My P3 runs at.... zero Mhz (bug rpt)

Just booted into 2.5.67-BK-latest (plus my __builtin_memcpy patch). Everything seems to be running just fine, so naturally one must nitpick little things like being told my CPU is running at 0.000 Mhz. :)

2.4.x reports the Mhz correctly. I don't recall if 2.5.x did in the past, but can check if needed.

sh-2.05b$ cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 7
model name : Pentium III (Katmai)
stepping : 3
cpu MHz : 0.000
cache size : 512 KB
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 2
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 sep mtrr pge mca cmov pat pse36 mmx fxsr sse
bogomips : 494.59

Selected dmesg bits:
Calibrating delay loop... 494.59 BogoMIPS
Memory: 644596k/655296k available (2192k kernel code, 9960k reserved, 694k data, 296k init, 0k highmem)
CPU: L1 I cache: 16K, L1 D cache: 16K
CPU: L2 cache: 512K
CPU serial number disabled.
CPU: After generic, caps: 0383f9ff 00000000 00000000 00000000
Intel machine check architecture supported.
Intel machine check reporting enabled on CPU#0.
CPU: Intel Pentium III (Katmai) stepping 03
Enabling fast FPU save and restore... done.
Enabling unmasked SIMD FPU exception support... done.
Checking 'hlt' instruction... OK.
IA-32 Microcode Update Driver: v1.11 <[email protected]>
Enabling SEP on CPU 0

From: Jeff Garzik <[email protected]>
Date: Thu, Apr 17, 2003 at 10:10:53PM -0400

fwiw, my Athlon XP2400 does the same in 2.5.67-ac1:

processor : 0
vendor_id : AuthenticAMD
cpu family : 6
model : 8
model name : AMD Athlon(tm) XP 2400+
stepping : 1
cpu MHz : 0.000
cache size : 256 KB
bogomips : 1970.17

Interesting enough, even the bogomips have halved w.r.t. earlier 2.5.x kernels and 2.4.2x kernels. Booting a 2.4.21-pre7 kernel on the same machine gets me close to 4000 bogomips, as is proper for a 2 GHz cpu.

The ioapic code does report the correct bus-speed however.

Ah well, they were bogus to begin with.

"I am not to be trifled with!"
"I didn't bring a trifle," Hawk said to Fisher. "Did you think to
bring a trifle?" "Knew I forgot something," said Fisher.
	Simon R Green - Beyond the Blue Moon
Debian (Unstable) GNU/Linux 2.5.67-ac1 1970 bogomips load av: 1.48 1.62 1.79

Missing parentheses in mach_countup().

diff -puN include/asm-i386/mach-default/mach_timer.h~a include/asm-i386/mach-default/mach_timer.h
@@ -40,7 +40,7 @@ static inline void mach_countup(unsigned
 {
 	*count = 0L;
 	do {
-		*count++;
+		(*count)++;
 	} while ((inb_p(0x61) & 0x20) == 0);
 }
 

_

Curious. Do either of you have any cpufreq bits enabled? If so, does it go away if you disable them?
That frobs with cpu_khz, so it *could* be not initialising it someplace. Especially if your hardware turns out to be unsupported by any of the cpufreq backend drivers..

Dave

Dave> Curious. Do either of you have any cpufreq bits enabled? Dave> If so, does it go away if you disable them?
Dave> That frobs with cpu_khz, so it *could* be not initialising Dave> it someplace. Especially if your hardware turns out to be Dave> unsupported by any of the cpufreq backend drivers..
It does not help me with 2.5.67-ac2 + pcmcia patch. I get 0.000 MHz, 589.82 BogoMIPS with or without CPUFreq. It did the same thing with 2.5.67-ac1 (did not test w/o CPUFreq).

The box is 600 MHz PIII (Coppermine) in Dell Inspiron 5000. On the plus side I kind of like the # of insns per clock cycle ;-)

As Jeff said it's almost OK otherwise; it hangs on boot without the pcmcia patch and I saw USB Storage oopses.

Petr

Its a bug in the mach- patches. Someone sent a fix to l/k

Yep. Linus committed a better fix, too. :)

Jeff