Re: When Good DIMMS go Bad (or how I fixed my sig11)
On Mon, 7 Aug 2000, David Scheidt wrote: > On Mon, 7 Aug 2000, Warner Losh wrote: > > :In message <[EMAIL PROTECTED]> David >Scheidt writes: > :: convince people that their memory is bad. The only reliable way to test > :: memory is with a hardware testor, or swapping known good memory in. > : > :Yes. while (1) do ; make world; done is a close second to a hardware > :tester. > > Ah, that tells you have a problem. It unfortunatly, doesn't distinguish > a bad memory module from a bad memory bus. One of my abits blew up a bit > ago with SIGSEGVs, I swapped memory in and around till I got to the point > that I realized that as long as I didn't populate the last DIMM slot, it > worked fine. It's not long for this earth, that machine. Many motherboards are unstable when you populate all DIMM slots. You generally have limitations to what types of DIMMs you can use (i.e. single-sided only, registered only, etc.) when you do populate all of them. The manual _should_ specify these limitations. -- Chris Dillon - [EMAIL PROTECTED] - [EMAIL PROTECTED] FreeBSD: The fastest and most stable server OS on the planet. For Intel x86 and Alpha architectures. ( http://www.freebsd.org ) To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: When Good DIMMS go Bad (or how I fixed my sig11)
On Mon, 7 Aug 2000, David Scheidt wrote: >Ah, that tells you have a problem. It unfortunatly, doesn't distinguish >a bad memory module from a bad memory bus. One of my abits blew up a bit >ago with SIGSEGVs, I swapped memory in and around till I got to the point >that I realized that as long as I didn't populate the last DIMM slot, it >worked fine. It's not long for this earth, that machine. Reminds me of the 4-5 SIMM pair on Tyan Tomcat P5 SMP motherboards. They're notorious for not working. I've got a pair of those boards and cannot put exactly 6 SIMMs despite the claims of requiring pairs. It *really* wants SIMMs installed 4 at a time. They were nice boards otherwise though. Brandon D. Valentine -- bandix at looksharp.net | bandix at structbio.vanderbilt.edu "Truth suffers from too much analysis." -- Ancient Fremen Saying To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: When Good DIMMS go Bad (or how I fixed my sig11)
On Mon, 7 Aug 2000, Warner Losh wrote: :In message <[EMAIL PROTECTED]> David :Scheidt writes: :: convince people that their memory is bad. The only reliable way to test :: memory is with a hardware testor, or swapping known good memory in. : :Yes. while (1) do ; make world; done is a close second to a hardware :tester. Ah, that tells you have a problem. It unfortunatly, doesn't distinguish a bad memory module from a bad memory bus. One of my abits blew up a bit ago with SIGSEGVs, I swapped memory in and around till I got to the point that I realized that as long as I didn't populate the last DIMM slot, it worked fine. It's not long for this earth, that machine. David To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: When Good DIMMS go Bad (or how I fixed my sig11)
In message <[EMAIL PROTECTED]> David Scheidt writes: : convince people that their memory is bad. The only reliable way to test : memory is with a hardware testor, or swapping known good memory in. Yes. while (1) do ; make world; done is a close second to a hardware tester. I can't tell you the number of times I've had flakey systems that made people sure FreeBSD was busted. A new CPU, mobo or memory fixed these right up. Troubleshooting that can be interesting... Warner To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: When Good DIMMS go Bad (or how I fixed my sig11)
On Fri, 4 Aug 2000, Mike Muir wrote: :Stephen Hocking wrote: :> :> About a week ago, I complained of mysterious Sig 11s during a make world. :> After some experimentation, a PC100 DIMM was found to be better suited for a :> 66MHz memory bus in another machine, who obligingly donated a DIMM in return :> that actually works with a 100MHz bus. I think the trip from Australia and :> this Texas heat finally pushed the dodgy one over the edge. : :Have you tried any memory testing routines such as memtest86 ? Its the :only you write to a floppy and it runs before any bootstrap kicks in -- :independant of the OS -- and takes around 18 hours for a single pass. It :appears to be quite a comprehensive torture test. If so, how did that Software memory testers don't work. They may sometimes find problems, true, but if they don't, it doesn't mean the memory is good. Lots of failures are only triggered by certain access paterns, which is why it's so hard to convince people that their memory is bad. The only reliable way to test memory is with a hardware testor, or swapping known good memory in. David To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: When Good DIMMS go Bad (or how I fixed my sig11)
Stephen Hocking wrote: > > About a week ago, I complained of mysterious Sig 11s during a make world. > After some experimentation, a PC100 DIMM was found to be better suited for a > 66MHz memory bus in another machine, who obligingly donated a DIMM in return > that actually works with a 100MHz bus. I think the trip from Australia and > this Texas heat finally pushed the dodgy one over the edge. Have you tried any memory testing routines such as memtest86 ? Its the only you write to a floppy and it runs before any bootstrap kicks in -- independant of the OS -- and takes around 18 hours for a single pass. It appears to be quite a comprehensive torture test. If so, how did that dodgy DIMM perform? (The reason I ask is that I'm interested in knowing if these tests can reveal the problems that building world did in your situation.) -mike To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
When Good DIMMS go Bad (or how I fixed my sig11)
About a week ago, I complained of mysterious Sig 11s during a make world. After some experimentation, a PC100 DIMM was found to be better suited for a 66MHz memory bus in another machine, who obligingly donated a DIMM in return that actually works with a 100MHz bus. I think the trip from Australia and this Texas heat finally pushed the dodgy one over the edge. Stephen -- The views expressed above are not those of PGS Tensor. "We've heard that a million monkeys at a million keyboards could produce the Complete Works of Shakespeare; now, thanks to the Internet, we know this is not true."Robert Wilensky, University of California To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message