Back to home page

DOS ain't dead

Forum index page

Log in | Register

Back to the forum
Board view  Mix view

DPMILD32 issues - spawn trouble: solution found ??? (DOSX)

posted by Japheth Homepage, Germany (South), 20.02.2008, 12:13

> In other forum you mentioned DPMILDR=8 only, so I missed the HDPMI=32 :-(

It's explained very lengthy in COMPAT.TXT ...

> 1. DPMILDR=8 : but why not suppress this hack, check for address space &
> fixups, and load into same if possible, and into a new one otherwise ?

I can't remember the reasons, but there WERE strong reasons why this didn't work. If you can supply a working patch ...

> 2. HDPMI=32 : OK, making this the default would hog low memory :-(

It's not just the memory. It's simply not DPMI 0.9 compatible, would make DPMI v0.9 TSRs impossible. And one of my favorite toys - debxxf - also runs as separate client and wants direct access to the debuggee's memory, IDT and LDT.

> Anyway, in DOS there is no true multitaxing, just spawning, A spawns B,
> thus A is suspended while B is running, when B has done, A resumes ... and
> the obvious (or did I miss something ?) solution arises: when A spawns B,
> just allocate a memory block with size of A at highest free / some high
> and free address, move A there, free the famous throne at $400000, load B
> into there, run B, and when B has done just move A back and resume it
> :clap:

Yes, that's possible, but is just another hack. For Win32 programs, the HDPMI=8 "hack" is preferable IMO since it is very stable and, more important, this appproach will finally allow to run several clients/tasks concurrently. Some Win32 applications need that (i.e. CygWin).

---
MS-DOS forever!

 

Complete thread:

Back to the forum
Board view  Mix view
22049 Postings in 2034 Threads, 396 registered users, 251 users online (0 registered, 251 guests)
DOS ain't dead | Admin contact
RSS Feed
powered by my little forum