Also, someone may want to take a look at this web page:
http://www.sandpile.org/80x86/cpuid.shtml
It's got some stuff under the AMD section that I'm going hack together an
implementation for. Basically, it appears that the x86 capabilities flags
are not completely accurate with the K[56]. There appear to be some
differences (and between the new K6 models too...). I'll have something
soonish, though I've GOT to finish a patch for tin first...
--- cut here ---
*** linux/arch/i386/kernel/setup.c.orig Wed Sep 10 09:06:32 1997
--- linux/arch/i386/kernel/setup.c Wed Sep 10 09:08:42 1997
***************
*** 257,263 ****
{
static const char *model[] = {
"SSA5 (PR-75, PR-90, PR-100)", "5k86 (PR-120, PR-133)",
! "5k86 (PR-166)", "5k86 (PR-200)", "", "", "K6"
};
if (nr < sizeof(model)/sizeof(char *))
return model[nr];
--- 257,264 ----
{
static const char *model[] = {
"SSA5 (PR-75, PR-90, PR-100)", "5k86 (PR-120, PR-133)",
! "5k86 (PR-166)", "5k86 (PR-200)", "", "",
! "K6 (PR-166 .. PR-266)", "K6 (PR-166 .. PR-300)"
};
if (nr < sizeof(model)/sizeof(char *))
return model[nr];
--- cut here ---
Pat
-- +----------------------------------------------------------------------------+ | Patrick St. Jean '97 XLH 883 psj@cgmlarson.com | | Programmer & Systems Administrator +1 713-977-4177 x106 | | Larson Software Technology http://www.cgmlarson.com | +----------------------------------------------------------------------------+