自学教程：C++ stride函数代码示例

voidvec4_generator::generate_gs_set_write_offset(struct brw_reg dst,        struct brw_reg src0,        struct brw_reg src1){    /* From p22 of volume 4 part 2 of the Ivy Bridge PRM (2.4.3.1 Message     * Header: M0.3):     *     *     Slot 0 Offset. This field, after adding to the Global Offset field     *     in the message descriptor, specifies the offset (in 256-bit units)     *     from the start of the URB entry, as referenced by URB Handle 0, at     *     which the data will be accessed.     *     * Similar text describes DWORD M0.4, which is slot 1 offset.     *     * Therefore, we want to multiply DWORDs 0 and 4 of src0 (the x components     * of the register for geometry shader invocations 0 and 1) by the     * immediate value in src1, and store the result in DWORDs 3 and 4 of dst.     *     * We can do this with the following EU instruction:     *     *     mul(2) dst.3<1>UD src0<8;2,4>UD src1   { Align1 WE_all }     */    brw_push_insn_state(p);    brw_set_access_mode(p, BRW_ALIGN_1);    brw_set_mask_control(p, BRW_MASK_DISABLE);    brw_MUL(p, suboffset(stride(dst, 2, 2, 1), 3), stride(src0, 8, 2, 4),            src1);    brw_set_access_mode(p, BRW_ALIGN_16);    brw_pop_insn_state(p);}

static void emit_pixel_xy(struct brw_compile *p,			  const struct brw_reg *dst,			  GLuint mask,			  const struct brw_reg *arg0){   struct brw_reg r1 = brw_vec1_grf(1, 0);   struct brw_reg r1_uw = retype(r1, BRW_REGISTER_TYPE_UW);   brw_set_compression_control(p, BRW_COMPRESSION_NONE);   /* Calculate pixel centers by adding 1 or 0 to each of the    * micro-tile coordinates passed in r1.    */   if (mask & WRITEMASK_X) {      brw_ADD(p,	      vec16(retype(dst[0], BRW_REGISTER_TYPE_UW)),	      stride(suboffset(r1_uw, 4), 2, 4, 0),	      brw_imm_v(0x10101010));   }   if (mask & WRITEMASK_Y) {      brw_ADD(p,	      vec16(retype(dst[1], BRW_REGISTER_TYPE_UW)),	      stride(suboffset(r1_uw,5), 2, 4, 0),	      brw_imm_v(0x11001100));   }   brw_set_compression_control(p, BRW_COMPRESSION_COMPRESSED);}

void stencilBlitzStencilVersion(BenchmarkExt<int>& bench){    bench.beginImplementation("Blitz++ Stencil");    while (!bench.doneImplementationBenchmark())    {        int N = bench.getParameter();        cout << "Blitz++ Stencil: N = " << N << endl;        cout.flush();        long iters = bench.getIterations();        Array<double,3> A(N,N,N), B(N,N,N);        initializeRandomDouble(A.data(), N*N*N, A.stride(thirdDim));        initializeRandomDouble(B.data(), N*N*N, B.stride(thirdDim));        TinyVector<int,2> size = N-2;        generateFastTraversalOrder(size);        double c = 1/7.;       	;        bench.start();        for (long i=0; i < iters; ++i)        {            Range I(1,N-2), J(1,N-2), K(1,N-2);	    applyStencil(test1stencil(),A,B);	    applyStencil(test1stencil(),B,A);        }        bench.stop();    }    bench.endImplementation();}

static void emit_pixel_xy(struct brw_wm_compile *c,		struct prog_instruction *inst){    struct brw_reg r1 = brw_vec1_grf(1, 0);    struct brw_reg r1_uw = retype(r1, BRW_REGISTER_TYPE_UW);    struct brw_reg dst0, dst1;    struct brw_compile *p = &c->func;    GLuint mask = inst->DstReg.WriteMask;    dst0 = get_dst_reg(c, inst, 0, 1);    dst1 = get_dst_reg(c, inst, 1, 1);    /* Calculate pixel centers by adding 1 or 0 to each of the     * micro-tile coordinates passed in r1.     */    if (mask & WRITEMASK_X) {	brw_ADD(p,		vec8(retype(dst0, BRW_REGISTER_TYPE_UW)),		stride(suboffset(r1_uw, 4), 2, 4, 0),		brw_imm_v(0x10101010));    }    if (mask & WRITEMASK_Y) {	brw_ADD(p,		vec8(retype(dst1, BRW_REGISTER_TYPE_UW)),		stride(suboffset(r1_uw, 5), 2, 4, 0),		brw_imm_v(0x11001100));    }}

static unsigned longsize_yuv420(int w, int h){    unsigned yPitch = stride(w);    return h * (yPitch + (yPitch >> 1));}

void Data::print() const{	printf("%p + %d, (%d", mData, offset(), size(0));	for(int i=1;i<highestDim(*this);++i){		printf(",%d", size(i));	}	printf("):%+d, %s %s/n", stride(), typeToString(type()).c_str(), toToken().c_str());}

void const* ms::GLPixelBuffer::as_argb_8888(){    if (pixels_need_y_flip)    {        auto const stride_val = stride().as_uint32_t();        auto const height = size_.height.as_uint32_t();        std::vector<char> tmp(stride_val);        for (unsigned int i = 0; i < height / 2; i++)        {            /* Store line i */            tmp.assign(&pixels[i * stride_val], &pixels[(i + 1) * stride_val]);            /* Copy line height - i - 1 to line i */            copy_and_convert_pixel_line(&pixels[(height - i - 1) * stride_val],                                        &pixels[i * stride_val]);            /* Copy stored line (i) to height - i - 1 */            copy_and_convert_pixel_line(tmp.data(),                                        &pixels[(height - i - 1) * stride_val]);        }        /* Process middle line if there is one */        if (height % 2 == 1)        {            copy_and_convert_pixel_line(&pixels[(height / 2) * stride_val],                                        &pixels[(height / 2) * stride_val]);        }        pixels_need_y_flip = false;    }    return pixels.data();}

	Image& Image::flip()	{		// 1. パラメ
C++ strim函数代码示例
C++ strict_strtoul函数代码示例