Newsgroups: comp.os.msdos.djgpp From: gaminer AT undergrad DOT math DOT uwaterloo DOT ca (Glen Miner) Subject: Re: Locking all code/data Sender: news AT undergrad DOT math DOT uwaterloo DOT ca (news spool owner) Message-ID: Date: Thu, 29 May 1997 19:32:13 GMT References: <5mg06f$r99 AT news DOT epcc DOT edu> Nntp-Posting-Host: lagrange.uwaterloo.ca Organization: University of Waterloo Lines: 25 To: djgpp AT delorie DOT com DJ-Gateway: from newsgroup comp.os.msdos.djgpp Precedence: bulk >> I'm working on a program that makes extensive use of hardware driver >> interrupts and speed sensitive code. It is required that several >> functions and variables be locked so that they never get swapped into >> virtual memory. I'm lazy: I don't want to lock each one seperately, I >> want all of the code and data locked right off the bat (any newly >> allocated memory can be unlocked, and really ought to be in my case). >_CTR0_FLAG_LOCK_MEMORY does indeed do what you want; it locks all memory >that your program uses, including all functions and all data. Okay, that's cool then. So long as I never have to worry about pages being swapped out, I'm happy. >Alternatively, you can use CWSDPR0 (a Ring 0 version of CWSDPMI) of PMODE\DJ. But what if the system was already running a DPMI host? Wouldn't it be used instaed of whatever I stub on? In that case, I would have no guarantees that the extender I ship would in fact get used, right? Thanks, Peace -- ===[ Gabo / [ABC] : gaminer AT undergrad DOT math DOT uwaterloo DOT ca ]=================== Latest ABC Shogi: http://www.undergrad.math.uwaterloo.ca/~gaminer/shogi.html "What Greenpeace spends in a year General Motors spends in four hours" -Moby