In this paper, we revisit the library OS porting approach in a new setting---Intel SGX. We first quantitatively evaluate the performance impact of enclave transitions on SGX programs, identifying it as a performance bottleneck for any library OSes that aim to support system-intensive SGX applications. We then present the design and implementation of SGXKernel, an in-enclave library OS, with highlight on its switchless design, which obviates the needs for enclave transitions.The results show that SGXKernel significantly outperforms a state-of-the-art library OS that has been ported to SGX.
Authors: Hongliang Tian, Tsinghua University, Beijing, China and Intel; Yong Zhang, Tsinghua University, Beijing, China; Chunxiao Xing, Tsinghua University, Beijing, China and Research Institute of Information Technology, Tsinghua National Laboratory for Information Science and Technology, Tsinghua University; Shoumeng Yan, Intel Corporation
From the Computing Frontiers Conference, Siena, Italy, May 15-17, 2017
Imagem do ícone:
Sometimes you may have completely separate instances for an MPI job. You can connect these separate instances using techniques such as MPI_Comm_accept/MPI_Comm_connect, which creates an intercommunicator. But in more complex situations, you may have a significant number of separate intercommunicators, and want to send data between arbitrary ranks, or perform collective operations across all ranks, or other functions which are most efficiently handled over a single intracommunicator.
Nmake and ifort are not recognized from a command window, however using Intel Fortran under Visual Studio works perfectly.
Follow below checklist to troubleshooting Visual Studio command environmental issues:
1. Verify whether ifort and nmake are installed correctly:
For Visual Studio 2017, nmake is installed at:
C:\Program Files (x86)\Microsoft Visual Studio\2017\Professional\VC\Tools\MSVC\14.10.25017\bin\HostX64\x64\nmake.exe