3f7694f804
Adapt these for use on the Milan.
2001-04-11 14:45:07 +00:00
5dd91b669c
add VR4101 specific BCU registers.
...
add VR4122 specific BCU registers.
add cpu specific comments.
2001-04-11 08:52:59 +00:00
68d21310de
opt_vr41x1.h -> opt_vr41xx.h.
...
add 'options VR41XX' and kill 'options VR41X1'.
add 'options VR4101' (not yet supported)
add 'options VR4102'
add 'options VR4111'
add 'options VR4121'
add 'options VR4122' (not yet supported)
add 'options VR4181' (not yet supported)
these are for config level size optimization.
currenty, no optimization codes are exist.
NOTE: there are some incompatibility in VR4181 & VR4101 &
VR4102,4111,4121 group & VR4122.
GENERIC (for VR41XX series) has options VR4102, VR4111, VR4121.
2001-04-11 08:23:26 +00:00
4290b6b288
Make sure that this will compile on a big-endian system.
2001-04-11 01:47:18 +00:00
d039e70428
Remove the "asc" driver. It's never worked, and we should have an MI version
...
sometime.
2001-04-10 21:33:00 +00:00
247d880fc9
The Milan uses the U0 pin to switch the pci-bridge between little & big
...
endian mode. That's why I moved the 'wired' flag to U1.
2001-04-10 19:51:00 +00:00
dcc1c6a9fb
Take NFDISA into account.
2001-04-10 19:36:20 +00:00
4c8caed00d
Apply a bandage to delay() so that it takes a more suitable time on
...
StrongARM Risc PCs. It really should be replaced entirely, but I've got more
important bugs to fix.
2001-04-10 19:08:42 +00:00
edde8b993b
Add a DIAGNOSTIC check to see if delay() waits long enough (it doesn't on
...
my machine).
2001-04-10 18:58:56 +00:00
dc0e725f9d
full[tc]flush from GLOBAL -> ASLOCAL.
2001-04-10 13:59:10 +00:00
7d1452855c
Fix bus_space_write_multi_* in the memory case; the asm constraints
...
were wrong.
2001-04-10 12:59:32 +00:00
b27eda4ead
Added.
2001-04-10 12:46:26 +00:00
1dfde059ab
Use fb_setsize_eeprom to do some intialization of the
...
fb structure.
2001-04-10 12:44:12 +00:00
bb09d2bc78
Fixed a bug where we were trying to map the board RAM
...
using the physical address of the board registers.
2001-04-10 12:42:51 +00:00
ca414a50ad
In prom_init, when making our local copy of the bootparams,
...
be sure to calculate the shift in terms of chars, not struct
bootparams, and don't shift undefined arguments.
2001-04-10 12:40:53 +00:00
8efc1dcff9
Fixed has_context to return true on the kernel pmap,
...
which always has context zero. This was causing panics
under DIAGNOSTIC.
2001-04-10 12:39:56 +00:00
857e8fbc6b
The mainbus doesn't define a device name in the attach args
...
after attaching the special busses, so just match when it
isn't, instead of calling strcmp with it.
2001-04-10 12:37:49 +00:00
c6871b7720
Make softintr_handler return one, not zero, to properly
...
claim the interrupt. Added special bus_space_probe
handling to sun2_bus_map.
2001-04-10 12:33:09 +00:00
af0a9202ce
Under DIAGNOSTIC, disabled a test that was using an
...
instruction not available on the 68010. Also removed
more old sun3 cache-related code.
2001-04-10 12:29:10 +00:00
c4973c2333
In fpu_probe, now be sure to return FPU_NONE. Returning
...
FPU_UNKNOWN was causing confusion in cpu_switch.
2001-04-10 12:27:40 +00:00
0cb40c9174
Define a very small UBC_NWINS to keep UBC from eating all kernel VM.
2001-04-10 12:24:05 +00:00
0050ddd2b2
Be paranoid, #undef PAGER_MAP_SIZE before defining it.
2001-04-10 12:23:14 +00:00
eeea1bcd05
g/c transfer_check
2001-04-10 10:09:45 +00:00
3e3afaaac3
Regen. again...
2001-04-10 07:03:49 +00:00
e3d2d361fd
The Hades also uses the TT030 like SCSI driver. Remove 'mouse' here...
2001-04-10 06:46:01 +00:00
2d38da2932
Initial MILAN generic config
2001-04-10 06:39:36 +00:00
e409b5d5df
Regen.
2001-04-10 06:39:10 +00:00
99047718f0
fdisa needs-flag...
2001-04-10 06:37:08 +00:00
30f5357c6e
Move dma.c away from the mainbus floppy and make it a separate entity with
...
tag 'mbdma'.
2001-04-10 06:36:08 +00:00
e159fdc910
files.atari:
...
Defopt motherboard type and make ym2149.c dependend on 'standard' atari
boards.
Configfiles:
Add pseudo device mouse. It got dropped in the moves by accident.
2001-04-10 06:28:38 +00:00
dfaa29355b
Add a config file for the Milan. Note that you cannot build a working Milan
...
kernel yet. There are still quite some bits missing in the repository.
2001-04-10 05:56:25 +00:00
1e08d983e3
Move GENERIC to GENERIC.in . This avoids confusion since GENERIC is not a
...
standalone config file anymore.
2001-04-10 05:52:14 +00:00
422ddeb55b
In cpu_fork(), initialize u_pcb.pcb_context[8] so that the newly
...
spawned kthread/process runs at IPL_0 instead of whatever IPL the
parent was running at.
This appears to fix the NTP clock stability problems observed on some
alpha systems; the clock appears stable even when there's heavy
raidframe (i.e., kthread-intensive) I/O under way.
2001-04-10 00:27:00 +00:00
cd24bc13e7
Account for the number of counter ticks that elapse while statintr()
...
processing takes place. This prevents stat clock skew by a factor
approximately proportional to stathz (not a serious problem if stathz==100;
but considerable already if stathz==1000).
2001-04-09 21:28:50 +00:00
2c2e7aca80
Add macro to convert a counter register value to microseconds.
2001-04-09 21:17:33 +00:00
986e7f8f1c
Interrupt latency fix from sparc port. (Hm. One insn instead of 5).
2001-04-09 21:15:42 +00:00
b6c1c5d49a
Altentries should also be marked as functions, to keep linker happy.
2001-04-09 20:31:20 +00:00
3e6eaba006
The old vidc console code can't handle all display depths
2001-04-09 19:33:22 +00:00
ba53dd71d7
Re-enable setmode() in wscons implementation.
2001-04-09 19:23:32 +00:00
b592059273
Just some pretty printing for 640 width screens
2001-04-09 19:22:42 +00:00
a0daff5647
proc_trampoline(): mave the new process run at base interrupt priority level.
...
This is especially important for in-kernel processes started from here,
which would otherwise keep running at splclock().
2001-04-09 18:18:48 +00:00
e0d02d24d5
Mark syscalls as functions.
2001-04-09 15:52:59 +00:00
e9d5de6e98
Patch the hardware scrolling to be only used when using VRAM until it is
...
working perfectly with DRAM only.
2001-04-09 13:59:44 +00:00
3aabea5c70
Remove '_KERNEL' conditionalizing. This break userland apps.
...
Noted by Steve Allan.
2001-04-09 05:35:32 +00:00
2a7597f49e
Make this compile w/o ukbd.
2001-04-08 08:53:31 +00:00
ed0aea79a7
Make the NC start run again ... it didnt detect the start of the DRAM good
...
enough.
2001-04-07 20:57:24 +00:00
179671f564
automatically restart dma for unexpected shutdowns
2001-04-07 13:02:55 +00:00
b2e3cc1a2e
continue to tweak nextdma detection of xfer length
...
add check to esp driver for incomplete transfers, but still does not handle them.
2001-04-07 11:29:50 +00:00
824065dc0a
Don't call "is_a1200()" again after its result was stored in the "softc"
...
structure.
2001-04-07 05:09:27 +00:00
adc78de568
Regen.
2001-04-06 21:55:11 +00:00
leo