On Tue, 13 Feb 2018 17:46:28 -0800, Charles Mills wrote:

>And no real reason a program could not support 64ki-1 bytes. Not looking at 
>the specs for "standard linkage" but there is no real good reason to treat the 
>length as signed.
> 
Tradition?  ("real good reason"?)

A while ago, I experimented.

BPXBATCH reasonably processes x'FFFF'.

Rexx address LINKMVS considers >=x'8000' a syntax error.

HLASM reasonably processes x'7FFF'.  At x'8000' it issues several hundred 
thousand
lines of error messages then program checks.  Ugh!  Sign extension?

BPXWDYN reasonably processes X'37FF'.  At x'4000' it fails, expecting a 
null-terminated
string rather than a CALL-style parm.  This has been documented as a 
restriction ever
since my RCF.  The reporting is not ideal.

>Beyond that would require a significant change in the linkage.
>
Yes.

-- gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN

Reply via email to