kenw@noah.arc.CDN (Ken Wallewein) (07/23/87)
The August/86 and June/87 issues of the VAX Professional contain articles on the PDA device. Summary: it's slower than an RA81. Reason: it doesn't use DMA, and it MOVB rather that MOVCx to move data, to avoid page faults in IPLFORK. Incidentally, I don't understand why there couldn't be some way to predict when page faulting would occur, so repeated MOVCx instructions could be used. Looks like you'd be better off optimizing your caching. BTW, from what I can see (I'm no expert), INSTALL works for data files, too. /kenw
jjo@calmasd.GE.COM (Jay Olson) (07/23/87)
In article <8707230126.AA12895@ucbvax.Berkeley.EDU>, David G. Dugal (DUGAL%ECL1@STAR.STANFORD.EDU) posts a uuencoded version of a "public domain" RAMdisk driver. In article <8707230820.AA19500@ucbvax.Berkeley.EDU>, Glenn Everhart (EVERHART%ARISIA@rca.COM) posts a disassembled version of this driver. Based on inspection of the disassembled code, I would say you are looking at the standard PDDRIVER, which has come with every VMS system since 4.2. As the author of the VAX/VMS PDDRIVER, I can assure you that I can recognize the code I wrote. For further evidence, just do an ANALYZE/IMAGE SYS$SYSTEM:PDDRIVER.EXE and note that the link date/time is identical to that in the disassembled listing. If this is the case, this code was furnished under terms of a license agreement and is NOT in the public domain. Jay Olson a consultant working at GE Calma San Diego, CA ...sdcsvax!calmasd!jjo jjo@calmasd.GE.COM