comparison i386/dsputil_h264_template_ssse3.c @ 8031:eebc7209c47f libavcodec

Convert asm keyword into __asm__. Neither the asm() nor the __asm__() keyword is part of the C99 standard, but while GCC accepts the former in C89 syntax, it is not accepted in C99 unless GNU extensions are turned on (with -fasm). The latter form is accepted in any syntax as an extension (without requiring further command-line options). Sun Studio C99 compiler also does not accept asm() while accepting __asm__(), albeit reporting warnings that it's not valid C99 syntax.

8030:a512ac8fa540

    assert(x<8 && y<8 && x>=0 && y>=0);

    if(y==0 || x==0)
    {
        /* 1 dimensional filter only */
        asm volatile(
            "movd %0, %%xmm7 \n\t"
            "movq %1, %%xmm6 \n\t"
            "pshuflw $0, %%xmm7, %%xmm7 \n\t"
            "movlhps %%xmm6, %%xmm6 \n\t"
            "movlhps %%xmm7, %%xmm7 \n\t"
            :: "r"(255*(x+y)+8), "m"(*(rnd?&ff_pw_4:&ff_pw_3))
        );

        if(x) {
            asm volatile(
                "1: \n\t"
                "movq (%1), %%xmm0 \n\t"
                "movq 1(%1), %%xmm1 \n\t"
                "movq (%1,%3), %%xmm2 \n\t"
                "movq 1(%1,%3), %%xmm3 \n\t"

                "jg 1b \n\t"
                :"+r"(dst), "+r"(src), "+r"(h)
                :"r"((x86_reg)stride)
            );
        } else {
            asm volatile(
                "1: \n\t"
                "movq (%1), %%xmm0 \n\t"
                "movq (%1,%3), %%xmm1 \n\t"
                "movdqa %%xmm1, %%xmm2 \n\t"
                "movq (%1,%3,2), %%xmm3 \n\t"

        }
        return;
    }

    /* general case, bilinear */
    asm volatile(
        "movd %0, %%xmm7 \n\t"
        "movd %1, %%xmm6 \n\t"
        "movdqa %2, %%xmm5 \n\t"
        "pshuflw $0, %%xmm7, %%xmm7 \n\t"
        "pshuflw $0, %%xmm6, %%xmm6 \n\t"
        "movlhps %%xmm7, %%xmm7 \n\t"
        "movlhps %%xmm6, %%xmm6 \n\t"
        :: "r"((x*255+8)*(8-y)), "r"((x*255+8)*y), "m"(*(rnd?&ff_pw_32:&ff_pw_28))
    );

    asm volatile(
        "movq (%1), %%xmm0 \n\t"
        "movq 1(%1), %%xmm1 \n\t"
        "punpcklbw %%xmm1, %%xmm0 \n\t"
        "add %3, %1 \n\t"
        "1: \n\t"

    );
}

static void H264_CHROMA_MC4_TMPL(uint8_t *dst/*align 4*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y)
{
    asm volatile(
        "movd %0, %%mm7 \n\t"
        "movd %1, %%mm6 \n\t"
        "movq %2, %%mm5 \n\t"
        "pshufw $0, %%mm7, %%mm7 \n\t"
        "pshufw $0, %%mm6, %%mm6 \n\t"
        :: "r"((x*255+8)*(8-y)), "r"((x*255+8)*y), "m"(ff_pw_32)
    );

    asm volatile(
        "movd (%1), %%mm0 \n\t"
        "punpcklbw 1(%1), %%mm0 \n\t"
        "add %3, %1 \n\t"
        "1: \n\t"
        "movd (%1), %%mm1 \n\t"

8031:eebc7209c47f

    assert(x<8 && y<8 && x>=0 && y>=0);

    if(y==0 || x==0)
    {
        /* 1 dimensional filter only */
        __asm__ volatile(
            "movd %0, %%xmm7 \n\t"
            "movq %1, %%xmm6 \n\t"
            "pshuflw $0, %%xmm7, %%xmm7 \n\t"
            "movlhps %%xmm6, %%xmm6 \n\t"
            "movlhps %%xmm7, %%xmm7 \n\t"
            :: "r"(255*(x+y)+8), "m"(*(rnd?&ff_pw_4:&ff_pw_3))
        );

        if(x) {
            __asm__ volatile(
                "1: \n\t"
                "movq (%1), %%xmm0 \n\t"
                "movq 1(%1), %%xmm1 \n\t"
                "movq (%1,%3), %%xmm2 \n\t"
                "movq 1(%1,%3), %%xmm3 \n\t"

                "jg 1b \n\t"
                :"+r"(dst), "+r"(src), "+r"(h)
                :"r"((x86_reg)stride)
            );
        } else {
            __asm__ volatile(
                "1: \n\t"
                "movq (%1), %%xmm0 \n\t"
                "movq (%1,%3), %%xmm1 \n\t"
                "movdqa %%xmm1, %%xmm2 \n\t"
                "movq (%1,%3,2), %%xmm3 \n\t"

        }
        return;
    }

    /* general case, bilinear */
    __asm__ volatile(
        "movd %0, %%xmm7 \n\t"
        "movd %1, %%xmm6 \n\t"
        "movdqa %2, %%xmm5 \n\t"
        "pshuflw $0, %%xmm7, %%xmm7 \n\t"
        "pshuflw $0, %%xmm6, %%xmm6 \n\t"
        "movlhps %%xmm7, %%xmm7 \n\t"
        "movlhps %%xmm6, %%xmm6 \n\t"
        :: "r"((x*255+8)*(8-y)), "r"((x*255+8)*y), "m"(*(rnd?&ff_pw_32:&ff_pw_28))
    );

    __asm__ volatile(
        "movq (%1), %%xmm0 \n\t"
        "movq 1(%1), %%xmm1 \n\t"
        "punpcklbw %%xmm1, %%xmm0 \n\t"
        "add %3, %1 \n\t"
        "1: \n\t"

    );
}

static void H264_CHROMA_MC4_TMPL(uint8_t *dst/*align 4*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y)
{
    __asm__ volatile(
        "movd %0, %%mm7 \n\t"
        "movd %1, %%mm6 \n\t"
        "movq %2, %%mm5 \n\t"
        "pshufw $0, %%mm7, %%mm7 \n\t"
        "pshufw $0, %%mm6, %%mm6 \n\t"
        :: "r"((x*255+8)*(8-y)), "r"((x*255+8)*y), "m"(ff_pw_32)
    );

    __asm__ volatile(
        "movd (%1), %%mm0 \n\t"
        "punpcklbw 1(%1), %%mm0 \n\t"
        "add %3, %1 \n\t"
        "1: \n\t"
        "movd (%1), %%mm1 \n\t"