config/i386/avx512bwintrin.h | 32 +++---
> > gcc/config/i386/avx512fintrin.h | 58 +++++++----
> > gcc/config/i386/avx512vlbwintrin.h | 36 ++++---
> > gcc/config/i386/avx512vlintrin.h | 112 +++++++++++----------
> > gcc/testsuite/gcc.target/i386/pr109173-1.c | 57 +++++++++++
> > gcc/testsuite/gcc.target/i386/pr109174-1.c | 45 +++++++++
> > 6 files changed, 236 insertions(+), 104 deletions(-)
> > create mode 100644 gcc/testsuite/gcc.target/i386/pr109173-1.c
> > create mode 100644 gcc/testsuite/gcc.target/i386/pr109174-1.c
> >
> > diff --git a/gcc/config/i386/avx512bwintrin.h
> b/gcc/config/i386/avx512bwintrin.h
> > index 89790f7917b..791d4e35f32 100644
> > --- a/gcc/config/i386/avx512bwintrin.h
> > +++ b/gcc/config/i386/avx512bwintrin.h
> > @@ -2880,7 +2880,7 @@ _mm512_maskz_dbsad_epu8 (__mmask32 __U,
> __m512i __A, __m512i __B,
> >
> > extern __inline __m512i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm512_srli_epi16 (__m512i __A, const int __imm)
> > +_mm512_srli_epi16 (__m512i __A, const unsigned int __imm)
> > {
> > return (__m512i) __builtin_ia32_psrlwi512_mask ((__v32hi) __A, __imm,
> > (__v32hi)
> > @@ -2891,7 +2891,7 @@ _mm512_srli_epi16 (__m512i __A, const int
> __imm)
> > extern __inline __m512i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm512_mask_srli_epi16 (__m512i __W, __mmask32 __U, __m512i __A,
> > - const int __imm)
> > + const unsigned int __imm)
> > {
> > return (__m512i) __builtin_ia32_psrlwi512_mask ((__v32hi) __A, __imm,
> > (__v32hi) __W,
> > @@ -2910,7 +2910,7 @@ _mm512_maskz_srli_epi16 (__mmask32 __U,
> __m512i __A, const int __imm)
> >
> > extern __inline __m512i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm512_slli_epi16 (__m512i __A, const int __B)
> > +_mm512_slli_epi16 (__m512i __A, const unsigned int __B)
> > {
> > return (__m512i) __builtin_ia32_psllwi512_mask ((__v32hi) __A, __B,
> > (__v32hi)
> > @@ -2921,7 +2921,7 @@ _mm512_slli_epi16 (__m512i __A, const int __B)
> > extern __inline __m512i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm512_mask_slli_epi16 (__m512i __W, __mmask32 __U, __m512i __A,
> > - const int __B)
> > + const unsigned int __B)
> > {
> > return (__m512i) __builtin_ia32_psllwi512_mask ((__v32hi) __A, __B,
> > (__v32hi) __W,
> > @@ -2930,7 +2930,7 @@ _mm512_mask_slli_epi16 (__m512i __W,
> __mmask32 __U, __m512i __A,
> >
> > extern __inline __m512i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm512_maskz_slli_epi16 (__mmask32 __U, __m512i __A, const int __B)
> > +_mm512_maskz_slli_epi16 (__mmask32 __U, __m512i __A, const
> unsigned int __B)
> > {
> > return (__m512i) __builtin_ia32_psllwi512_mask ((__v32hi) __A, __B,
> > (__v32hi)
> > @@ -3008,7 +3008,7 @@ _mm512_maskz_shufflelo_epi16 (__mmask32
> __U, __m512i __A,
> >
> > extern __inline __m512i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm512_srai_epi16 (__m512i __A, const int __imm)
> > +_mm512_srai_epi16 (__m512i __A, const unsigned int __imm)
> > {
> > return (__m512i) __builtin_ia32_psrawi512_mask ((__v32hi) __A, __imm,
> > (__v32hi)
> > @@ -3019,7 +3019,7 @@ _mm512_srai_epi16 (__m512i __A, const int
> __imm)
> > extern __inline __m512i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm512_mask_srai_epi16 (__m512i __W, __mmask32 __U, __m512i __A,
> > - const int __imm)
> > + const unsigned int __imm)
> > {
> > return (__m512i) __builtin_ia32_psrawi512_mask ((__v32hi) __A, __imm,
> > (__v32hi) __W,
> > @@ -3028,7 +3028,7 @@ _mm512_mask_srai_epi16 (__m512i __W,
> __mmask32 __U, __m512i __A,
> >
> > extern __inline __m512i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm512_maskz_srai_epi16 (__mmask32 __U, __m512i __A, const int
> __imm)
> > +_mm512_maskz_srai_epi16 (__mmask32 __U, __m512i __A, const
> unsigned int __imm)
> > {
> > return (__m512i) __builtin_ia32_psrawi512_mask ((__v32hi) __A, __imm,
> > (__v32hi)
> > @@ -3196,28 +3196,28 @@ _mm512_bsrli_epi128 (__m512i __A, const int
> __N)
> >
> > #define _mm512_srli_epi16(A, B) \
> > ((__m512i) __builtin_ia32_psrlwi512_mask ((__v32hi)(__m512i)(A), \
> > - (int)(B), (__v32hi)_mm512_setzero_si512 (), (__mmask32)-1))
> > + (unsigned int)(B), (__v32hi)_mm512_setzero_si512 (), (__mmask32)-1))
> >
> > #define _mm512_mask_srli_epi16(W, U, A, B) \
> > ((__m512i) __builtin_ia32_psrlwi512_mask ((__v32hi)(__m512i)(A), \
> > - (int)(B), (__v32hi)(__m512i)(W), (__mmask32)(U)))
> > + (unsigned int)(B), (__v32hi)(__m512i)(W), (__mmask32)(U)))
> >
> > #define _mm512_maskz_srli_epi16(U, A, B) \
> > ((__m512i) __builtin_ia32_psrlwi512_mask ((__v32hi)(__m512i)(A), \
> > (int)(B), (__v32hi)_mm512_setzero_si512 (), (__mmask32)(U)))
> >
> > #define _mm512_slli_epi16(X, C) \
> > - ((__m512i)__builtin_ia32_psllwi512_mask ((__v32hi)(__m512i)(X),
> (int)(C),\
> > + ((__m512i)__builtin_ia32_psllwi512_mask ((__v32hi)(__m512i)(X),
> (unsigned int)(C),\
> > (__v32hi)(__m512i)_mm512_setzero_si512 (), \
> > (__mmask32)-1))
> >
> > #define _mm512_mask_slli_epi16(W, U, X, C) \
> > - ((__m512i)__builtin_ia32_psllwi512_mask ((__v32hi)(__m512i)(X),
> (int)(C),\
> > + ((__m512i)__builtin_ia32_psllwi512_mask ((__v32hi)(__m512i)(X),
> (unsigned int)(C),\
> > (__v32hi)(__m512i)(W),\
> > (__mmask32)(U)))
> >
> > #define _mm512_maskz_slli_epi16(U, X, C) \
> > - ((__m512i)__builtin_ia32_psllwi512_mask ((__v32hi)(__m512i)(X),
> (int)(C),\
> > + ((__m512i)__builtin_ia32_psllwi512_mask ((__v32hi)(__m512i)(X),
> (unsigned int)(C),\
> > (__v32hi)(__m512i)_mm512_setzero_si512 (), \
> > (__mmask32)(U)))
> >
> > @@ -3257,15 +3257,15 @@ _mm512_bsrli_epi128 (__m512i __A, const int
> __N)
> >
> > #define _mm512_srai_epi16(A, B) \
> > ((__m512i) __builtin_ia32_psrawi512_mask ((__v32hi)(__m512i)(A), \
> > - (int)(B), (__v32hi)_mm512_setzero_si512 (), (__mmask32)-1))
> > + (unsigned int)(B), (__v32hi)_mm512_setzero_si512 (), (__mmask32)-1))
> >
> > #define _mm512_mask_srai_epi16(W, U, A, B) \
> > ((__m512i) __builtin_ia32_psrawi512_mask ((__v32hi)(__m512i)(A), \
> > - (int)(B), (__v32hi)(__m512i)(W), (__mmask32)(U)))
> > + (unsigned int)(B), (__v32hi)(__m512i)(W), (__mmask32)(U)))
> >
> > #define _mm512_maskz_srai_epi16(U, A, B) \
> > ((__m512i) __builtin_ia32_psrawi512_mask ((__v32hi)(__m512i)(A), \
> > - (int)(B), (__v32hi)_mm512_setzero_si512 (), (__mmask32)(U)))
> > + (unsigned int)(B), (__v32hi)_mm512_setzero_si512 (),
> (__mmask32)(U)))
> >
> > #define _mm512_mask_blend_epi16(__U, __A, __W) \
> > ((__m512i) __builtin_ia32_blendmw_512_mask ((__v32hi) (__A), \
> > diff --git a/gcc/config/i386/avx512fintrin.h
> b/gcc/config/i386/avx512fintrin.h
> > index 89b321970cc..277260c3655 100644
> > --- a/gcc/config/i386/avx512fintrin.h
> > +++ b/gcc/config/i386/avx512fintrin.h
> > @@ -1038,17 +1038,20 @@ _mm512_maskz_slli_epi64 (__mmask8 __U,
> __m512i __A, unsigned int __B)
> > }
> > #else
> > #define _mm512_slli_epi64(X, C) \
> > - ((__m512i) __builtin_ia32_psllqi512_mask ((__v8di)(__m512i)(X),
> (int)(C),\
> > - (__v8di)(__m512i)_mm512_undefined_epi32 (),\
> > + ((__m512i) __builtin_ia32_psllqi512_mask ((__v8di)(__m512i)(X),
> \
> > + (unsigned int)(C), \
> > + (__v8di)(__m512i)_mm512_undefined_epi32 (),
> \
> > (__mmask8)-1))
> >
> > #define _mm512_mask_slli_epi64(W, U, X, C) \
> > - ((__m512i) __builtin_ia32_psllqi512_mask ((__v8di)(__m512i)(X),
> (int)(C),\
> > - (__v8di)(__m512i)(W),\
> > + ((__m512i) __builtin_ia32_psllqi512_mask ((__v8di)(__m512i)(X),
> \
> > + (unsigned int)(C), \
> > + (__v8di)(__m512i)(W), \
> > (__mmask8)(U)))
> >
> > #define _mm512_maskz_slli_epi64(U, X, C) \
> > - ((__m512i) __builtin_ia32_psllqi512_mask ((__v8di)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_psllqi512_mask ((__v8di)(__m512i)(X),
> \
> > + (unsigned int)(C),
> \
> > (__v8di)(__m512i)_mm512_setzero_si512 (),\
> > (__mmask8)(U)))
> > #endif
> > @@ -1117,17 +1120,20 @@ _mm512_maskz_srli_epi64 (__mmask8 __U,
> __m512i __A, unsigned int __B)
> > }
> > #else
> > #define _mm512_srli_epi64(X, C) \
> > - ((__m512i) __builtin_ia32_psrlqi512_mask ((__v8di)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_psrlqi512_mask ((__v8di)(__m512i)(X), \
> > + (unsigned int)(C), \
> > (__v8di)(__m512i)_mm512_undefined_epi32 (),\
> > (__mmask8)-1))
> >
> > #define _mm512_mask_srli_epi64(W, U, X, C) \
> > - ((__m512i) __builtin_ia32_psrlqi512_mask ((__v8di)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_psrlqi512_mask ((__v8di)(__m512i)(X), \
> > + (unsigned int)(C), \
> > (__v8di)(__m512i)(W),\
> > (__mmask8)(U)))
> >
> > #define _mm512_maskz_srli_epi64(U, X, C) \
> > - ((__m512i) __builtin_ia32_psrlqi512_mask ((__v8di)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_psrlqi512_mask ((__v8di)(__m512i)(X),
> \
> > + (unsigned int)(C), \
> > (__v8di)(__m512i)_mm512_setzero_si512 (),\
> > (__mmask8)(U)))
> > #endif
> > @@ -1196,17 +1202,20 @@ _mm512_maskz_srai_epi64 (__mmask8 __U,
> __m512i __A, unsigned int __B)
> > }
> > #else
> > #define _mm512_srai_epi64(X, C) \
> > - ((__m512i) __builtin_ia32_psraqi512_mask ((__v8di)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_psraqi512_mask ((__v8di)(__m512i)(X), \
> > + (unsigned int)(C), \
> > (__v8di)(__m512i)_mm512_undefined_epi32 (),\
> > (__mmask8)-1))
> >
> > #define _mm512_mask_srai_epi64(W, U, X, C) \
> > - ((__m512i) __builtin_ia32_psraqi512_mask ((__v8di)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_psraqi512_mask ((__v8di)(__m512i)(X), \
> > + (unsigned int)(C), \
> > (__v8di)(__m512i)(W),\
> > (__mmask8)(U)))
> >
> > #define _mm512_maskz_srai_epi64(U, X, C) \
> > - ((__m512i) __builtin_ia32_psraqi512_mask ((__v8di)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_psraqi512_mask ((__v8di)(__m512i)(X), \
> > + (unsigned int)(C), \
> > (__v8di)(__m512i)_mm512_setzero_si512 (),\
> > (__mmask8)(U)))
> > #endif
> > @@ -1275,17 +1284,20 @@ _mm512_maskz_slli_epi32 (__mmask16 __U,
> __m512i __A, unsigned int __B)
> > }
> > #else
> > #define _mm512_slli_epi32(X, C) \
> > - ((__m512i) __builtin_ia32_pslldi512_mask ((__v16si)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_pslldi512_mask ((__v16si)(__m512i)(X), \
> > + (unsigned int)(C), \
> > (__v16si)(__m512i)_mm512_undefined_epi32 (),\
> > (__mmask16)-1))
> >
> > #define _mm512_mask_slli_epi32(W, U, X, C) \
> > - ((__m512i) __builtin_ia32_pslldi512_mask ((__v16si)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_pslldi512_mask ((__v16si)(__m512i)(X), \
> > + (unsigned int)(C), \
> > (__v16si)(__m512i)(W),\
> > (__mmask16)(U)))
> >
> > #define _mm512_maskz_slli_epi32(U, X, C) \
> > - ((__m512i) __builtin_ia32_pslldi512_mask ((__v16si)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_pslldi512_mask ((__v16si)(__m512i)(X), \
> > + (unsigned int)(C), \
> > (__v16si)(__m512i)_mm512_setzero_si512 (),\
> > (__mmask16)(U)))
> > #endif
> > @@ -1354,17 +1366,20 @@ _mm512_maskz_srli_epi32 (__mmask16 __U,
> __m512i __A, unsigned int __B)
> > }
> > #else
> > #define _mm512_srli_epi32(X, C) \
> > - ((__m512i) __builtin_ia32_psrldi512_mask ((__v16si)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_psrldi512_mask ((__v16si)(__m512i)(X),
> \
> > + (unsigned int)(C), \
> > (__v16si)(__m512i)_mm512_undefined_epi32 (),\
> > (__mmask16)-1))
> >
> > #define _mm512_mask_srli_epi32(W, U, X, C) \
> > - ((__m512i) __builtin_ia32_psrldi512_mask ((__v16si)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_psrldi512_mask ((__v16si)(__m512i)(X),
> \
> > + (unsigned int)(C), \
> > (__v16si)(__m512i)(W),\
> > (__mmask16)(U)))
> >
> > #define _mm512_maskz_srli_epi32(U, X, C) \
> > - ((__m512i) __builtin_ia32_psrldi512_mask ((__v16si)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_psrldi512_mask ((__v16si)(__m512i)(X),
> \
> > + (unsigned int)(C), \
> > (__v16si)(__m512i)_mm512_setzero_si512 (),\
> > (__mmask16)(U)))
> > #endif
> > @@ -1433,17 +1448,20 @@ _mm512_maskz_srai_epi32 (__mmask16 __U,
> __m512i __A, unsigned int __B)
> > }
> > #else
> > #define _mm512_srai_epi32(X, C) \
> > - ((__m512i) __builtin_ia32_psradi512_mask ((__v16si)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_psradi512_mask ((__v16si)(__m512i)(X), \
> > + (unsigned int)(C), \
> > (__v16si)(__m512i)_mm512_undefined_epi32 (),\
> > (__mmask16)-1))
> >
> > #define _mm512_mask_srai_epi32(W, U, X, C) \
> > - ((__m512i) __builtin_ia32_psradi512_mask ((__v16si)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_psradi512_mask ((__v16si)(__m512i)(X), \
> > + (unsigned int)(C), \
> > (__v16si)(__m512i)(W),\
> > (__mmask16)(U)))
> >
> > #define _mm512_maskz_srai_epi32(U, X, C) \
> > - ((__m512i) __builtin_ia32_psradi512_mask ((__v16si)(__m512i)(X),
> (int)(C),\
> > + ((__m512i) __builtin_ia32_psradi512_mask ((__v16si)(__m512i)(X), \
> > + (unsigned int)(C), \
> > (__v16si)(__m512i)_mm512_setzero_si512 (),\
> > (__mmask16)(U)))
> > #endif
> > diff --git a/gcc/config/i386/avx512vlbwintrin.h
> b/gcc/config/i386/avx512vlbwintrin.h
> > index 3654cf214c4..b9d248b807b 100644
> > --- a/gcc/config/i386/avx512vlbwintrin.h
> > +++ b/gcc/config/i386/avx512vlbwintrin.h
> > @@ -1759,7 +1759,7 @@ _mm_maskz_shufflelo_epi16 (__mmask8 __U,
> __m128i __A, const int __imm)
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm256_mask_srai_epi16 (__m256i __W, __mmask16 __U, __m256i __A,
> > - const int __imm)
> > + const unsigned int __imm)
> > {
> > return (__m256i) __builtin_ia32_psrawi256_mask ((__v16hi) __A, __imm,
> > (__v16hi) __W,
> > @@ -1768,7 +1768,7 @@ _mm256_mask_srai_epi16 (__m256i __W,
> __mmask16 __U, __m256i __A,
> >
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm256_maskz_srai_epi16 (__mmask16 __U, __m256i __A, const int
> __imm)
> > +_mm256_maskz_srai_epi16 (__mmask16 __U, __m256i __A, const
> unsigned int __imm)
> > {
> > return (__m256i) __builtin_ia32_psrawi256_mask ((__v16hi) __A, __imm,
> > (__v16hi)
> > @@ -1779,7 +1779,7 @@ _mm256_maskz_srai_epi16 (__mmask16 __U,
> __m256i __A, const int __imm)
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm_mask_srai_epi16 (__m128i __W, __mmask8 __U, __m128i __A,
> > - const int __imm)
> > + const unsigned int __imm)
> > {
> > return (__m128i) __builtin_ia32_psrawi128_mask ((__v8hi) __A, __imm,
> > (__v8hi) __W,
> > @@ -1788,7 +1788,7 @@ _mm_mask_srai_epi16 (__m128i __W,
> __mmask8 __U, __m128i __A,
> >
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm_maskz_srai_epi16 (__mmask8 __U, __m128i __A, const int __imm)
> > +_mm_maskz_srai_epi16 (__mmask8 __U, __m128i __A, const unsigned
> int __imm)
> > {
> > return (__m128i) __builtin_ia32_psrawi128_mask ((__v8hi) __A, __imm,
> > (__v8hi)
> > @@ -1799,7 +1799,7 @@ _mm_maskz_srai_epi16 (__mmask8 __U,
> __m128i __A, const int __imm)
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm256_mask_slli_epi16 (__m256i __W, __mmask16 __U, __m256i __A,
> > - int __B)
> > + unsigned int __B)
> > {
> > return (__m256i) __builtin_ia32_psllwi256_mask ((__v16hi) __A, __B,
> > (__v16hi) __W,
> > @@ -1808,7 +1808,7 @@ _mm256_mask_slli_epi16 (__m256i __W,
> __mmask16 __U, __m256i __A,
> >
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm256_maskz_slli_epi16 (__mmask16 __U, __m256i __A, int __B)
> > +_mm256_maskz_slli_epi16 (__mmask16 __U, __m256i __A, unsigned int
> __B)
> > {
> > return (__m256i) __builtin_ia32_psllwi256_mask ((__v16hi) __A, __B,
> > (__v16hi)
> > @@ -1818,7 +1818,7 @@ _mm256_maskz_slli_epi16 (__mmask16 __U,
> __m256i __A, int __B)
> >
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm_mask_slli_epi16 (__m128i __W, __mmask8 __U, __m128i __A, int
> __B)
> > +_mm_mask_slli_epi16 (__m128i __W, __mmask8 __U, __m128i __A,
> unsigned int __B)
> > {
> > return (__m128i) __builtin_ia32_psllwi128_mask ((__v8hi) __A, __B,
> > (__v8hi) __W,
> > @@ -1827,7 +1827,7 @@ _mm_mask_slli_epi16 (__m128i __W, __mmask8
> __U, __m128i __A, int __B)
> >
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm_maskz_slli_epi16 (__mmask8 __U, __m128i __A, int __B)
> > +_mm_maskz_slli_epi16 (__mmask8 __U, __m128i __A, unsigned int __B)
> > {
> > return (__m128i) __builtin_ia32_psllwi128_mask ((__v8hi) __A, __B,
> > (__v8hi)
> > @@ -1859,19 +1859,19 @@ _mm_maskz_slli_epi16 (__mmask8 __U,
> __m128i __A, int __B)
> >
> > #define _mm256_mask_srai_epi16(W, U, A, B) \
> > ((__m256i) __builtin_ia32_psrawi256_mask ((__v16hi)(__m256i)(A), \
> > - (int)(B), (__v16hi)(__m256i)(W), (__mmask16)(U)))
> > + (unsigned int)(B), (__v16hi)(__m256i)(W), (__mmask16)(U)))
> >
> > #define _mm256_maskz_srai_epi16(U, A, B) \
> > ((__m256i) __builtin_ia32_psrawi256_mask ((__v16hi)(__m256i)(A), \
> > - (int)(B), (__v16hi)_mm256_setzero_si256 (), (__mmask16)(U)))
> > + (unsigned int)(B), (__v16hi)_mm256_setzero_si256 (),
> (__mmask16)(U)))
> >
> > #define _mm_mask_srai_epi16(W, U, A, B) \
> > ((__m128i) __builtin_ia32_psrawi128_mask ((__v8hi)(__m128i)(A), \
> > - (int)(B), (__v8hi)(__m128i)(W), (__mmask8)(U)))
> > + (unsigned int)(B), (__v8hi)(__m128i)(W), (__mmask8)(U)))
> >
> > #define _mm_maskz_srai_epi16(U, A, B) \
> > ((__m128i) __builtin_ia32_psrawi128_mask ((__v8hi)(__m128i)(A), \
> > - (int)(B), (__v8hi)_mm_setzero_si128(), (__mmask8)(U)))
> > + (unsigned int)(B), (__v8hi)_mm_setzero_si128(), (__mmask8)(U)))
> >
> > #define _mm256_mask_shufflehi_epi16(W, U, A, B) \
> > ((__m256i) __builtin_ia32_pshufhw256_mask ((__v16hi)(__m256i)(A),
> (int)(B), \
> > @@ -1931,12 +1931,14 @@ _mm_maskz_slli_epi16 (__mmask8 __U,
> __m128i __A, int __B)
> > (__mmask16)(U)))
> >
> > #define _mm_mask_slli_epi16(W, U, X, C) \
> > - ((__m128i)__builtin_ia32_psllwi128_mask ((__v8hi)(__m128i)(X),
> (int)(C),\
> > + ((__m128i)__builtin_ia32_psllwi128_mask ((__v8hi)(__m128i)(X), \
> > + (unsigned int)(C),\
> > (__v8hi)(__m128i)(W),\
> > (__mmask8)(U)))
> >
> > #define _mm_maskz_slli_epi16(U, X, C) \
> > - ((__m128i)__builtin_ia32_psllwi128_mask ((__v8hi)(__m128i)(X),
> (int)(C),\
> > + ((__m128i)__builtin_ia32_psllwi128_mask ((__v8hi)(__m128i)(X), \
> > + (unsigned int)(C),\
> > (__v8hi)(__m128i)_mm_setzero_si128 (),\
> > (__mmask8)(U)))
> >
> > @@ -1947,12 +1949,14 @@ _mm_maskz_slli_epi16 (__mmask8 __U,
> __m128i __A, int __B)
> > (__mmask16)-1))
> >
> > #define _mm256_mask_slli_epi16(W, U, X, C) \
> > - ((__m256i)__builtin_ia32_psllwi256_mask ((__v16hi)(__m256i)(X),
> (int)(C),\
> > + ((__m256i)__builtin_ia32_psllwi256_mask ((__v16hi)(__m256i)(X),
> \
> > + (unsigned int)(C),
> \
> > (__v16hi)(__m256i)(W),\
> > (__mmask16)(U)))
> >
> > #define _mm256_maskz_slli_epi16(U, X, C) \
> > - ((__m256i)__builtin_ia32_psllwi256_mask ((__v16hi)(__m256i)(X),
> (int)(C),\
> > + ((__m256i)__builtin_ia32_psllwi256_mask ((__v16hi)(__m256i)(X),
> \
> > + (unsigned int)(C),
> \
> > (__v16hi)(__m256i)_mm256_setzero_si256 (),\
> > (__mmask16)(U)))
> >
> > diff --git a/gcc/config/i386/avx512vlintrin.h
> b/gcc/config/i386/avx512vlintrin.h
> > index 4a717a7e52f..8b055867986 100644
> > --- a/gcc/config/i386/avx512vlintrin.h
> > +++ b/gcc/config/i386/avx512vlintrin.h
> > @@ -10565,7 +10565,7 @@ _mm_maskz_fixupimm_ps (__mmask8 __U,
> __m128 __A, __m128 __B,
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm256_mask_srli_epi32 (__m256i __W, __mmask8 __U, __m256i __A,
> > - const int __imm)
> > + const unsigned int __imm)
> > {
> > return (__m256i) __builtin_ia32_psrldi256_mask ((__v8si) __A, __imm,
> > (__v8si) __W,
> > @@ -10574,7 +10574,7 @@ _mm256_mask_srli_epi32 (__m256i __W,
> __mmask8 __U, __m256i __A,
> >
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm256_maskz_srli_epi32 (__mmask8 __U, __m256i __A, const int
> __imm)
> > +_mm256_maskz_srli_epi32 (__mmask8 __U, __m256i __A, const
> unsigned int __imm)
> > {
> > return (__m256i) __builtin_ia32_psrldi256_mask ((__v8si) __A, __imm,
> > (__v8si)
> > @@ -10585,7 +10585,7 @@ _mm256_maskz_srli_epi32 (__mmask8 __U,
> __m256i __A, const int __imm)
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm_mask_srli_epi32 (__m128i __W, __mmask8 __U, __m128i __A,
> > - const int __imm)
> > + const unsigned int __imm)
> > {
> > return (__m128i) __builtin_ia32_psrldi128_mask ((__v4si) __A, __imm,
> > (__v4si) __W,
> > @@ -10594,7 +10594,7 @@ _mm_mask_srli_epi32 (__m128i __W,
> __mmask8 __U, __m128i __A,
> >
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm_maskz_srli_epi32 (__mmask8 __U, __m128i __A, const int __imm)
> > +_mm_maskz_srli_epi32 (__mmask8 __U, __m128i __A, const unsigned
> int __imm)
> > {
> > return (__m128i) __builtin_ia32_psrldi128_mask ((__v4si) __A, __imm,
> > (__v4si)
> > @@ -10605,7 +10605,7 @@ _mm_maskz_srli_epi32 (__mmask8 __U,
> __m128i __A, const int __imm)
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm256_mask_srli_epi64 (__m256i __W, __mmask8 __U, __m256i __A,
> > - const int __imm)
> > + const unsigned int __imm)
> > {
> > return (__m256i) __builtin_ia32_psrlqi256_mask ((__v4di) __A, __imm,
> > (__v4di) __W,
> > @@ -10614,7 +10614,7 @@ _mm256_mask_srli_epi64 (__m256i __W,
> __mmask8 __U, __m256i __A,
> >
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm256_maskz_srli_epi64 (__mmask8 __U, __m256i __A, const int
> __imm)
> > +_mm256_maskz_srli_epi64 (__mmask8 __U, __m256i __A, const
> unsigned int __imm)
> > {
> > return (__m256i) __builtin_ia32_psrlqi256_mask ((__v4di) __A, __imm,
> > (__v4di)
> > @@ -10625,7 +10625,7 @@ _mm256_maskz_srli_epi64 (__mmask8 __U,
> __m256i __A, const int __imm)
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm_mask_srli_epi64 (__m128i __W, __mmask8 __U, __m128i __A,
> > - const int __imm)
> > + const unsigned int __imm)
> > {
> > return (__m128i) __builtin_ia32_psrlqi128_mask ((__v2di) __A, __imm,
> > (__v2di) __W,
> > @@ -10634,7 +10634,7 @@ _mm_mask_srli_epi64 (__m128i __W,
> __mmask8 __U, __m128i __A,
> >
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm_maskz_srli_epi64 (__mmask8 __U, __m128i __A, const int __imm)
> > +_mm_maskz_srli_epi64 (__mmask8 __U, __m128i __A, const unsigned
> int __imm)
> > {
> > return (__m128i) __builtin_ia32_psrlqi128_mask ((__v2di) __A, __imm,
> > (__v2di)
> > @@ -12059,7 +12059,7 @@ _mm256_maskz_cvtps_ph (__mmask8 __U,
> __m256 __A, const int __I)
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm256_mask_srai_epi32 (__m256i __W, __mmask8 __U, __m256i __A,
> > - const int __imm)
> > + const unsigned int __imm)
> > {
> > return (__m256i) __builtin_ia32_psradi256_mask ((__v8si) __A, __imm,
> > (__v8si) __W,
> > @@ -12068,7 +12068,7 @@ _mm256_mask_srai_epi32 (__m256i __W,
> __mmask8 __U, __m256i __A,
> >
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm256_maskz_srai_epi32 (__mmask8 __U, __m256i __A, const int
> __imm)
> > +_mm256_maskz_srai_epi32 (__mmask8 __U, __m256i __A, const
> unsigned int __imm)
> > {
> > return (__m256i) __builtin_ia32_psradi256_mask ((__v8si) __A, __imm,
> > (__v8si)
> > @@ -12079,7 +12079,7 @@ _mm256_maskz_srai_epi32 (__mmask8 __U,
> __m256i __A, const int __imm)
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm_mask_srai_epi32 (__m128i __W, __mmask8 __U, __m128i __A,
> > - const int __imm)
> > + const unsigned int __imm)
> > {
> > return (__m128i) __builtin_ia32_psradi128_mask ((__v4si) __A, __imm,
> > (__v4si) __W,
> > @@ -12088,7 +12088,7 @@ _mm_mask_srai_epi32 (__m128i __W,
> __mmask8 __U, __m128i __A,
> >
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm_maskz_srai_epi32 (__mmask8 __U, __m128i __A, const int __imm)
> > +_mm_maskz_srai_epi32 (__mmask8 __U, __m128i __A, const unsigned
> int __imm)
> > {
> > return (__m128i) __builtin_ia32_psradi128_mask ((__v4si) __A, __imm,
> > (__v4si)
> > @@ -12098,7 +12098,7 @@ _mm_maskz_srai_epi32 (__mmask8 __U,
> __m128i __A, const int __imm)
> >
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm256_srai_epi64 (__m256i __A, const int __imm)
> > +_mm256_srai_epi64 (__m256i __A, const unsigned int __imm)
> > {
> > return (__m256i) __builtin_ia32_psraqi256_mask ((__v4di) __A, __imm,
> > (__v4di)
> > @@ -12109,7 +12109,7 @@ _mm256_srai_epi64 (__m256i __A, const int
> __imm)
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm256_mask_srai_epi64 (__m256i __W, __mmask8 __U, __m256i __A,
> > - const int __imm)
> > + const unsigned int __imm)
> > {
> > return (__m256i) __builtin_ia32_psraqi256_mask ((__v4di) __A, __imm,
> > (__v4di) __W,
> > @@ -12118,7 +12118,7 @@ _mm256_mask_srai_epi64 (__m256i __W,
> __mmask8 __U, __m256i __A,
> >
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm256_maskz_srai_epi64 (__mmask8 __U, __m256i __A, const int
> __imm)
> > +_mm256_maskz_srai_epi64 (__mmask8 __U, __m256i __A, const
> unsigned int __imm)
> > {
> > return (__m256i) __builtin_ia32_psraqi256_mask ((__v4di) __A, __imm,
> > (__v4di)
> > @@ -12128,7 +12128,7 @@ _mm256_maskz_srai_epi64 (__mmask8 __U,
> __m256i __A, const int __imm)
> >
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm_srai_epi64 (__m128i __A, const int __imm)
> > +_mm_srai_epi64 (__m128i __A, const unsigned int __imm)
> > {
> > return (__m128i) __builtin_ia32_psraqi128_mask ((__v2di) __A, __imm,
> > (__v2di)
> > @@ -12139,7 +12139,7 @@ _mm_srai_epi64 (__m128i __A, const int
> __imm)
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm_mask_srai_epi64 (__m128i __W, __mmask8 __U, __m128i __A,
> > - const int __imm)
> > + const unsigned int __imm)
> > {
> > return (__m128i) __builtin_ia32_psraqi128_mask ((__v2di) __A, __imm,
> > (__v2di) __W,
> > @@ -12148,7 +12148,7 @@ _mm_mask_srai_epi64 (__m128i __W,
> __mmask8 __U, __m128i __A,
> >
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm_maskz_srai_epi64 (__mmask8 __U, __m128i __A, const int __imm)
> > +_mm_maskz_srai_epi64 (__mmask8 __U, __m128i __A, const unsigned
> int __imm)
> > {
> > return (__m128i) __builtin_ia32_psraqi128_mask ((__v2di) __A, __imm,
> > (__v2di)
> > @@ -12158,7 +12158,7 @@ _mm_maskz_srai_epi64 (__mmask8 __U,
> __m128i __A, const int __imm)
> >
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm_mask_slli_epi32 (__m128i __W, __mmask8 __U, __m128i __A, int
> __B)
> > +_mm_mask_slli_epi32 (__m128i __W, __mmask8 __U, __m128i __A,
> unsigned int __B)
> > {
> > return (__m128i) __builtin_ia32_pslldi128_mask ((__v4si) __A, __B,
> > (__v4si) __W,
> > @@ -12167,7 +12167,7 @@ _mm_mask_slli_epi32 (__m128i __W,
> __mmask8 __U, __m128i __A, int __B)
> >
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm_maskz_slli_epi32 (__mmask8 __U, __m128i __A, int __B)
> > +_mm_maskz_slli_epi32 (__mmask8 __U, __m128i __A, unsigned int __B)
> > {
> > return (__m128i) __builtin_ia32_pslldi128_mask ((__v4si) __A, __B,
> > (__v4si)
> > @@ -12177,7 +12177,7 @@ _mm_maskz_slli_epi32 (__mmask8 __U,
> __m128i __A, int __B)
> >
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm_mask_slli_epi64 (__m128i __W, __mmask8 __U, __m128i __A, int
> __B)
> > +_mm_mask_slli_epi64 (__m128i __W, __mmask8 __U, __m128i __A,
> unsigned int __B)
> > {
> > return (__m128i) __builtin_ia32_psllqi128_mask ((__v2di) __A, __B,
> > (__v2di) __W,
> > @@ -12186,7 +12186,7 @@ _mm_mask_slli_epi64 (__m128i __W,
> __mmask8 __U, __m128i __A, int __B)
> >
> > extern __inline __m128i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm_maskz_slli_epi64 (__mmask8 __U, __m128i __A, int __B)
> > +_mm_maskz_slli_epi64 (__mmask8 __U, __m128i __A, unsigned int __B)
> > {
> > return (__m128i) __builtin_ia32_psllqi128_mask ((__v2di) __A, __B,
> > (__v2di)
> > @@ -12197,7 +12197,7 @@ _mm_maskz_slli_epi64 (__mmask8 __U,
> __m128i __A, int __B)
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm256_mask_slli_epi32 (__m256i __W, __mmask8 __U, __m256i __A,
> > - int __B)
> > + unsigned int __B)
> > {
> > return (__m256i) __builtin_ia32_pslldi256_mask ((__v8si) __A, __B,
> > (__v8si) __W,
> > @@ -12206,7 +12206,7 @@ _mm256_mask_slli_epi32 (__m256i __W,
> __mmask8 __U, __m256i __A,
> >
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm256_maskz_slli_epi32 (__mmask8 __U, __m256i __A, int __B)
> > +_mm256_maskz_slli_epi32 (__mmask8 __U, __m256i __A, unsigned int
> __B)
> > {
> > return (__m256i) __builtin_ia32_pslldi256_mask ((__v8si) __A, __B,
> > (__v8si)
> > @@ -12217,7 +12217,7 @@ _mm256_maskz_slli_epi32 (__mmask8 __U,
> __m256i __A, int __B)
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > _mm256_mask_slli_epi64 (__m256i __W, __mmask8 __U, __m256i __A,
> > - int __B)
> > + unsigned int __B)
> > {
> > return (__m256i) __builtin_ia32_psllqi256_mask ((__v4di) __A, __B,
> > (__v4di) __W,
> > @@ -12226,7 +12226,7 @@ _mm256_mask_slli_epi64 (__m256i __W,
> __mmask8 __U, __m256i __A,
> >
> > extern __inline __m256i
> > __attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
> > -_mm256_maskz_slli_epi64 (__mmask8 __U, __m256i __A, int __B)
> > +_mm256_maskz_slli_epi64 (__mmask8 __U, __m256i __A, unsigned int
> __B)
> > {
> > return (__m256i) __builtin_ia32_psllqi256_mask ((__v4di) __A, __B,
> > (__v4di)
> > @@ -12864,73 +12864,81 @@ _mm256_permutex_pd (__m256d __X,
> const int __M)
> >
> > #define _mm256_mask_srli_epi32(W, U, A, B) \
> > ((__m256i) __builtin_ia32_psrldi256_mask ((__v8si)(__m256i)(A), \
> > - (int)(B), (__v8si)(__m256i)(W), (__mmask8)(U)))
> > + (unsigned int)(B), (__v8si)(__m256i)(W), (__mmask8)(U)))
> >
> > #define _mm256_maskz_srli_epi32(U, A, B) \
> > ((__m256i) __builtin_ia32_psrldi256_mask ((__v8si)(__m256i)(A), \
> > - (int)(B), (__v8si)_mm256_setzero_si256 (), (__mmask8)(U)))
> > + (unsigned int)(B), (__v8si)_mm256_setzero_si256 (), (__mmask8)(U)))
> >
> > #define _mm_mask_srli_epi32(W, U, A, B) \
> > ((__m128i) __builtin_ia32_psrldi128_mask ((__v4si)(__m128i)(A), \
> > - (int)(B), (__v4si)(__m128i)(W), (__mmask8)(U)))
> > + (unsigned int)(B), (__v4si)(__m128i)(W), (__mmask8)(U)))
> >
> > #define _mm_maskz_srli_epi32(U, A, B) \
> > ((__m128i) __builtin_ia32_psrldi128_mask ((__v4si)(__m128i)(A), \
> > - (int)(B), (__v4si)_mm_setzero_si128 (), (__mmask8)(U)))
> > + (unsigned int)(B), (__v4si)_mm_setzero_si128 (), (__mmask8)(U)))
> >
> > #define _mm256_mask_srli_epi64(W, U, A, B) \
> > ((__m256i) __builtin_ia32_psrlqi256_mask ((__v4di)(__m256i)(A), \
> > - (int)(B), (__v4di)(__m256i)(W), (__mmask8)(U)))
> > + (unsigned int)(B), (__v4di)(__m256i)(W), (__mmask8)(U)))
> >
> > #define _mm256_maskz_srli_epi64(U, A, B) \
> > ((__m256i) __builtin_ia32_psrlqi256_mask ((__v4di)(__m256i)(A), \
> > - (int)(B), (__v4di)_mm256_setzero_si256 (), (__mmask8)(U)))
> > + (unsigned int)(B), (__v4di)_mm256_setzero_si256 (), (__mmask8)(U)))
> >
> > #define _mm_mask_srli_epi64(W, U, A, B) \
> > ((__m128i) __builtin_ia32_psrlqi128_mask ((__v2di)(__m128i)(A), \
> > - (int)(B), (__v2di)(__m128i)(W), (__mmask8)(U)))
> > + (unsigned int)(B), (__v2di)(__m128i)(W), (__mmask8)(U)))
> >
> > #define _mm_maskz_srli_epi64(U, A, B) \
> > ((__m128i) __builtin_ia32_psrlqi128_mask ((__v2di)(__m128i)(A), \
> > - (int)(B), (__v2di)_mm_setzero_si128 (), (__mmask8)(U)))
> > + (unsigned int)(B), (__v2di)_mm_setzero_si128 (), (__mmask8)(U)))
> >
> > #define _mm256_mask_slli_epi32(W, U, X, C) \
> > - ((__m256i)__builtin_ia32_pslldi256_mask ((__v8si)(__m256i)(X),
> (int)(C),\
> > + ((__m256i)__builtin_ia32_pslldi256_mask ((__v8si)(__m256i)(X),
> \
> > + (unsigned int)(C),
> \
> > (__v8si)(__m256i)(W), \
> > (__mmask8)(U)))
> >
> > #define _mm256_maskz_slli_epi32(U, X, C) \
> > - ((__m256i)__builtin_ia32_pslldi256_mask ((__v8si)(__m256i)(X),
> (int)(C),\
> > + ((__m256i)__builtin_ia32_pslldi256_mask ((__v8si)(__m256i)(X),
> \
> > + (unsigned int)(C),
> \
> > (__v8si)(__m256i)_mm256_setzero_si256 (), \
> > (__mmask8)(U)))
> >
> > #define _mm256_mask_slli_epi64(W, U, X, C) \
> > - ((__m256i)__builtin_ia32_psllqi256_mask ((__v4di)(__m256i)(X),
> (int)(C),\
> > + ((__m256i)__builtin_ia32_psllqi256_mask ((__v4di)(__m256i)(X),
> \
> > + (unsigned int)(C), \
> > (__v4di)(__m256i)(W), \
> > (__mmask8)(U)))
> >
> > #define _mm256_maskz_slli_epi64(U, X, C) \
> > - ((__m256i)__builtin_ia32_psllqi256_mask ((__v4di)(__m256i)(X),
> (int)(C),\
> > + ((__m256i)__builtin_ia32_psllqi256_mask ((__v4di)(__m256i)(X),
> \
> > + (unsigned int)(C), \
> > (__v4di)(__m256i)_mm256_setzero_si256 (), \
> > (__mmask8)(U)))
> >
> > #define _mm_mask_slli_epi32(W, U, X, C) \
> > - ((__m128i)__builtin_ia32_pslldi128_mask ((__v4si)(__m128i)(X),
> (int)(C),\
> > + ((__m128i)__builtin_ia32_pslldi128_mask ((__v4si)(__m128i)(X), \
> > + (unsigned int)(C), \
> > (__v4si)(__m128i)(W),\
> > (__mmask8)(U)))
> >
> > #define _mm_maskz_slli_epi32(U, X, C) \
> > - ((__m128i)__builtin_ia32_pslldi128_mask ((__v4si)(__m128i)(X),
> (int)(C),\
> > + ((__m128i)__builtin_ia32_pslldi128_mask ((__v4si)(__m128i)(X), \
> > + (unsigned int)(C), \
> > (__v4si)(__m128i)_mm_setzero_si128 (),\
> > (__mmask8)(U)))
> >
> > #define _mm_mask_slli_epi64(W, U, X, C) \
> > - ((__m128i)__builtin_ia32_psllqi128_mask ((__v2di)(__m128i)(X),
> (int)(C),\
> > + ((__m128i)__builtin_ia32_psllqi128_mask ((__v2di)(__m128i)(X),
> \
> > + (unsigned int)(C), \
> > (__v2di)(__m128i)(W),\
> > (__mmask8)(U)))
> >
> > #define _mm_maskz_slli_epi64(U, X, C) \
> > - ((__m128i)__builtin_ia32_psllqi128_mask ((__v2di)(__m128i)(X),
> (int)(C),\
> > + ((__m128i)__builtin_ia32_psllqi128_mask ((__v2di)(__m128i)(X),
> \
> > + (unsigned int)(C), \
> > (__v2di)(__m128i)_mm_setzero_si128 (),\
> > (__mmask8)(U)))
> >
> > @@ -13634,43 +13642,43 @@ _mm256_permutex_pd (__m256d __X,
> const int __M)
> >
> > #define _mm256_mask_srai_epi32(W, U, A, B) \
> > ((__m256i) __builtin_ia32_psradi256_mask ((__v8si)(__m256i)(A), \
> > - (int)(B), (__v8si)(__m256i)(W), (__mmask8)(U)))
> > + (unsigned int)(B), (__v8si)(__m256i)(W), (__mmask8)(U)))
> >
> > #define _mm256_maskz_srai_epi32(U, A, B) \
> > ((__m256i) __builtin_ia32_psradi256_mask ((__v8si)(__m256i)(A), \
> > - (int)(B), (__v8si)_mm256_setzero_si256 (), (__mmask8)(U)))
> > + (unsigned int)(B), (__v8si)_mm256_setzero_si256 (), (__mmask8)(U)))
> >
> > #define _mm_mask_srai_epi32(W, U, A, B) \
> > ((__m128i) __builtin_ia32_psradi128_mask ((__v4si)(__m128i)(A), \
> > - (int)(B), (__v4si)(__m128i)(W), (__mmask8)(U)))
> > + (unsigned int)(B), (__v4si)(__m128i)(W), (__mmask8)(U)))
> >
> > #define _mm_maskz_srai_epi32(U, A, B) \
> > ((__m128i) __builtin_ia32_psradi128_mask ((__v4si)(__m128i)(A), \
> > - (int)(B), (__v4si)_mm_setzero_si128 (), (__mmask8)(U)))
> > + (unsigned int)(B), (__v4si)_mm_setzero_si128 (), (__mmask8)(U)))
> >
> > #define _mm256_srai_epi64(A, B) \
> > ((__m256i) __builtin_ia32_psraqi256_mask ((__v4di)(__m256i)(A), \
> > - (int)(B), (__v4di)_mm256_setzero_si256 (), (__mmask8)-1))
> > + (unsigned int)(B), (__v4di)_mm256_setzero_si256 (), (__mmask8)-1))
> >
> > #define _mm256_mask_srai_epi64(W, U, A, B) \
> > ((__m256i) __builtin_ia32_psraqi256_mask ((__v4di)(__m256i)(A), \
> > - (int)(B), (__v4di)(__m256i)(W), (__mmask8)(U)))
> > + (unsigned int)(B), (__v4di)(__m256i)(W), (__mmask8)(U)))
> >
> > #define _mm256_maskz_srai_epi64(U, A, B) \
> > ((__m256i) __builtin_ia32_psraqi256_mask ((__v4di)(__m256i)(A), \
> > - (int)(B), (__v4di)_mm256_setzero_si256 (), (__mmask8)(U)))
> > + (unsigned int)(B), (__v4di)_mm256_setzero_si256 (), (__mmask8)(U)))
> >
> > #define _mm_srai_epi64(A, B) \
> > ((__m128i) __builtin_ia32_psraqi128_mask ((__v2di)(__m128i)(A), \
> > - (int)(B), (__v2di)_mm_setzero_si128 (), (__mmask8)-1))
> > + (unsigned int)(B), (__v2di)_mm_setzero_si128 (), (__mmask8)-1))
> >
> > #define _mm_mask_srai_epi64(W, U, A, B) \
> > ((__m128i) __builtin_ia32_psraqi128_mask ((__v2di)(__m128i)(A), \
> > - (int)(B), (__v2di)(__m128i)(W), (__mmask8)(U)))
> > + (unsigned int)(B), (__v2di)(__m128i)(W), (__mmask8)(U)))
> >
> > #define _mm_maskz_srai_epi64(U, A, B) \
> > ((__m128i) __builtin_ia32_psraqi128_mask ((__v2di)(__m128i)(A), \
> > - (int)(B), (__v2di)_mm_setzero_si128 (), (__mmask8)(U)))
> > + (unsigned int)(B), (__v2di)_mm_setzero_si128 (), (__mmask8)(U)))
> >
> > #define _mm256_mask_permutex_pd(W, U, A, B) \
> > ((__m256d) __builtin_ia32_permdf256_mask ((__v4df)(__m256d)(A),
> \
> > diff --git a/gcc/testsuite/gcc.target/i386/pr109173-1.c
> b/gcc/testsuite/gcc.target/i386/pr109173-1.c
> > new file mode 100644
> > index 00000000000..286939e56be
> > --- /dev/null
> > +++ b/gcc/testsuite/gcc.target/i386/pr109173-1.c
> > @@ -0,0 +1,57 @@
> > +/* PR target/109173 */
> > +/* { dg-do compile } */
> > +/* { dg-options "-c -Wsign-conversion -Werror -mavx512bw -mavx512vl -
> O2" } */
> > +
> > +#include <immintrin.h>
> > +
> > +extern unsigned int bar();
> > +
> > +void foo()
> > +{
> > + __m128i a1, w1;
> > + __m256i a2, w2;
> > + __mmask8 u;
> > +
> > + _mm256_mask_srli_epi32(w2, u, a2, bar());
> > + _mm256_maskz_srli_epi32(u, a2, bar());
> > + _mm_mask_srli_epi32(w1, u, a1, bar());
> > + _mm_maskz_srli_epi32(u, a1, bar());
> > +
> > + _mm256_mask_srli_epi64(w2, u, a2, bar());
> > + _mm256_maskz_srli_epi64(u, a2, bar());
> > + _mm_mask_srli_epi64(w1, u, a1, bar());
> > + _mm_maskz_srli_epi64(u, a1, bar());
> > +
> > + _mm256_mask_srai_epi32(w2, u, a2, bar());
> > + _mm256_maskz_srai_epi32(u, a2, bar());
> > + _mm_mask_srai_epi32(w1, u, a1, bar());
> > + _mm_maskz_srai_epi32(u, a1, bar());
> > +
> > + _mm256_srai_epi64(a2, bar());
> > + _mm256_mask_srai_epi64(w2, u, a2, bar());
> > + _mm256_maskz_srai_epi64(u, a2, bar());
> > + _mm_srai_epi64(a1, bar());
> > + _mm_mask_srai_epi64(w1, u, a1, bar());
> > + _mm_maskz_srai_epi64(u, a1, bar());
> > +
> > + _mm256_mask_slli_epi32(w2, u, a2, bar());
> > + _mm256_maskz_slli_epi32(u, a2, bar());
> > + _mm_mask_slli_epi32(w1, u, a1, bar());
> > + _mm_maskz_slli_epi32(u, a1, bar());
> > +
> > + _mm256_mask_slli_epi64(w2, u, a2, bar());
> > + _mm256_maskz_slli_epi64(u, a2, bar());
> > + _mm_mask_slli_epi64(w1, u, a1, bar());
> > + _mm_maskz_slli_epi64(u, a1, bar());
> > +
> > + _mm256_mask_srai_epi16(w2, u, a2, bar());
> > + _mm256_maskz_srai_epi16(u, a2, bar());
> > + _mm_mask_srai_epi16(w1, u, a1, bar());
> > + _mm_maskz_srai_epi16(u, a1, bar());
> > +
> > + _mm256_mask_slli_epi16(w2, u, a2, bar());
> > + _mm256_maskz_slli_epi16(u, a2, bar());
> > + _mm_mask_slli_epi16(w1, u, a1, bar());
> > + _mm_maskz_slli_epi16(u, a1, bar());
> > +}
> > +
> > diff --git a/gcc/testsuite/gcc.target/i386/pr109174-1.c
> b/gcc/testsuite/gcc.target/i386/pr109174-1.c
> > new file mode 100644
> > index 00000000000..119721bccf0
> > --- /dev/null
> > +++ b/gcc/testsuite/gcc.target/i386/pr109174-1.c
> > @@ -0,0 +1,45 @@
> > +/* PR target/109174 */
> > +/* { dg-do compile } */
> > +/* { dg-options "-c -Wsign-conversion -Werror -mavx512bw -O2" } */
> > +
> > +#include <immintrin.h>
> > +
> > +extern unsigned int bar();
> > +
> > +void foo()
> > +{
> > + __m512i a, w;
> > + __mmask32 u1;
> > + __mmask16 u2;
> > + __mmask8 u3;
> > +
> > + _mm512_slli_epi64(a, bar());
> > + _mm512_mask_slli_epi64(w, u3, a, bar());
> > + _mm512_maskz_slli_epi64(u3, a, bar());
> > + _mm512_slli_epi32(a, bar());
> > + _mm512_mask_slli_epi32(w, u2, a, bar());
> > + _mm512_maskz_slli_epi32(u2, a, bar());
> > + _mm512_slli_epi16(a, bar());
> > + _mm512_mask_slli_epi16(w, u1, a, bar());
> > + _mm512_maskz_slli_epi16(u1, a, bar());
> > +
> > + _mm512_srai_epi64(a, bar());
> > + _mm512_mask_srai_epi64(w, u3, a, bar());
> > + _mm512_maskz_srai_epi64(u3, a, bar());
> > + _mm512_srai_epi32(a, bar());
> > + _mm512_mask_srai_epi32(w, u2, a, bar());
> > + _mm512_maskz_srai_epi32(u2, a, bar());
> > + _mm512_srai_epi16(a, bar());
> > + _mm512_mask_srai_epi16(w, u1, a, bar());
> > + _mm512_maskz_srai_epi16(u1, a, bar());
> > +
> > + _mm512_srli_epi64(a, bar());
> > + _mm512_mask_srli_epi64(w, u3, a, bar());
> > + _mm512_maskz_srli_epi64(u3, a, bar());
> > + _mm512_srli_epi32(a, bar());
> > + _mm512_mask_srli_epi32(w, u2, a, bar());
> > + _mm512_maskz_srli_epi32(u2, a, bar());
> > + _mm512_srli_epi16(a, bar());
> > + _mm512_mask_srli_epi16(w, u1, a, bar());
> > +}
> > +
> > --
> > 2.31.1
> >
>
>
> --
> BR,
> Hongtao
prev parent reply other threads:[~2023-05-25 8:33 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-25 2:54 Hu, Lin1
2023-05-25 3:39 ` Hongtao Liu
2023-05-25 7:52 ` Hu, Lin1
2023-05-25 8:33 ` Liu, Hongtao [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=SA1PR11MB675758BDFE74E22E82BD17F3E5469@SA1PR11MB6757.namprd11.prod.outlook.com \
--to=hongtao.liu@intel.com \
--cc=crazylht@gmail.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=lin1.hu@intel.com \
--cc=ubizjak@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).