Re: [Qemu-devel] [PATCH v3 2/5] target-ppc: add vector extract instructions

[David Gibson]

On Thu, Aug 11, 2016 at 01:06:45PM +0530, Rajalakshmi Srinivasaraghavan wrote:
> The following vector extract instructions are added from ISA 3.0.
> 
> vextractub - Vector Extract Unsigned Byte
> vextractuh - Vector Extract Unsigned Halfword
> vextractuw - Vector Extract Unsigned Word
> vextractd - Vector Extract Unsigned Doubleword
> 
> Signed-off-by: Rajalakshmi Srinivasaraghavan <address@hidden>
> ---
>  target-ppc/helper.h             |    4 ++++
>  target-ppc/int_helper.c         |   26 ++++++++++++++++++++++++++
>  target-ppc/translate/vmx-impl.c |   10 ++++++++++
>  target-ppc/translate/vmx-ops.c  |   10 +++++++---
>  4 files changed, 47 insertions(+), 3 deletions(-)
> 
> diff --git a/target-ppc/helper.h b/target-ppc/helper.h
> index 0923779..59e7b88 100644
> --- a/target-ppc/helper.h
> +++ b/target-ppc/helper.h
> @@ -250,6 +250,10 @@ DEF_HELPER_2(vspltisw, void, avr, i32)
>  DEF_HELPER_3(vspltb, void, avr, avr, i32)
>  DEF_HELPER_3(vsplth, void, avr, avr, i32)
>  DEF_HELPER_3(vspltw, void, avr, avr, i32)
> +DEF_HELPER_3(vextractub, void, avr, avr, i32)
> +DEF_HELPER_3(vextractuh, void, avr, avr, i32)
> +DEF_HELPER_3(vextractuw, void, avr, avr, i32)
> +DEF_HELPER_3(vextractd, void, avr, avr, i32)
>  DEF_HELPER_3(vinsertb, void, avr, avr, i32)
>  DEF_HELPER_3(vinserth, void, avr, avr, i32)
>  DEF_HELPER_3(vinsertw, void, avr, avr, i32)
> diff --git a/target-ppc/int_helper.c b/target-ppc/int_helper.c
> index 3f8e439..a917bd5 100644
> --- a/target-ppc/int_helper.c
> +++ b/target-ppc/int_helper.c
> @@ -1819,6 +1819,32 @@ VINSERT(h, u16, 3)
>  VINSERT(w, u32, 1)
>  VINSERT(d, u64, 0)
>  #undef VINSERT
> +#if defined(HOST_WORDS_BIGENDIAN)
> +#define VEXTRACT(suffix, element, index)                                     
> \
> +    void helper_vextract##suffix(ppc_avr_t *r, ppc_avr_t *b, uint32_t splat) 
> \

splat is not a good parameter name here, since the element is being
extracted, rather than splatted.

> +    {                                                                        
> \
> +        uint32_t s = sizeof(r->element[0]) * index;                          
> \
> +        ppc_avr_t result = { .u64 = { 0, 0 } };                              
> \
> +        memcpy(&result.element[index], &b->u8[splat],                        
> \
> +               sizeof(result.element[0]));                                   
> \
> +        *r = result;                                                         
> \
> +    }
> +#else
> +#define VEXTRACT(suffix, element, index)                                     
> \
> +    void helper_vextract##suffix(ppc_avr_t *r, ppc_avr_t *b, uint32_t splat) 
> \
> +    {                                                                        
> \
> +        ppc_avr_t result = { .u64 = { 0, 0 } };                              
> \
> +        uint32_t s = (16 - splat) - sizeof(r->element[0]);                   
> \
> +        uint32_t d = (ARRAY_SIZE(r->element) - index) - 1;
> \

Same comments on the index value as for vinsert*.

> +        memcpy(&result.element[d], &b->u8[s], sizeof(result.element[0]));    
> \
> +        *r = result;                                                         
> \
> +    }
> +#endif
> +VEXTRACT(ub, u8, 7)
> +VEXTRACT(uh, u16, 3)
> +VEXTRACT(uw, u32, 1)
> +VEXTRACT(d, u64, 0)
> +#undef VEXTRACT
>  
>  #define VSPLTI(suffix, element, splat_type)                     \
>      void helper_vspltis##suffix(ppc_avr_t *r, uint32_t splat)   \
> diff --git a/target-ppc/translate/vmx-impl.c b/target-ppc/translate/vmx-impl.c
> index f6a97ac..766a645 100644
> --- a/target-ppc/translate/vmx-impl.c
> +++ b/target-ppc/translate/vmx-impl.c
> @@ -648,6 +648,10 @@ static void glue(gen_, name)(DisasContext *ctx)          
>                \
>  GEN_VXFORM_UIMM(vspltb, 6, 8);
>  GEN_VXFORM_UIMM(vsplth, 6, 9);
>  GEN_VXFORM_UIMM(vspltw, 6, 10);
> +GEN_VXFORM_UIMM_SPLAT(vextractub, 6, 8, 15);
> +GEN_VXFORM_UIMM_SPLAT(vextractuh, 6, 9, 14);
> +GEN_VXFORM_UIMM_SPLAT(vextractuw, 6, 10, 12);
> +GEN_VXFORM_UIMM_SPLAT(vextractd, 6, 11, 8);
>  GEN_VXFORM_UIMM_SPLAT(vinsertb, 6, 12, 15);
>  GEN_VXFORM_UIMM_SPLAT(vinserth, 6, 13, 14);
>  GEN_VXFORM_UIMM_SPLAT(vinsertw, 6, 14, 12);
> @@ -656,6 +660,12 @@ GEN_VXFORM_UIMM_ENV(vcfux, 5, 12);
>  GEN_VXFORM_UIMM_ENV(vcfsx, 5, 13);
>  GEN_VXFORM_UIMM_ENV(vctuxs, 5, 14);
>  GEN_VXFORM_UIMM_ENV(vctsxs, 5, 15);
> +GEN_VXFORM_DUAL(vspltb, PPC_NONE, PPC2_ALTIVEC_207,
> +                      vextractub, PPC_NONE, PPC2_ISA300);
> +GEN_VXFORM_DUAL(vsplth, PPC_NONE, PPC2_ALTIVEC_207,
> +                      vextractuh, PPC_NONE, PPC2_ISA300);
> +GEN_VXFORM_DUAL(vspltw, PPC_NONE, PPC2_ALTIVEC_207,
> +                      vextractuw, PPC_NONE, PPC2_ISA300);
>  GEN_VXFORM_DUAL(vspltisb, PPC_NONE, PPC2_ALTIVEC_207,
>                        vinsertb, PPC_NONE, PPC2_ISA300);
>  GEN_VXFORM_DUAL(vspltish, PPC_NONE, PPC2_ALTIVEC_207,
> diff --git a/target-ppc/translate/vmx-ops.c b/target-ppc/translate/vmx-ops.c
> index ca69e56..aafe70b 100644
> --- a/target-ppc/translate/vmx-ops.c
> +++ b/target-ppc/translate/vmx-ops.c
> @@ -197,6 +197,13 @@ GEN_VXRFORM_DUAL(vcmpbfp, vcmpgtsd, 3, 15, PPC_ALTIVEC, 
> PPC_NONE)
>  #define GEN_VXFORM_DUAL_INV(name0, name1, opc2, opc3, inval0, inval1, type) \
>  GEN_OPCODE_DUAL(name0##_##name1, 0x04, opc2, opc3, inval0, inval1, type, \
>                                                                 PPC_NONE)
> +GEN_VXFORM_DUAL_INV(vspltb, vextractub, 6, 8, 0x00000000, 0x100000,
> +                                               PPC2_ALTIVEC_207),
> +GEN_VXFORM_DUAL_INV(vsplth, vextractuh, 6, 9, 0x00000000, 0x100000,
> +                                               PPC2_ALTIVEC_207),
> +GEN_VXFORM_DUAL_INV(vspltw, vextractuw, 6, 10, 0x00000000, 0x100000,
> +                                               PPC2_ALTIVEC_207),
> +GEN_VXFORM_300_EXT(vextractd, 6, 11, 0x100000),
>  GEN_VXFORM_DUAL_INV(vspltisb, vinsertb, 6, 12, 0x00000000, 0x100000,
>                                                 PPC2_ALTIVEC_207),
>  GEN_VXFORM_DUAL_INV(vspltish, vinserth, 6, 13, 0x00000000, 0x100000,
> @@ -226,9 +233,6 @@ GEN_VXFORM_NOA(vrfiz, 5, 9),
>  
>  #define GEN_VXFORM_UIMM(name, opc2, opc3)                               \
>      GEN_HANDLER(name, 0x04, opc2, opc3, 0x00000000, PPC_ALTIVEC)
> -GEN_VXFORM_UIMM(vspltb, 6, 8),
> -GEN_VXFORM_UIMM(vsplth, 6, 9),
> -GEN_VXFORM_UIMM(vspltw, 6, 10),
>  GEN_VXFORM_UIMM(vcfux, 5, 12),
>  GEN_VXFORM_UIMM(vcfsx, 5, 13),
>  GEN_VXFORM_UIMM(vctuxs, 5, 14),

-- 
David Gibson                    | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au  | minimalist, thank you.  NOT _the_ _other_
                                | _way_ _around_!
http://www.ozlabs.org/~dgibson

signature.asc
Description: PGP signature