Date: Mon, 29 Jan 2001 07:20:55 +1100 From: Peter Jeremy <peter.jeremy@alcatel.com.au> To: freebsd-alpha@freebsd.org Subject: Multia problems Message-ID: <20010129072055.X52423@gsmx07.alcatel.com.au>
next in thread | raw e-mail | index | archive | help
I have a Multia VX40 that is displaying normal `dodgy hardware' symptoms - specifically random segmentation violations whilst trying to make world. Having poked around, I've found that writing 0x0000000000000000/ 0xFFFFFFFFFFFFFFFF to alternate longs shows up the problem more reliably: Bits 27-31 (the top 5 bits of the low word) occasionally read as 0 instead of 1 (but never the opposite). Swapping the memory modules around does not affect the problem - which pretty well clears the RAM. Based on re-reading and cache-flushing, the problem seems to be the CPU reading the L2 (external) cache. The problem does not seem to occur using 0x5555555555555555/0xAAAAAAAAAAAAAAAA as a pattern. I've already replaced the (notorious) 74F623 and heat does not seem to be a problem. Does anyone have any suggestions on how to further narrow down my problem? Peter To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010129072055.X52423>