vcon in a dsect

2020-09-10 Thread pgs4ibmm...@pacbell.net
What is the best way to represent a VCON in a DSECT? A DS V requires some value: TESTDSECT TESTVCON DS 1VL4(DUMMY) or TESTVCON DS 1AL4(*-*) [TESTVCON DS 1VL4(*-*) generates assembler error message] Either way you get the 4-byte placeholder, but is there a preferre

USING support for HIGH-WORD facility

2020-07-20 Thread pgs4ibmm...@pacbell.net
It seems the HIGH-WORD facility (from the year 2010) provides a limited amount of instructions that act only on the high 32 bits. The Principles of Operation terminology “effectively provides sixteen additional 32-bit registers”. HLASM doesn’t seem to provide a way to specify the a USING with o

Re: Probably dumb assembler question

2019-07-11 Thread pgs4ibmm...@pacbell.net
Is the 'XDCBL ' value a multiple of 8? The STORAGE OBTAIN will round up to multiple of 8; the STORAGE RELEASE might be using the actual 'XDCBL' length.

Re: Probably dumb assembler question

2019-07-11 Thread pgs4ibmm...@pacbell.net
Why don't you get a dump of the 878 and try to debug? Maybe you can find 1900 chunks of the same storage eye-catchers.

Re: Any real need for sequence numbers in 73-80 any more?

2017-12-13 Thread pgs4ibmm...@pacbell.net
One must remember that the the machine (CPU) does not care about how the source looks. Once the machine gets the code, it is just opcodes. There is no 'SET SOURCE CODE IS BEAUTIFUL' instruction to make it run faster. FB 80 files, with or without sequence numbers, work for me as a source reposito