Re: AMD K5 with Linux

Danny ter Haar (
18 Jul 1996 22:03:31 +0200

In article <>,
<> wrote:
>Ok, we kicked the M1 around about a month ago, now how about
>AMD's K5. Based on AMD's docs it can execute twice as
>many instructions per clock as the P5 and is P5 plug-in
>compatible. Anyone tried it with Linux and is it SMP plug-in
>capable as well?
> Melvin

We installed with a AMD K5 today.
Several weird errors occured after about 15 minutes of operation:
divide errors on several processes. (even ls)
But after a fresh reboot it would be fine for about 10-15 minutes.
Just like a register running over.

We "fixed" it though:

We compiled a kernel wit copro support and started a kernel
with the no387 option, thus forcing the copro emulation and the
errors didn't return (yet).

The motherboard we installed ahd the latest triton chipset
and a dimm memory connector.

Here are the bootup messages, look at the bogomips ;-)

Console: 16 point font, 400 scans
Console: colour VGA+ 80x25, 1 virtual console (max 63)
pcibios_init : BIOS32 Service Directory structure at 0x000fdb60
pcibios_init : BIOS32 Service Directory entry at 0xfdb70
pcibios_init : PCI BIOS revision 2.10 entry at 0xfdb91
Probing PCI hardware.
Warning : Unknown PCI device (8086:7030). Please read include/linux/pci.h
Calibrating delay loop.. ok - 149.91 BogoMIPS
Memory: 6792k/8192k available (640k kernel code, 384k reserved, 376k data)
Swansea University Computer Society NET3.035 for Linux 2.0
NET3: Unix domain sockets 0.12 for Linux NET3.035.
Swansea University Computer Society TCP/IP for NET3.034
IP Protocols: ICMP, UDP, TCP
Checking 'hlt' instruction... Ok.
Linux version 2.0.7 (miquels@Q) (gcc version 2.7.2) #1 Thu Jul 18 16:35:22 MET DST 1996
Serial driver version 4.13 with no serial options enabled
tty00 at 0x03f8 (irq = 4) is a 16550A
tty01 at 0x02f8 (irq = 3) is a 16550A
lp1 at 0x0378, (polling)
hda: QUANTUM FIREBALL1080A, 1039MB w/83kB Cache, LBA, CHS=528/64/63
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
Floppy drive(s): fd0 is 1.44M
Started kswapd v
FDC 0 is an 8272A

It seems there is a bug in the AMD or in the kernel somewhere ;-)