On 2016-11-27, at 11:09, Steve Thompson wrote: > Yes, that did not exist when I was doing all that. And I have not looked at > the exit doc for HLASM for several years now. So I can't speak to the Unix > I/O control block interface. > It must have major/minor device numbers plus I-number which doesn't all fit in 32 bits returned by BLDL, ao it must be an ephemeral token.
There was one major APAR just before I started using the exits. Something like providing notification of MEXIT/MEND so the next call to the input exit could switch back to the parent file. And after BPAM support happened, there were two APARs for me concerning mixed SYSLIB concatenations. One fixed a program check and the other fixed POINT's misbehaving on nested COPY instructions. We've since switched to a cross assembler so I don't know whether they've stayed fixed. I have a mixed SYSEXEC concatenation. Unsupported but useful I get occasional program checks, but never where they cause data loss. In fact, I was able to reproduce one of my SYSEXEC program checks using HLASM SYSLIB and thus get support. -- gil
