Home
last modified time | relevance | path

Searched refs:vv (Results 1 – 16 of 16) sorted by relevance

/libCEED/backends/memcheck/
H A Dceed-memcheck-restriction.c49 … CeedScalar *__restrict__ vv) { in CeedElemRestrictionApplyStridedNoTranspose_Memcheck_Core() argument
63 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyStridedNoTranspose_Memcheck_Core()
75 … CeedScalar *__restrict__ vv) { in CeedElemRestrictionApplyOffsetNoTranspose_Memcheck_Core() argument
83 …vv[elem_size * (k * block_size + e * num_comp) + i - v_offset] = uu[impl->offsets[i + e * elem_siz… in CeedElemRestrictionApplyOffsetNoTranspose_Memcheck_Core()
93 … const CeedScalar *__restrict__ uu, CeedScalar *__restrict__ vv) { in CeedElemRestrictionApplyOrientedNoTranspose_Memcheck_Core() argument
101 vv[elem_size * (k * block_size + e * num_comp) + i - v_offset] = in CeedElemRestrictionApplyOrientedNoTranspose_Memcheck_Core()
112 … const CeedScalar *__restrict__ uu, CeedScalar *__restrict__ vv) { 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()
130 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
140 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedNoTranspose_Memcheck_Core()
[all …]
/libCEED/backends/avx/
H A Dceed-avx-tensor.c56 rtype vv[JJ][CC / 4]; // Output tile to be held in registers in CeedTensorContract_Avx_Blocked() local
58 … for (CeedInt cc = 0; cc < CC / 4; cc++) vv[jj][cc] = loadu(&v[(a * J + j + jj) * C + c + cc * 4]); in CeedTensorContract_Avx_Blocked()
64 fmadd(vv[jj][cc], tqv, loadu(&u[(a * B + b) * C + c + cc * 4])); in CeedTensorContract_Avx_Blocked()
69 … for (CeedInt cc = 0; cc < CC / 4; cc++) storeu(&v[(a * J + j + jj) * C + c + cc * 4], vv[jj][cc]); in CeedTensorContract_Avx_Blocked()
78 rtype vv[JJ][CC / 4]; // Output tile to be held in registers in CeedTensorContract_Avx_Blocked() local
81 … for (CeedInt cc = 0; cc < CC / 4; cc++) vv[jj][cc] = loadu(&v[(a * J + j + jj) * C + c + cc * 4]); in CeedTensorContract_Avx_Blocked()
88 fmadd(vv[jj][cc], tqv, loadu(&u[(a * B + b) * C + c + cc * 4])); in CeedTensorContract_Avx_Blocked()
93 … for (CeedInt cc = 0; cc < CC / 4; cc++) storeu(&v[(a * J + j + jj) * C + c + cc * 4], vv[jj][cc]); in CeedTensorContract_Avx_Blocked()
121 rtype vv[JJ]; // Output tile to be held in registers in CeedTensorContract_Avx_Remainder() local
123 for (CeedInt jj = 0; jj < JJ; jj++) vv[jj] = loadu(&v[(a * J + j + jj) * C + c]); in CeedTensorContract_Avx_Remainder()
[all …]
/libCEED/tests/
H A Dt411-qfunction-f.f9012 real*8 vv(q) variable
39 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
41 if (abs(vv(i+voffset)-(i-1)*(i-1)) > 1.0D-14) then
43 write(*,*) 'v(i)=',vv(i+voffset),', u(i)=',(i-1)*(i-1)
47 call ceedvectorrestorearrayread(v,vv,voffset,err)
H A Dt412-qfunction-f.f9013 real*8 vv(q*s) variable
40 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
42 if (abs(vv(i+voffset)-(i-1)*(i-1)) > 1.0D-12) then
44 write(*,*) 'v(i)=',vv(i+voffset),', u(i)=',(i-1)*(i-1)
48 call ceedvectorrestorearrayread(v,vv,voffset,err)
H A Dt410-qfunction-f.f9014 real*8 vv(q) variable
62 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
64 if (abs(vv(i+voffset)-vvv(i)) > 1.0D-14) then
66 write(*,*) 'v(i)=',vv(i+voffset),', vv(i)=',vvv(i)
70 call ceedvectorrestorearrayread(v,vv,voffset,err)
H A Dt400-qfunction-f.f9018 real*8 vv(q)
74 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
76 if (abs(vv(i+voffset)-vvv(i)) > 1.0D-14) then
78 write(*,*) 'v(i)=',vv(i+voffset),', vv(i)=',vvv(i)
82 call ceedvectorrestorearrayread(v,vv,voffset,err)
H A Dt401-qfunction-f.f9019 real*8 vv(q)
86 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
88 if (abs(vv(i+voffset)-ctxdata(5)*vvv(i)) > 1.0D-14) then
90 write(*,*) 'v(i)=',vv(i+voffset),', 5*vv(i)=',ctxdata(5)*vvv(i)
94 call ceedvectorrestorearrayread(v,vv,voffset,err)
H A Dt530-operator-f.f9030 real*8 arrx(d*ndofs),aa(nqpts),qq(nqpts),vv(ndofs)
149 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
152 total=total+vv(voffset+i)
159 call ceedvectorrestorearrayread(v,vv,voffset,err)
172 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
175 total=total+vv(voffset+i)
182 call ceedvectorrestorearrayread(v,vv,voffset,err)
H A Dt532-operator-f.f9031 real*8 arrx(d*ndofs),vv(ndofs)
167 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
170 total=total+vv(voffset+i)
177 call ceedvectorrestorearrayread(v,vv,voffset,err)
213 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
216 total=total+vv(voffset+i)
223 call ceedvectorrestorearrayread(v,vv,voffset,err)
H A Dt531-operator-f.f9031 real*8 arrx(d*ndofs),uu(ndofs),vv(ndofs),vvlin(ndofs)
168 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
171 if (abs(vv(voffset+i)-vvlin(vlinoffset+i))>1.0d-14) then
173 write(*,*) 'Error: Linearized operator computed v[i] = ',vv(voffset+i),&
178 call ceedvectorrestorearrayread(v,vv,voffset,err)
H A Dt533-operator-f.f9030 real*8 arrx(d*ndofs),aa(nqpts),uu(ndofs),vv(ndofs),atrue(ndofs)
141 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
142 atrue(i)=vv(voffset+i)
143 call ceedvectorrestorearrayread(v,vv,voffset,err)
H A Dt534-operator-f.f9030 real*8 arrx(d*ndofs),aa(nqpts),uu(ndofs),vv(ndofs),atrue(ndofs)
145 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
146 atrue(i)=vv(voffset+i)
147 call ceedvectorrestorearrayread(v,vv,voffset,err)
H A Dt535-operator-f.f9030 real*8 arrx(d*ndofs),aa(nqpts),uu(ndofs),vv(ndofs),atrue(ndofs)
176 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
177 atrue(i)=vv(voffset+i)
178 call ceedvectorrestorearrayread(v,vv,voffset,err)
H A Dt536-operator-f.f9034 real*8 arrx(d*ndofs),aa(nqpts),uu(ndofs),vv(ndofs),atrue(ndofs)
198 call ceedvectorgetarrayread(v,ceed_mem_host,vv,voffset,err)
199 atrue(i)=vv(voffset+i)
200 call ceedvectorrestorearrayread(v,vv,voffset,err)
/libCEED/backends/ref/
H A Dceed-ref-restriction.c22 CeedScalar *__restrict__ vv) { in CeedElemRestrictionApplyStridedNoTranspose_Ref_Core() argument
34 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyStridedNoTranspose_Ref_Core()
49 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyStridedNoTranspose_Ref_Core()
62 … const CeedScalar *__restrict__ uu, CeedScalar *__restrict__ vv) { in CeedElemRestrictionApplyOffsetNoTranspose_Ref_Core() argument
70 …vv[elem_size * (k * block_size + e * num_comp) + i - v_offset] = uu[impl->offsets[i + e * elem_siz… in CeedElemRestrictionApplyOffsetNoTranspose_Ref_Core()
80 … const CeedScalar *__restrict__ uu, CeedScalar *__restrict__ vv) { in CeedElemRestrictionApplyOrientedNoTranspose_Ref_Core() argument
88 vv[elem_size * (k * block_size + e * num_comp) + i - v_offset] = in CeedElemRestrictionApplyOrientedNoTranspose_Ref_Core()
99 … const CeedScalar *__restrict__ uu, CeedScalar *__restrict__ vv) { in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core() argument
109 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core()
117 vv[e * elem_size * num_comp + (k * elem_size + n) * block_size + j - v_offset] = in CeedElemRestrictionApplyCurlOrientedNoTranspose_Ref_Core()
[all …]
/libCEED/python/
H A Dtest-wheel.sh10 python -m pytest test-*.py --ceed /cpu/self/opt/blocked -vv