On Sat, 14 Aug 2021 20:02:42 +0200, Bernd Oppolzer wrote: >IIRC, DELETE does not really DELETE the module, if the use count (which >is in some of >the control blocks) is not zero. The use count is incremented on every >LOAD, which refers >to the load module and decremented on every DELETE. > Does the storage for all those control blocks and the load modules remain in use, reclaimed exactly when that use count reaches zero?
>If there are more copies of the same module (in the NORENT case), there >must be >multiple control blocks, too (IMO), so that every copy has its own use count >(which is always 1, I guess). > >Maybe I am mixing up the use count with another count, called >responsibility count; >there are two of them, IIRC. > >One of them records the LOAD / DELETE sequences, and the other, AFAIK, >records the LINK / XCTL / ATTACH uses. It is important to know that normal >CALL does not change this count, because the op sys does not know about >CALLs to modules (using machine instructions). > >It would be nice if someone who knows better could correct me on the >details. > And document (RCF) the answer if appropriate. >> On Sat, 14 Aug 2021 08:39:02 -0400, Peter Relson wrote: >>> ... [for NORRENT} every time the system is asked to load the module, even >>> if there >>> is already a usable copy in storage, another copy is fetched. ... >>> >>> And it is possible that there could be procedural dependencies on there >>> being a single copy.. >>> >> If the programmer issues DELETE, which copy is deleted? Thanks, gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
