From:	Charles Sandmann <sandmann AT new-orleans DOT NeoSoft DOT com>
Subject:	QDPMI VM bug workaround
To:	djgpp AT sun DOT soe DOT clarkson DOT edu (DJGPP Mailing List)
Date:	Thu, 22 Dec 1994 21:57:04 -0600 (CST)

In the DPMI-only GO32 testbed, I hacked "initsize" into the GO32 env
variable and was able to run V1.12's pagetest under QDPMI+VM with many more
MB's than my machine really has (BTW, the VM algorithm in QDPMI sucks,
it pages any time I use over 50% of the real memory with VM on?!?).

This test code doesn't work real well in the real world, since 
nested programs (make+gcc+cc1) then take 3*initsize memory.  So, you 
end up with a huge swapfile which is rarely (but too frequently) used.
But this workaround is probably better than no workaround.

It should would be easier if Quarterdeck had just done resize memory
block right to begin with (or fix it and offer free patches ...)

Anyway.

- Raw text -