On Tue, 2 Sep 2025 09:55:05 -0500, Paul Gilmartin <[email protected]> wrote:
>On Mon, 1 Sep 2025 11:47:49 -0500, Paul Edwards wrote:
>> ...
>>>Can it deal with the prevalent FAT-32 USB flash drives from
>>>an Intel Windows system?
>>
>>No, this is a pure EBCDIC system. It does indeed use FAT-32,
>>but it is EBCDIC FAT-32.
>>
>That shouldn't be an ultimate obstacle.
You are correct.
> I have used
>NFS with a Solaris (ASCII) server and MVS (EBCDIC)
>client. Software performs the 819<->1047 translation
>for both directories and member data.
Sure. I absolutely could do that, and probably will do that
at some point. I just haven't done that yet because it hasn't
been priority.
My current priority is to see if I can run this:
* LG R15,=VD(TWO)
LG R15,XXX
...
LTORG
DS 0D
XXX DC V(TWO)
DC A(0)
through as370, which produces 32-bit ELF, but knows
the 64-bit instructions, and doesn't accept VD and this:
D:\devel\binutils\gas\config>cvs diff tc-i370.c
Index: tc-i370.c
===================================================================
RCS file: \cvsroot/binutils/gas/config/tc-i370.c,v
retrieving revision 1.69
diff -r1.69 tc-i370.c
1228a1229,1232
> if (*input_line_pointer == 'D')
> {
> input_line_pointer++;
> }
1425c1429
< align2 = 2;
---
> align2 = /*2;*/ 3;
1528c1532
< for (i=0; i<npeat; i++) emit_expr (&xexp, 4);
---
> for (i=0; i<npeat*2; i++) emit_expr (&xexp, 4);
2176a2181,2189
> if (*input_line_pointer == 'D')
> {
> ++input_line_pointer; /* Skip D */
> cons_len = 8;
> }
> else
> {
> cons_len = 4;
> }
2184c2197
< add_to_lit_pool (xexp, sym_name, 4);
---
> add_to_lit_pool (xexp, sym_name, cons_len);
D:\devel\binutils\gas\config>
didn't make it accept VD, and I'm reluctant to spend any
more brainpower, for free, on copyrighted freeware, which
means I instead need to handcode the workaround VD
replacement, and then get pdld to accept a parameter to
force all 32-bit V references into 64-bit V references to
produce a 64-bit MVS load module.
So then I will be in a position to create 64-bit load modules
on any system.
And I have already demonstrated I can load 64-bit MVS
load modules.
So I will be in a good position.
>Cygwin? WINE? Bash?
What about them?
>If your motivation is EBCDIC advocacy,
One of my motivations is to ensure I have portable C
code, and that includes no assumptions about ASCII,
and building an EBCDIC version of anything should
be trivial.
> you're moving against the current.
I don't really care where the current is going.
Well, I care to the same extent that this guy does:
https://www.quora.com/Why-did-you-leave-a-job-as-a-software-engineer/answer/Jeff-Sturm-2
Probably more than he does. He simply left. I went
back to basics to see if I could create something I
was comfortable with, and that didn't match the
description he gave.
BFN. Paul.
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
