With great thanks to Simon Kirby from Stormix, my ping problems (and several
others!) are now resolved. Apparently, this was brought up earlier in the list
and I just missed it (or wasn't subscribed then). My problem came from the two
CPUs being dissimilar speeds.
Here's a patch which resolved my problem by disabling the tsc correction code
(or something). It is against 2.2.14 (clean). Sometime soon, I'll create a
patch which makes this an option in the make menuconfig. I'd like to see code
added to the kernel which realises the CPUs are different and automatically
disables the TSC code. It shouldn't be hard to do as the kernel realises on
bootup that the speeds are different. I'm afraid I don't have enough kernel
experience to write this myself, though.
$ diff -u time.c~ time.c
--- time.c~ Thu Apr 29 12:53:41 1999
+++ time.c Wed Mar 1 00:38:25 2000
@@ -110,7 +110,6 @@
#define TICK_SIZE tick
-#ifndef CONFIG_X86_TSC
/* This function must be called with interrupts disabled
* It was inspired by Steve McCanne's microtime-i386 for BSD. -- jrs
@@ -224,12 +223,7 @@
}
static unsigned long (*do_gettimeoffset)(void) = do_slow_gettimeoffset;
-
-#else
-
-#define do_gettimeoffset() do_fast_gettimeoffset()
-
-#endif
+#define do_gettimeoffset() do_slow_gettimeoffset()
/*
* This version of gettimeofday has microsecond resolution
@@ -643,7 +637,9 @@
*/
dodgy_tsc();
-
+
+ fast_gettimeoffset_quotient = calibrate_tsc();
+#if 0
if (boot_cpu_data.x86_capability & X86_FEATURE_TSC) {
unsigned long tsc_quotient = calibrate_tsc();
if (tsc_quotient) {
@@ -667,6 +663,7 @@
}
}
}
+#endif
#ifdef CONFIG_VISWS
printk("Starting Cobalt Timer system clock\n");
--
Christopher Thompson http://hypocrite.org/
The unexamined life is not worth living
-
Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/dmentre/smp-howto/
To Unsubscribe: send "unsubscribe linux-smp" to [EMAIL PROTECTED]
