werner@ut-ngp.UUCP (Werner Uhrig) (08/24/86)
I just received Dr Dobbs Journal (DDJ) .. September 1986, and a recent discussion in net.micro.mac of reliability-questions when using a Mac in critical applications (due the fact of the lack of parity-bits in memory) came to mind, when I read the title of an article on page 112: Memory test with the 68000 - THE RIGHT TO ASSEMBLE: The Worm Memory Test (by Jan W. Steinman). which includes the code ready to run on a Tektronix 4404, with only 2 system dependent instruction [claim of the author, not mine] he seems to have used it on the Amiga, also, but no mention of the MAC is made [I only scanned the article, though]. to quote a few lines: "The test uses a dynamically executing program as the actual test data Unlike previous programs of this type, however, this worm has a special twist - it is able to overlay itself while it is executing thanks to the MC68000's pre-fetch register. .... ... I originally developed the MC68000 Worm test for an embedded processor application that was having dynamic RAM refresh problems. It was discovered that conventional RAM tests, which move smoothly up through consecutive addresses, were masking the problem by unintentionally providing software refresh. The test is not long enough to cause a complete cycle of all a dynamic RAM's row-address strobe (RAS) lines and was able to help diagnose the problems." I doubt that I'll get to do anything more to try to install this on the MAC, but I'm sure (well, hope) someone out there is tempted. I know, DDJ makes their code available online somewhere, but I can't find the number anywhere in the magazine (or my phone-files on the computer for that matter) but I found a sales pitch in the magazine for their code on floppies; $15 - call (800)528-6050 ext.4001 Cheers, ---Werner
werner@ut-ngp.UUCP (Werner Uhrig) (08/24/86)
> I know, DDJ makes their code available online somewhere, but I can't find the > number anywhere in the magazine ... for those with CompuServe accounts, "go ddj" gets you this months sources ... this info was provided not at a logical place (with the sources, with the ad for the sources on floppies, or in the masthead) but in an editorial column somewhere ...
jans@tekecs.UUCP (Jan Steinman) (08/25/86)
In article <3867@ut-ngp.UUCP> werner@ut-ngp.UUCP (Werner Uhrig) writes: >I just received Dr Dobbs Journal (DDJ) .. September 1986... Memory test >with the 68000 - THE RIGHT TO ASSEMBLE: The Worm Memory Test (by Jan W. >Steinman). which includes the code ready to run on a Tektronix 4404, with >only 2 system dependent instruction... I know, DDJ makes their code >available online somewhere... The system dependent parts of the Worm have to do with turning off interrupts, which is likely to be different on different systems. On the Amiga, it is a single instruction. If anyone wants the Worm source code, I will gladly mail it out. If I get more than a few replies, I'll stick it in net.sources with a pointer here. -- :::::: Artificial Intelligence Machines --- Smalltalk Project :::::: :::::: Jan Steinman Box 1000, MS 60-405 (w)503/685-2956 :::::: :::::: tektronix!tekecs!jans Wilsonville, OR 97070 (h)503/657-7703 ::::::