IMHO this is not a good idea, because the C/C++ runtime probably
did more things than simply construct a DCB and open it; IMO it
allocated storage for buffers etc., too.
And if you "steal" the open DCB from it, you will leave the C/C++
runtime in an undefined state,
at best, even if you plan not to use the C/C++ file functions after the
fopen() call;
you will have a storage leak IMO.
And: no, I have no idea how to get the DCB from the C/C++ FILE structure.
(I am a compiler maintainer myself, and I cannot imagine someone
successfully
using the DCBs, which are part of my Pascal File control structures, and
doing I/O using them,
this way "fooling" the Pascal runtime system ... but, who knows).
Kind regards
Bernd
Am 14.07.2021 um 16:44 schrieb Larry Slaten:
I am attempting to use the C/C++ FOPEN function to open a SYSLIB DD
pointing to a concatenation of executable libraries. Then extract the
DCB address so that I can use it to populate the LOAD and BLDL macro.
This would make it easier to keep the root program RENT. I realize
that I can code the DCB in the root assembler program and do what is
necessary to keep the program RENT. But, using the C/C++ functions
would make it simpler. I have used the FOPEN function to open the
data set, retrieved the file pointer. The file pointer points to
another 31-bit address. I have dumped the storage around that 31-bit
address, it still doesn't look like a DCB. I do see "SYSLIB" at +84
bytes, so I appear to be close to the answer. Does anybody know to
obtain the DCB address using the file pointer returned by FOPEN?
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
