Lines Matching refs:elem_size
20 CeedInt elem_size, num_comp, num_elem; in CeedElemRestrictionGetBackendStrides_Memcheck() local
22 CeedCallBackend(CeedElemRestrictionGetElementSize(rstr, &elem_size)); in CeedElemRestrictionGetBackendStrides_Memcheck()
28 strides[2] = num_comp * elem_size; in CeedElemRestrictionGetBackendStrides_Memcheck()
47 … CeedInt start, CeedInt stop, CeedInt num_elem, CeedInt elem_size, in CeedElemRestrictionApplyStridedNoTranspose_Memcheck_Core() argument
61 for (CeedSize n = 0; n < elem_size; n++) { in CeedElemRestrictionApplyStridedNoTranspose_Memcheck_Core()
63 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyStridedNoTranspose_Memcheck_Core()
74 … CeedInt elem_size, CeedSize v_offset, const CeedScalar *__restrict__ uu, in CeedElemRestrictionApplyOffsetNoTranspose_Memcheck_Core() argument
82 CeedPragmaSIMD for (CeedSize i = 0; i < elem_size * block_size; i++) { in CeedElemRestrictionApplyOffsetNoTranspose_Memcheck_Core()
83 …vv[elem_size * (k * block_size + e * num_comp) + i - v_offset] = uu[impl->offsets[i + e * elem_siz… in CeedElemRestrictionApplyOffsetNoTranspose_Memcheck_Core()
92 … CeedInt stop, CeedInt num_elem, CeedInt elem_size, CeedSize v_offset, in CeedElemRestrictionApplyOrientedNoTranspose_Memcheck_Core() argument
100 CeedPragmaSIMD for (CeedSize i = 0; i < elem_size * block_size; i++) { in CeedElemRestrictionApplyOrientedNoTranspose_Memcheck_Core()
101 vv[elem_size * (k * block_size + e * num_comp) + i - v_offset] = in CeedElemRestrictionApplyOrientedNoTranspose_Memcheck_Core()
102 …uu[impl->offsets[i + e * elem_size] + k * comp_stride] * (impl->orients[i + e * elem_size] ? -1.0 … in CeedElemRestrictionApplyOrientedNoTranspose_Memcheck_Core()
111 … CeedInt stop, CeedInt num_elem, CeedInt elem_size, CeedSize v_offset, in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core() argument
122 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
123 uu[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
124 impl->curl_orients[j + (3 * n + 1) * block_size + e * 3 * elem_size] + in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
125 uu[impl->offsets[j + (n + 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
126 impl->curl_orients[j + (3 * n + 2) * block_size + e * 3 * elem_size]; in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
128 for (n = 1; n < elem_size - 1; n++) { in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
130 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
131 uu[impl->offsets[j + (n - 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
132 impl->curl_orients[j + (3 * n + 0) * block_size + e * 3 * elem_size] + in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
133 uu[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
134 impl->curl_orients[j + (3 * n + 1) * block_size + e * 3 * elem_size] + in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
135 uu[impl->offsets[j + (n + 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
136 impl->curl_orients[j + (3 * n + 2) * block_size + e * 3 * elem_size]; in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
140 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
141 uu[impl->offsets[j + (n - 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
142 impl->curl_orients[j + (3 * n + 0) * block_size + e * 3 * elem_size] + in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
143 uu[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
144 impl->curl_orients[j + (3 * n + 1) * block_size + e * 3 * elem_size]; in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
153 …CeedInt num_elem, CeedInt elem_size, CeedSize v_offset, const CeedScalar *__restrict__ uu, CeedSca… in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core() argument
163 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
164 uu[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
165 abs(impl->curl_orients[j + (3 * n + 1) * block_size + e * 3 * elem_size]) + in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
166 uu[impl->offsets[j + (n + 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
167 abs(impl->curl_orients[j + (3 * n + 2) * block_size + e * 3 * elem_size]); in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
169 for (n = 1; n < elem_size - 1; n++) { in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
171 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
172 uu[impl->offsets[j + (n - 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
173 abs(impl->curl_orients[j + (3 * n + 0) * block_size + e * 3 * elem_size]) + in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
174 uu[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
175 abs(impl->curl_orients[j + (3 * n + 1) * block_size + e * 3 * elem_size]) + in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
176 uu[impl->offsets[j + (n + 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
177 abs(impl->curl_orients[j + (3 * n + 2) * block_size + e * 3 * elem_size]); in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
181 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
182 uu[impl->offsets[j + (n - 1) * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
183 abs(impl->curl_orients[j + (3 * n + 0) * block_size + e * 3 * elem_size]) + in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
184 uu[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] * in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
185 abs(impl->curl_orients[j + (3 * n + 1) * block_size + e * 3 * elem_size]); in CeedElemRestrictionApplyCurlOrientedUnsignedNoTranspose_Memcheck_Core()
193 … CeedInt start, CeedInt stop, CeedInt num_elem, CeedInt elem_size, in CeedElemRestrictionApplyStridedTranspose_Memcheck_Core() argument
207 for (CeedSize n = 0; n < elem_size; n++) { in CeedElemRestrictionApplyStridedTranspose_Memcheck_Core()
210 uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset]; in CeedElemRestrictionApplyStridedTranspose_Memcheck_Core()
220 … CeedInt elem_size, CeedSize v_offset, const CeedScalar *__restrict__ uu, in CeedElemRestrictionApplyOffsetTranspose_Memcheck_Core() argument
228 for (CeedSize i = 0; i < elem_size * block_size; i += block_size) { in CeedElemRestrictionApplyOffsetTranspose_Memcheck_Core()
233 vv_loc = uu[elem_size * (k * block_size + e * num_comp) + j - v_offset]; in CeedElemRestrictionApplyOffsetTranspose_Memcheck_Core()
234 CeedPragmaAtomic vv[impl->offsets[j + e * elem_size] + k * comp_stride] += vv_loc; in CeedElemRestrictionApplyOffsetTranspose_Memcheck_Core()
244 … CeedInt elem_size, CeedSize v_offset, const CeedScalar *__restrict__ uu, in CeedElemRestrictionApplyOrientedTranspose_Memcheck_Core() argument
252 for (CeedSize i = 0; i < elem_size * block_size; i += block_size) { in CeedElemRestrictionApplyOrientedTranspose_Memcheck_Core()
257 …vv_loc = uu[elem_size * (k * block_size + e * num_comp) + j - v_offset] * (impl->orients[j + e * e… in CeedElemRestrictionApplyOrientedTranspose_Memcheck_Core()
258 CeedPragmaAtomic vv[impl->offsets[j + e * elem_size] + k * comp_stride] += vv_loc; in CeedElemRestrictionApplyOrientedTranspose_Memcheck_Core()
268 … CeedInt stop, CeedInt num_elem, CeedInt elem_size, CeedSize v_offset, in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core() argument
282 vv_loc[j] = uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
283 impl->curl_orients[j + (3 * n + 1) * block_size + e * 3 * elem_size] + in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
284 … uu[e * elem_size * num_comp + (k * elem_size + n + 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
285 impl->curl_orients[j + (3 * n + 3) * block_size + e * 3 * elem_size]; in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
288 …CeedPragmaAtomic vv[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] += vv_loc… in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
290 for (n = 1; n < elem_size - 1; n++) { in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
292 … vv_loc[j] = uu[e * elem_size * num_comp + (k * elem_size + n - 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
293 impl->curl_orients[j + (3 * n - 1) * block_size + e * 3 * elem_size] + in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
294 … uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
295 impl->curl_orients[j + (3 * n + 1) * block_size + e * 3 * elem_size] + in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
296 … uu[e * elem_size * num_comp + (k * elem_size + n + 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
297 impl->curl_orients[j + (3 * n + 3) * block_size + e * 3 * elem_size]; in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
300 …CeedPragmaAtomic vv[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] += vv_loc… in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
304 … vv_loc[j] = uu[e * elem_size * num_comp + (k * elem_size + n - 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
305 impl->curl_orients[j + (3 * n - 1) * block_size + e * 3 * elem_size] + in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
306 uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
307 impl->curl_orients[j + (3 * n + 1) * block_size + e * 3 * elem_size]; in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
310 …CeedPragmaAtomic vv[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] += vv_loc… in CeedElemRestrictionApplyCurlOrientedTranspose_Memcheck_Core()
319 …CeedInt num_elem, CeedInt elem_size, CeedSize v_offset, const CeedScalar *__restrict__ uu, CeedSca… in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core() argument
332 vv_loc[j] = uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
333 abs(impl->curl_orients[j + (3 * n + 1) * block_size + e * 3 * elem_size]) + in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
334 … uu[e * elem_size * num_comp + (k * elem_size + n + 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
335 abs(impl->curl_orients[j + (3 * n + 3) * block_size + e * 3 * elem_size]); in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
338 …CeedPragmaAtomic vv[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] += vv_loc… in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
340 for (n = 1; n < elem_size - 1; n++) { in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
342 … vv_loc[j] = uu[e * elem_size * num_comp + (k * elem_size + n - 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
343 … abs(impl->curl_orients[j + (3 * n - 1) * block_size + e * 3 * elem_size]) + in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
344 … uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
345 … abs(impl->curl_orients[j + (3 * n + 1) * block_size + e * 3 * elem_size]) + in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
346 … uu[e * elem_size * num_comp + (k * elem_size + n + 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
347 abs(impl->curl_orients[j + (3 * n + 3) * block_size + e * 3 * elem_size]); in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
350 …CeedPragmaAtomic vv[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] += vv_loc… in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
354 … vv_loc[j] = uu[e * elem_size * num_comp + (k * elem_size + n - 1) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
355 abs(impl->curl_orients[j + (3 * n - 1) * block_size + e * 3 * elem_size]) + in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
356 uu[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] * in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
357 abs(impl->curl_orients[j + (3 * n + 1) * block_size + e * 3 * elem_size]); in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
360 …CeedPragmaAtomic vv[impl->offsets[j + n * block_size + e * elem_size] + k * comp_stride] += vv_loc… in CeedElemRestrictionApplyCurlOrientedUnsignedTranspose_Memcheck_Core()
395 CeedInt num_elem, elem_size; in CeedElemRestrictionApply_Memcheck_Core() local
402 CeedCallBackend(CeedElemRestrictionGetElementSize(rstr, &elem_size)); in CeedElemRestrictionApply_Memcheck_Core()
403 v_offset = start * block_size * elem_size * (CeedSize)num_comp; in CeedElemRestrictionApply_Memcheck_Core()
423 …idedTranspose_Memcheck_Core(rstr, num_comp, block_size, start, stop, num_elem, elem_size, v_offset, in CeedElemRestrictionApply_Memcheck_Core()
428 … elem_size, v_offset, uu, vv)); in CeedElemRestrictionApply_Memcheck_Core()
433 … elem_size, v_offset, uu, vv)); in CeedElemRestrictionApply_Memcheck_Core()
436 … elem_size, v_offset, uu, vv)); in CeedElemRestrictionApply_Memcheck_Core()
442 … elem_size, v_offset, uu, vv)); in CeedElemRestrictionApply_Memcheck_Core()
445 … num_elem, elem_size, v_offset, uu, vv)); in CeedElemRestrictionApply_Memcheck_Core()
448 … elem_size, v_offset, uu, vv)); in CeedElemRestrictionApply_Memcheck_Core()
463 …ApplyStridedNoTranspose_Memcheck_Core(rstr, num_comp, block_size, start, stop, num_elem, elem_size, in CeedElemRestrictionApply_Memcheck_Core()
468 … elem_size, v_offset, uu, vv)); in CeedElemRestrictionApply_Memcheck_Core()
473 … elem_size, v_offset, uu, vv)); in CeedElemRestrictionApply_Memcheck_Core()
476 … elem_size, v_offset, uu, vv)); in CeedElemRestrictionApply_Memcheck_Core()
482 … num_elem, elem_size, v_offset, uu, vv)); in CeedElemRestrictionApply_Memcheck_Core()
485 … num_elem, elem_size, v_offset, uu, vv)); in CeedElemRestrictionApply_Memcheck_Core()
488 … elem_size, v_offset, uu, vv)); in CeedElemRestrictionApply_Memcheck_Core()
643 …CeedInt num_elem, elem_size, num_block, block_size, num_comp, comp_stride, n… in CeedElemRestrictionCreate_Memcheck() local
649 CeedCallBackend(CeedElemRestrictionGetElementSize(rstr, &elem_size)); in CeedElemRestrictionCreate_Memcheck()
664 CeedInt e_layout[3] = {1, elem_size, elem_size * num_comp}, l_layout[3] = {0}; in CeedElemRestrictionCreate_Memcheck()
698 for (CeedInt i = 0; i < num_elem * elem_size; i++) { in CeedElemRestrictionCreate_Memcheck()
706 …sets = rstr_type == CEED_RESTRICTION_POINTS ? (num_elem + 1 + num_points) : (num_elem * elem_size); in CeedElemRestrictionCreate_Memcheck()