At 15:42 -0400 on 05/05/2015, Farley, Peter x23353 wrote about Re:
LNKEDT 64-bit mode assembler in AMODE 31:
Unless the new version truly needs to move more than 16M bytes at a
time, I would stay with MVCL. The only two benefits to using MVCLE
are support for 4GB lengths (32 bits instead of 24 bits) and the
padding byte specified as part of the instruction instead of loaded
in to the high byte of a register.
I suspect that "CPU-determined" number of bytes probably is not
documented. If IBM still published "Functional Specifications"
manuals for each CPU model it would probably be there, but they
don't publish those any more.
HTH
Peter
Aside from the "CPU-determined" number of bytes halt there is the
issue of page faults with the target (since you just read the source
I will assume it is still all in memory). I think this will be a
bigger factor than the need to test-cc/branch/reissue-the-move.
