Lines Matching full:k
31 for (CeedSize k = 0; k < num_comp; k++) { in CeedElemRestrictionApplyStridedNoTranspose_Ref_Core() local
34 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyStridedNoTranspose_Ref_Core()
35 … uu[n + k * elem_size + CeedIntMin(e + j, num_elem - 1) * elem_size * (CeedSize)num_comp]; in CeedElemRestrictionApplyStridedNoTranspose_Ref_Core()
46 for (CeedSize k = 0; k < num_comp; k++) { in CeedElemRestrictionApplyStridedNoTranspose_Ref_Core() local
49 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyStridedNoTranspose_Ref_Core()
50 … uu[n * strides[0] + k * strides[1] + CeedIntMin(e + j, num_elem - 1) * (CeedSize)strides[2]]; in CeedElemRestrictionApplyStridedNoTranspose_Ref_Core()
68 for (CeedSize k = 0; k < num_comp; k++) { in CeedElemRestrictionApplyOffsetNoTranspose_Ref_Core() local
70 …vv[elem_size * (k * block_size + e * num_comp) + i - v_offset] = uu[impl->offsets[i + e * elem_siz… in CeedElemRestrictionApplyOffsetNoTranspose_Ref_Core()
86 for (CeedSize k = 0; k < num_comp; k++) { in CeedElemRestrictionApplyOrientedNoTranspose_Ref_Core() local
88 vv[elem_size * (k * block_size + e * num_comp) + i - v_offset] = in CeedElemRestrictionApplyOrientedNoTranspose_Ref_Core()
89 …uu[impl->offsets[i + e * elem_size] + k * comp_stride] * (impl->orients[i + e * elem_size] ? -1.0 … in CeedElemRestrictionApplyOrientedNoTranspose_Ref_Core()
105 for (CeedSize k = 0; k < num_comp; k++) { in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core() local
109 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core()
110 uu[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core()
112 uu[impl->offsets[j + (n + 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core()
117 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core()
118 uu[impl->offsets[j + (n - 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core()
120 uu[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core()
122 uu[impl->offsets[j + (n + 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core()
127 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core()
128 uu[impl->offsets[j + (n - 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core()
130 uu[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core()
148 for (CeedSize k = 0; k < num_comp; k++) { in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Ref_Core() local
152 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Ref_Core()
153 uu[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Ref_Core()
155 uu[impl->offsets[j + (n + 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Ref_Core()
160 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Ref_Core()
161 uu[impl->offsets[j + (n - 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Ref_Core()
163 uu[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Ref_Core()
165 uu[impl->offsets[j + (n + 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Ref_Core()
170 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Ref_Core()
171 uu[impl->offsets[j + (n - 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Ref_Core()
173 uu[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Ref_Core()
193 for (CeedSize k = 0; k < num_comp; k++) { in CeedElemRestrictionApplyStridedTranspose_Ref_Core() local
196 …vv[n + k * elem_size + (e + j) * elem_size * num_comp] += uu[e * elem_size * num_comp + (k * elem_… in CeedElemRestrictionApplyStridedTranspose_Ref_Core()
207 for (CeedSize k = 0; k < num_comp; k++) { in CeedElemRestrictionApplyStridedTranspose_Ref_Core() local
210 vv[n * strides[0] + k * strides[1] + (e + j) * strides[2]] += in CeedElemRestrictionApplyStridedTranspose_Ref_Core()
211 uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset]; in CeedElemRestrictionApplyStridedTranspose_Ref_Core()
229 for (CeedSize k = 0; k < num_comp; k++) { in CeedElemRestrictionApplyOffsetTranspose_Ref_Core() local
235 vv_loc = uu[elem_size * (k * block_size + e * num_comp) + j - v_offset]; in CeedElemRestrictionApplyOffsetTranspose_Ref_Core()
236 CeedPragmaAtomic vv[impl->offsets[j + e * elem_size] + k * comp_stride] += vv_loc; in CeedElemRestrictionApplyOffsetTranspose_Ref_Core()
253 for (CeedSize k = 0; k < num_comp; k++) { in CeedElemRestrictionApplyOrientedTranspose_Ref_Core() local
259 …vv_loc = uu[elem_size * (k * block_size + e * num_comp) + j - v_offset] * (impl->orients[j + e * e… in CeedElemRestrictionApplyOrientedTranspose_Ref_Core()
260 CeedPragmaAtomic vv[impl->offsets[j + e * elem_size] + k * comp_stride] += vv_loc; in CeedElemRestrictionApplyOrientedTranspose_Ref_Core()
278 for (CeedSize k = 0; k < num_comp; k++) { in CeedElemRestrictionApplyCurlOrientedTranspose_Ref_Core() local
284 vv_loc[j] = uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Ref_Core()
286 … uu[e * elem_size * num_comp + (k * elem_size + n + 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Ref_Core()
290 …CeedPragmaAtomic vv[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] += vv_loc… in CeedElemRestrictionApplyCurlOrientedTranspose_Ref_Core()
294 … vv_loc[j] = uu[e * elem_size * num_comp + (k * elem_size + n - 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Ref_Core()
296 … uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Ref_Core()
298 … uu[e * elem_size * num_comp + (k * elem_size + n + 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Ref_Core()
302 …CeedPragmaAtomic vv[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] += vv_loc… in CeedElemRestrictionApplyCurlOrientedTranspose_Ref_Core()
306 … vv_loc[j] = uu[e * elem_size * num_comp + (k * elem_size + n - 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Ref_Core()
308 uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Ref_Core()
312 …CeedPragmaAtomic vv[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] += vv_loc… in CeedElemRestrictionApplyCurlOrientedTranspose_Ref_Core()
330 for (CeedSize k = 0; k < num_comp; k++) { in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Ref_Core() local
336 vv_loc[j] = uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Ref_Core()
338 … uu[e * elem_size * num_comp + (k * elem_size + n + 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Ref_Core()
342 …CeedPragmaAtomic vv[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] += vv_loc… in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Ref_Core()
346 … vv_loc[j] = uu[e * elem_size * num_comp + (k * elem_size + n - 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Ref_Core()
348 … uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Ref_Core()
350 … uu[e * elem_size * num_comp + (k * elem_size + n + 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Ref_Core()
354 …CeedPragmaAtomic vv[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] += vv_loc… in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Ref_Core()
358 … vv_loc[j] = uu[e * elem_size * num_comp + (k * elem_size + n - 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Ref_Core()
360 uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Ref_Core()
364 …CeedPragmaAtomic vv[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] += vv_loc… in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Ref_Core()