Re: [OMPI devel] Rationale behind memcpy chunk size (in smsc/xpmem)

2022-08-04 Thread Giorgos Katevainis via devel
testing the right things: do warmup > transfers, make sure > you're not swapping, make sure all the processes and memory are pinned > properly, make sure you're > on an otherwise-quiet machine, ... etc.  All the Usual Benchmarking Things. > > -- > Jeff Squyres > jsquy..

Re: [OMPI devel] Rationale behind memcpy chunk size (in smsc/xpmem)

2022-08-03 Thread Jeff Squyres (jsquyres) via devel
tc. All the Usual Benchmarking Things. -- Jeff Squyres jsquy...@cisco.com From: devel on behalf of Giorgos Katevainis via devel Sent: Thursday, July 28, 2022 9:33 AM To: Open MPI Developers Cc: Giorgos Katevainis Subject: [OMPI devel] Rationale behind memcpy chun

[OMPI devel] Rationale behind memcpy chunk size (in smsc/xpmem)

2022-07-28 Thread Giorgos Katevainis via devel
Hello all, I've come across the "memcpy_chunk_size" MCA parameter in smsc/xpmem, which effectively causes memory copies to take place in chunks (used in mca_smsc_xpmem_memmove()). The comment reads: "Maximum size to copy with a single call to memcpy. On some systems a smaller or larger number