Re: Fresh install won't compile requirement libraries for cvsup

On Wed, 13 Jun 2007, Andrew Falanga wrote:
On 6/13/07, Ian Smith <smithi@xxxxxxxxxxxxx> wrote:
On Tue, 12 Jun 2007 10:02:38 -0600 Andrew Falanga <af300wsm@xxxxxxxxx> wrote:

> a hard hang. Nothing worked. I could not even Alt+<num> to a
> different pseudo terminal. The system just hard hanged. I rebooted
> and tried the install again with the same result.

At the same place? If so, I'd tend to suspect memory rather than cpu.

No it did not stop at the same place. I still suspect memory versus
the cpu, especially considering that, apparently, there were several
generations of the K6. I didn't know this. For several years, I did
not have the time or money to play with hardware and therefore lost
touch with much of the hardware that was out there.

Andy, you're making me remember stuff I thought I'd about done with :)

> I'm wondering if it could be hardware, specifically memory. I've
> never seen a FreeBSD, OpenBSD or Linux (for that matter) hard hang on
> program compilation apart from hardware problems. Also of particular
[..]
> System configuration is as follows:
>
> AMD K6 700mHz
> 256mb RAM (PC 133)
> 13gb HDD

700MHz? Please show us the line from your /var/run/dmesg.boot showing
the exact cpu and clock. This will also indicate features & stepping

I'm still curious ..

That's a K6-2, though it doesn't say so there. From memory, the fastest
ever K6-2 was ~550MHz, but people did tend to wildly overclock them ..

525MHz, on reflection. I built a box for a friend on a budget and went
for the 450, which another friend had managed to screw up to about 600
with a huge heatsink and fan/s. These were Gigabyte mobos with 100MHz
FSB - I see you've been pointed to wikipedia re that - and PC133 RAM,
and both 'mine' are still rocking 8 years on, running standard spec ..

I'm still rather awed to hear 700MHz though, especially when you say:

pretty sure this was never over clocked. It was the secretaries
computer of the church I attend. I'm working on remaking the system
into a web server as the secretary was just given a laptop.

You'll need to find out the motherboard make/model and google up its
manual to have any chance of resetting bus speed / clock divider and
such. There's another whole trip about setting the cpu core voltage
for different speeds; overclocking these beasts involved black magic to
which whole websites were/are fanatically dedicated, gamer folks mostly.

If you really are running it at 700MHz (at what bus speed setting?) then
I'd treat it to a new heat sink with fresh thermal paste and a BIG fan.

Ok, sounds good. I'll see what I can find for this CPU.

Seeing you'll want to replace the no doubt well dried out thermal paste
anyway, K6-2s are clearly marked with notional speed and model numbers.

And sure it's best to run matched-speed memory. Your BIOS probably lets
you play with wait states and such, but the basic PCI bus speed might be
something weird if you've managed to crank the cpu up to 700MHz ..

How do wait states relate to memory speed? Please enlighten me. I
have an idea, but I'm only theorizing, I'd like to know what it really
means. If it's more in depth than one would like to type in a
response, sending a link is fine. I learned quite a bit on the "Sig
11" links given earlier.

This is drifting well past getting FreeBSD to behave under load on it,
but I googled "AMD K6 wait states" and got heaps of hits, including the
above. Basically, wait states delay the processor long enough to read
or write to comparatively slow memory devices; better to wait than burn.

You mentioned later trying some 100MHz memory, but you'll either need a
slower bus speed than 100MHz (or more / some wait states) for that, and
would be better off finding more PC133 RAM. Dumpster diving, anyone? :)

Cheers, Ian

_______________________________________________
freebsd-questions@xxxxxxxxxxx mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-questions
To unsubscribe, send any mail to "freebsd-questions-unsubscribe@xxxxxxxxxxx"

Relevant Pages

  • Re: How to develop a random number generation device
    ... single shared CPU; much less, in fact, since irrevelant interrupts, ... probably through shared memory and semaphores. ... switches are a reliability issue - processors don't have problems with them. ... If you shift the complexity to hardware, you'd get hardware that is expensive and buggy. ...
    (sci.electronics.design)
  • Re: What Linux distro to use for old Intel machine, that fits on CDs?
    ... CPU: Pentium 3, 933 MHz ... Memory: 512MB ... Considering the uncertainty of what hardware you have, ... path of getting a USB flash based system to boot on ancient hardware. ...
    (comp.os.linux.misc)
  • [PATCH] Document Linuxs memory barriers [try #3]
    ... The attached patch documents the Linux kernel's memory barriers. ... I've tried to get rid of the concept of memory accesses appearing on the bus; ... barring implicit enforcement by the CPU. ...
    (Linux-Kernel)
  • Re: read vs. mmap (or io vs. page faults)
    ... not fit in main memory, and there are overheads related to the heuristics ... But because the CPU is underutilized, ... reasonably sized user buffer). ... You have to measure the actual overhead to see what the actual cost is. ...
    (freebsd-questions)
  • Re: Hardware suggestions for a high volume Linux webserver?
    ... Number of servers, ... > memory, cpu, disk, etc. I'm trying to get an idea of how much we'd have ... > 4gig memory, coupled with a Cisco LocalDirector load balancer. ... The amount of hardware is mostly determined by the amount of interactivity ...
    (comp.os.linux.hardware)