Files
rocm-systems/tests/vcopy_MI100
T

50 lines
39 KiB
Plaintext
Raw Normal View History

2024-03-12 12:15:39 -05:00
ELF>p[ @˜’@8 @$"@@ @ hh¨¨ ¨  pKpKpKp[ p[ °° ] } } 8_8� 8� HØ@]@} @} ÐÐRåtd ] } } àPåtdHH H llQåtdÄÄ Ä /lib64/ld-linux-x86-64.so.2GNU ""8"G"_y“ËÐ×"els…’™ ¥¯¹¿Ôï÷ügu„´ÁØî‰‘ œÀl m  
 P5Px€Ž€¤€r'-t)’D‰)
Pqø—^Ó¯k mui ‚P&y œui ‚”‘–®´‘–¹%!%˜ C(EÕLyIk¶__libc_start_main__gmon_start____register_frame_info__cxa_finalize__deregister_frame_info__hipPopCallConfigurationhipLaunchKernel_ZSt4cout_ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKcexitstrlen_ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE9_M_createERmmmemcpy_ZNKSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE7compareEPKcstrtol_ZdlPvhipGetDeviceCounthipSetDeviceprintfmallocputshipMallochipMemcpyceilfhipDeviceSynchronize__hipPushCallConfigurationhipFreefree_ZSt19__throw_logic_errorPKc_ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l__assert_fail_Unwind_Resume_ZNSt8ios_base4InitC1Ev_ZNSt8ios_base4InitD1Ev__cxa_atexit__hipRegisterFatBinary__hipRegisterFunction__hipUnregisterFatBinary__gxx_personality_v0libamdhip64.so.6hip_4.2libstdc++.so.6GLIBCXX_3.4GLIBCXX_3.4.9GLIBCXX_3.4.21CXXABI_1.3libm.so.6GLIBC_2.2.5libgcc_s.so.1GCC_3.0libc.so.6GLIBC_2.14GLIBC_2.34/opt/rocm-6.0.2/lib/llvm/bin/../../../lib    ( 0 ‘ %`� h� p� x� €� ˆ� ��  ˜� 
 �  ¨�  °� 
¸� À� È� Ð� Ø� à� è� ð� ø� � � � �  � (� 0� 8� @� H�  P� &X� !`� "h� #p� $x� 'ÿÿÿÿÿÿÿÿÿÿÿÿÿÿñh㈵øä>@\ à\ Finished allocating vectors on the GPU-b(hipMalloc(&d_c, bytes))==hipSuccessbasic_string::_M_construct null not validFinished copying vectors to the GPU -d/--dev <value> Set the device ID [Default: 0]
-b/--blockSize <value> Set the block size
--device(hipFree(d_a))==hipSuccessOptional:
(hipMemcpy(d_b, h_b, bytes, hipMemcpyHostToDevice))==hipSuccess(hipMemcpy( h_c, d_c, bytes, hipMemcpyDeviceToHost))==hipSuccess(hipSetDevice(devId))==hipSuccess(hipFree(d_c))==hipSuccess--vec-dsw thinks it moved %f KB per wave
Error at position i %d, Expected: %f, Found: %f
-n/--numThreads <value> Set the num of threads
-h_Z7vecCopyPdS_S_iiFinished copying the output vector from the GPU to the CPU--multikernelReleasing GPU memory--helpvcopy testing on GCD %d
Usage: vcopy [OPTIONS]
-h/--help Display this help message
Launching the kernel on the GPUn > 0(hipMalloc(&d_b, bytes))==hipSuccessFinished allocating vectors on the CPURequired:
--blockSizeint main(int, char **)blockSize > 0./sample/vcopy.cpp--iter(hipMalloc(&d_a, bytes))==hipSuccess(hipFree(d_b))==hipSuccess -i/--iter <value> Set the num of iterations [Default: 1]
-nTotal threads: %d, Grid Size: %d block Size:%d, Wavefronts:%d:
(hipMemcpy(d_a, h_a, bytes, hipMemcpyHostToDevice))==hipSuccess(hipGetDeviceCount(&numGpuDevices))==hipSuccessFinished executing kernel_Z9vecCopy_2PdS_S_iiReleasing CPU memory-iÿÿœœ›Š·í__CLANG_OFFLOAD_BUNDLE__host-x86_64-unknown-linux--ðhipv4-amdgcn-amd-amdhsa--gfx908
„
m
AMDGPUƒ®amdhsa.kernels’Þ«.agpr_count¥.argsÜ„®.address_space¦global§.offset¥.size«.value_kind­global_buffer„®.address_space¦global§.offset¥.size«.value_kind­global_buffer„®.address_space¦global§.offset¥.size«.value_kind­global_bufferƒ§.offset¥.size«.value_kind¨by_valueƒ§.offset¥.size«.value_kind¨by_valueƒ§.offset ¥.size«.value_kind´hidden_block_count_xƒ§.offset$¥.size«.value_kind´hidden_block_count_yƒ§.offset(¥.size«.value_kind´hidden_block_count_zƒ§.offset,¥.size«.value_kind³hidden_group_size_xƒ§.offset.¥.size«.value_kind³hidden_group_size_yƒ§.offset0¥.size«.value_kind³hidden_group_size_zƒ§.offset2¥.size«.value_kind²hidden_remainder_xƒ§.offset4¥.size«.value_kind²hidden_remainder_yƒ§.offset6¥.size«.value_kind²hidden_remainder_zƒ§.offsetH¥.size«.value_kind¶hidden_global_offset_xƒ§.offsetP¥.size«.value_kind¶hidden_global_offset_yƒ§.offsetX¥.size«.value_kind¶hidden_global_offset_zƒ§.offset`¥.size«.value_kind°hidden_grid_dims¹.group_segment_fixed_size¶.kernarg_segment_alignµ.kernarg_segment_sizeÍ ©.language¨OpenCL C±.language_version’¸.max_flat_workgroup_sizeÍ¥.name²_Z7vecCopyPdS_S_ii».private_segment_fixed_size«.sgpr_count ±.sgpr_spill_count§.symbolµ_Z7vecCopyPdS_S_ii.kd¸.uniform_work_group_size³.uses_dynamic_stack«.vgpr_count±.vgpr_spill_count¯.wavefront_size@Þ«.agpr_count¥.argsÜ„®.address_space¦global§.offset¥.size«.value_kind­global_buffer„®.address_space¦global§.offset¥.size«.value_kind­global_buffer„®.address_space¦global§.offset¥.size«.value_kind­global_bufferƒ§.offset¥.size«.value_kind¨by_valueƒ§.offset¥.size«.value_kind¨by_valueƒ§.offset ¥.size«.value_kind´hidden_block_count_xƒ§.offset$¥.size«.value_kind´hidden_block_count_yƒ§.offset(¥.size«.value_kind´hidden_block_count_zƒ§.offset,¥.size«.value_kind³hidden_group_size_xƒ§.offset.¥.size«.value_kind³hidden_group_size_yƒ§.offset0¥.size«.value_kind³hidden_group_size_zƒ§.offset2¥.size«.value_kind²hidden_remainder_xƒ§.offset4¥.size«.value_kind²hidden_remainder_yƒ§.offset6¥.size«.value_kind²hidden_remainder_zƒ§.offsetH¥.size«.value_kind¶hidden_global_offset_xƒ§.offsetP¥.size«.value_kind¶hidden_global_offset_yƒ§.offsetX¥.size«.value_kind¶hidden_global_offset_zƒ§.offset`¥.size«.value_kind°hidden_grid_dims¹.group_segment_fixed_size¶.kernarg_segment_alignµ.kernarg_segment_sizeÍ ©.language¨OpenCL C±.language_version’¸.max_flat_workgroup_sizeÍ¥.name´_Z9vecCopy_2PdS_S_ii».private_segment_fixed_size«.sgpr_count ±.sgpr_spill_count§.symbol·_Z9vecCopy_2PdS_S_ii.kd¸.uniform_work_group_size³.uses_dynamic_stack«.vgpr_count±.vgpr_spill_count¯.wavefront_size@­amdhsa.target¹amdgcn-amd-amdhsa--gfx908®amdhsa.version’ €À@*!€?@W!@ƒ!A¯!B

X1s4¤tŠpŒˆh0‡|[äPéã‘Ë—QN_Z7vecCopyPdS_S_ii_Z7vecCopyPdS_S_ii.kd_Z9vecCopy_2PdS_S_ii_Z9vecCopy_2PdS_S_ii.kd_ZN17__HIP_CoordinatesI14__HIP_BlockIdxE1xE_ZN17__HIP_CoordinatesI14__HIP_BlockDimE1xE_ZN17__HIP_CoordinatesI15__HIP_ThreadIdxE1xE @A¯Œ A¯Œ À,BÀÀŒ¿ÿ†ÿÿ’hˆ}j €¾ˆ¿À‚ÀŸ"�ÒƒÀŒ¿~28€TÜ~28pŒ¿€tÜ�¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿€¿À,BÀÀŒ¿ÿ†ÿÿ’hˆ}j €¾ˆ¿À‚ÀŸ"�ÒƒÀŒ¿~28€TÜ~28pŒ¿€tÜ�¿ˆ Ð
ÜõþÿoH
ˆ
Linker: AMD LLD 17.0.0AMD clang version 17.0.0 (https://github.com/RadeonOpenCompute/llvm-project roc-6.0.2 24012 af27734ed982b52a9f1be0f035ac91726fc697e4)Ü€1 €À@*!€?@W!@ƒ!A¯!B.note.dynsym.gnu.hash.hash.dynstr.rodata.text.dynamic.comment.symtab.shstrtab.strtab_Z7vecCopyPdS_S_ii_Z7vecCopyPdS_S_ii.kd_Z9vecCopy_2PdS_S_ii_Z9vecCopy_2PdS_S_ii.kd_ZN17__HIP_CoordinatesI14__HIP_BlockIdxE1xE_ZN17__HIP_CoordinatesI14__HIP_BlockDimE1xE_ZN17__HIP_CoordinatesI15__HIP_ThreadIdxE1xE_DYNAMIC„
 ˆ ˆ ÀöÿÿoH
H
@ˆ
ˆ
HÐ
Ð
Ü'ÀÀƒ@/ €5€1€p>0ðžG�Ø OhaYÉåFPIH ;l hˆ˜ �´áÔ8ôØx<ظ!hH"ˆh"Pˆ" zRx �Ø&D4ðHáDA†C
 hUA†C
P  ˆ<–D€�ˆD�L ¬¸–D€�ˆD�LÐ4ƒAä! AZüH ŒAD0ƒ�A¸ A]4à zPLRxm  �ˆ$(¤   ABB B(B0A8G ƒŒ�Ž�†’.Y.D.a.Q.a.m.T.a.k.Q.a.n.Q.a.n.Q.U.h..f.O.o.K.K.K.K.J.U.U.U.«.j.].f.F.L.\.t.E.\.™.{.Q¨G°I .E.X.™.{.Q¨G°I J.b.J.^.`.O.R.R.M.J.H.H.Q8A0B(B BBAA .J.T.T.T.T.T.T.T.O.J.B.O.O.O.O.O.O.O.O.J.Y.Y.Y.Y.Y.Y.Y.Y.Y.Y.Y.Y.f.E.óú1íI‰Ñ^H‰âHƒäðPTE1À1ÉH�=�ÿ{#ôÌÌÌÌÌÌÌÌÌÌóúÃUH‰åÐ4¶ÀH…Àu.ÆÁ4H‹Z#H…ÀtH�Î4H‰ÆH�œìÿÿH‰Çèüë�]ÃUH‰åì4¶ÀH…Àu?ÆÝ4H‹#H…ÀtH‹*3H‰ÇèÒH‹ #H…ÀtH�GìÿÿH‰ÇèÇë�]ÃÌÌHƒìxH‰|$HH‰t$@H‰T$8‰L$D‰$H�D$HH‰D$PH�D$@H‰D$XH�D$8H‰D$`H�D$H‰D$hH‰àH‰D$pH�|$(H�t$H�T$H�L$èlH‹t$(‹T$0H‹L$D‹D$ L�L$P¿H ÿt$ÿt$èRH�ĈÃf.„HƒìxH‰|$HH‰t$@H‰T$8‰L$D‰$H�D$HH‰D$PH�D$@H‰D$XH�D$8H‰D$`H�D$H‰D$hH‰àH‰D$pH�|$(H�t$H�T$H�L$èÌ
H‹t$(‹T$0H‹L$D‹D$ L�L$P¿P ÿt$ÿt$è²
H�ĈÃf.„P¿‘ ¾u è 
¿‘ ¾7 è‘
¿‘ ¾­ è‚
¿‘ ¾1 ès
¿‘ ¾… èd
¿‘ ¾õ èU
¿‘ ¾Í èF
¿‘ ¾� è7
¿è=
ffff.„UAWAVAUATSH�ìèH‰t$…ÿŽ£ H�l$p‰øH‰D$ H÷ØH‰„$˜¸H‰D$H1ÀH‰„$àA½L�d$`H‰D$H‰„$�1ÀH‰D$@ë!fDH‹„$˜LèHÿÀIÿÅHƒø„†H‹D$N‹tèøH‰l$`M…ö„oL‰÷è• H‰ÃH‰D$(I‰ïHƒør!L‰çH�t$(1Òè… H‰ÅH‰D$`H‹L$(H‰L$pH…Ût,HƒûuAˆMëffffff.„H‰ïL‰öH‰ÚèR H‹D$(H‰D$hH‹L$`ƾB L‰çèB …Àt¾ L‰çè1 …Àu=L;l$ s6H‹D$J‹<è1öº
è! H‰„$�L‰ýH‹|$`H9ï…véóþÿÿ¾N L‰çèã …ÀL‰ýt¾ L‰çèÏ …Àu;L;l$ s4H‹D$J‹<è1öº
è¿ H‰D$H‹|$`H9ï…é—þÿÿ€¾a L‰çèƒ …Àt¾T L‰çèr …Àu>L;l$ s7H‹D$J‹<è1öº
èb H‰D$@H‹|$`H9ï…½é:þÿÿf.„¾† L‰çè# …Àt¾ L‰çè …Àu>L;l$ s7H‹D$J‹<è1öº
è H‰D$HH‹|$`H9ïuaéÞýÿÿfffff.„¾2 L‰çèÃ