579f97d5c5
The interrupt line registers for the devices wired directly to the CPU
...
are not meaningfull. Instead match them by {bus,device,pin}.
2000-04-09 00:13:27 +00:00
43a7ae04c9
Disabling PCI Memory Read Multiple improves Tulip transmit performance 30%.
2000-04-09 00:11:51 +00:00
1c5551f260
splsoftnet() should block softclock() too.
2000-04-03 11:44:19 +00:00
076e3ffa1c
Update.
2000-04-03 00:20:11 +00:00
0551ccb4d4
Typo.
2000-03-31 15:00:49 +00:00
278498c05f
Some cleanups.
2000-03-31 14:51:49 +00:00
fab7d466f3
Add little to-do list.
2000-03-30 22:14:17 +00:00
70abbd3c24
Crude kernel installer.
2000-03-30 21:46:16 +00:00
80e6d43253
Add asm.h like other ports.
2000-03-30 21:33:30 +00:00
471a03f8f6
Remove local hack.
2000-03-30 20:08:37 +00:00
712cb103ff
Customary GENERIC configuration.
...
Unfortunately, the Cobalt firmware seems to fail loading kernels larger
than about 2.5 megs before compression, so without a boot loader, this
is a rather amputated GENERIC.
2000-03-30 20:01:02 +00:00
0076cd46eb
- Nuke unnecessary cast.
...
- Change comments on cpu_startup() so as what it does (XXX found mostly
common across ports).
- Retain UNIX heritage of /* Good {morning,afternoon,evening,night} */.
2000-03-27 01:51:17 +00:00
795ee9db8e
Remove unnecessary bzero() op for proc0's USPACE as pmap_steal_memory()
...
returns a nullified area.
2000-03-27 01:24:55 +00:00
61609b54e2
Make sure proc0 PCB has spl0 condition in CP0 status register field.
...
cpu_fork() mistakenly created processes forked by proc0, including
kthreads, in splhigh condition, because [1] proc0's PCB was zero
cleared during initialization, and [2] value 0 in status register
field made processes to have splhigh condition when CPU tick was
assigned for them. This mostly doesn't matter as forked processes
dive immediately into user mode through proc_trampoline code path,
however, kthreads never do that and remain in splhigh.
Reported by Ethan Solomita <ethan@geocast.com>.
2000-03-25 10:14:13 +00:00
573160e03b
Revert previous.
2000-03-24 23:06:03 +00:00
c535ede30b
Move sysctl definitions from arch/mips to arch/foo.
2000-03-24 21:30:58 +00:00
b2386045f0
Protect against multiple inclusion.
2000-03-24 20:16:27 +00:00
0a40850f62
Let's try NCR_IOMAPPED.
2000-03-24 13:08:30 +00:00
608ce86497
- Have physical address for MIPS_PHYS_TO_KSEG1() macro.
...
- Make consistent function declarations.
- White spaces.
2000-03-23 08:09:54 +00:00
f3ca63ef9e
Note that this is just for compatibility.
2000-03-22 21:15:59 +00:00
5be5f12199
The firmware uses the clock in BCD mode.
2000-03-22 20:38:22 +00:00
793382b291
Revert to the same scheme as the other MBR-using ports.
2000-03-22 20:38:05 +00:00
7c35662066
add commented out option PCI_CONFIG_DUMP whever there's a PCIVERBOSE.
2000-03-22 00:58:16 +00:00
826e64bd17
Add math.h.
2000-03-21 02:48:17 +00:00
7db0d53d01
Add math.h.
2000-03-21 02:30:42 +00:00
28fba2d433
Blocking soft interrupts can be useful..
2000-03-21 02:27:50 +00:00
82f0c9bf4a
Add bounds_check_with_label().
2000-03-21 02:26:48 +00:00
b124a5ec8d
Extend the evil interrupt kludge to the second ethernet.
2000-03-21 01:05:53 +00:00
d71e4cbde1
Use correct MACHINE_ARCH.
2000-03-20 12:26:11 +00:00
4c547143d4
NetBSD/cobalt. Work-in-progress.
2000-03-19 23:07:43 +00:00
soren