Commit Graph

12159 Commits

Author SHA1 Message Date
is
e76096387d Make multichannel output work. The cc parameter to hw_start_output is
a byte counter, it seems. Somebody should really write man 9 audio_if.
While here, make debug output invisible #ifndef AUCCDEBUG.
1997-07-09 22:37:08 +00:00
pk
f3dc7ce0f4 Use setpgt4m(), not bzero(), to initialize a pmap's first-level page table.
Fix a slightly misleading comment.
1997-07-09 21:51:26 +00:00
matthias
bb66aaeadc * Make it possible to disable disconnect/reconnect on a per target base.
* Eliminate disable interrupts in ncr_pdma_in. No more fifo overflows at
  high baud rates :-) (Thank's to Ian Dall for this).
1997-07-09 19:28:25 +00:00
matthias
c5e3d310a6 When booting from a md device set RB_DFLTROOT in boothowto. 1997-07-09 19:28:20 +00:00
matthias
1c1450c97a Remove __VM_PMAP_HACK. 1997-07-09 19:27:54 +00:00
matthias
ed60e6a598 This wasn't in use for several years now. 1997-07-09 19:27:33 +00:00
matthias
147aa6816f Remove __VM_PMAP_HACK. 1997-07-09 19:26:34 +00:00
leo
1c21c4994d Make sure the card is probed once. 1997-07-09 14:38:02 +00:00
leo
89325630dd Introduce 'bootversion' in locore.s. Check this to decide if installing
a bootstrap with the running installboot makes sense.
1997-07-09 14:31:12 +00:00
thorpej
b4ab2c89f5 Oops, fix a slight oversight, and remove T_MACHCHECK handling (some
code for this was in my private tree at one time, and I must have
forgotten to g/c the constant from trap.h).  Reported by
enami tsugutomo <enami@but-b.or.jp>, PR #3845.
1997-07-09 07:01:29 +00:00
pk
94289e9990 In the turbosparc's "match" function, test for the `master' CPU, in case
someone ever builds a multi-processor turbosparc module.
1997-07-08 22:14:42 +00:00
pk
8eae6eb3b7 Fix a pasto in kvm_uncache() (sun4m)
Add TLB flushes when removing segments/regions.
Flush cache appropriately when cache-tag protection bits need invalidation.
1997-07-08 20:02:47 +00:00
pk
27d22f8faf Add FPU version 5 (turbosparc on-chip).
Re-arrange cpuattach() a bit so that more initialization is done before
dealing with the FPU.  Some more work is still required to de-couple
FPU detection from the autoconfiguration attach code in SMP systems.
1997-07-08 19:58:14 +00:00
mark
ec553ec8c0 New version of the ARM FPE core built. This core does not have all the
builtin veneer code. Instead this veneer code has been moved to kernel.
This removes the need for any runtime fixups of the fpe.
1997-07-08 19:41:28 +00:00
thorpej
4e8a131f3a When allocating space used by physical memory from the iomem extent,
use a value derived from biosbasemem, rather than a constant.  Pointed
out by Chris Demetriou <cgd@netbsd.org>
1997-07-08 17:55:44 +00:00
kleink
96914606be From sun3:
Correct handling for Trap #2 in SunOS executables,
 now that we know it is supposed to flush the cache.
 (Was thought to be "some obscure FPU operation".)
1997-07-08 16:56:31 +00:00
pk
99edebd566 fbrcons_{rows,cols} helper functions. 1997-07-07 23:31:30 +00:00
pk
af86c62d61 Get console tty size from raster console if applicable. 1997-07-07 23:30:22 +00:00
pk
7a959f1752 Add `maxusers' line and set it to 32. 1997-07-07 21:54:51 +00:00
pk
42f6c894f8 Raise max maxusers to 1024 1997-07-07 21:53:43 +00:00
pk
0360275f33 Clear relevant `cpuinfo' fields upon detection of a cloaked TurboSparc. 1997-07-07 20:43:25 +00:00
cgd
fce829ca4b mark prototypes for static inline functions as possibly unused
(with __attribute__ ((unused))), to avoid generating warnings when
compiling without optimization but with the default warning flags.
1997-07-07 20:37:34 +00:00
pk
2d50be8cb3 Add `turbosparc_getfaultstatus()' to be used instead of the generic srmmu item. 1997-07-07 20:06:42 +00:00
jonathan
1490cbcf7b Rewrite struct ecoff_symhdr using the same field ordering as GNU
binutils and the MipsCo toolchain, not the Alpha ordering (which has a
block of int32_t symbol counts and a block of long offsets) .
1997-07-07 19:37:33 +00:00
pk
3c1b9ba29b Install media functions only if something useful can be returned. 1997-07-07 16:28:44 +00:00
hannken
bcb9e18784 Fix bad parentheses in `getcpuinfo()'. 1997-07-07 08:44:01 +00:00
jonathan
65e2c70353 Force write-back of D-cache after doing DDB writes on mips3. Flushing
the Icache is not sufficient: a mips3 can write a new insn into
writeback L1 Dcache, leaving stale instructions in the mixed L2 cache.
1997-07-07 04:55:27 +00:00
jonathan
919bc0ce92 Typo in RCS id. 1997-07-07 03:57:55 +00:00
jonathan
d1ec048977 DDB for mips.
Add DDB interface to /sys/arch/mips/mips..
  Rework heuristic stack traceback to work with DDB.
  Add hooks  to print exception log from DDB.
  Add hooks from pmax console drivers:   call Debugger()
  after break from serial console, or 'DO' key from LK-xxx.
1997-07-07 03:54:24 +00:00
pk
27434aef27 Follow lock changes (DEBUG => LOCKDEBUG). 1997-07-06 23:57:16 +00:00
pk
4846e5ba47 Adjust for generic lock changes. 1997-07-06 23:52:52 +00:00
is
487476c2a9 Implement the RTC_OFFSET kernel option/rtc_offset variable for the Amiga port.
To get the old behaviour, set options RTC_OFFSET=0 in your kernel configuration
file.
1997-07-06 23:17:53 +00:00
ragge
759763742b Remove __VM_PMAP_HACK. 1997-07-06 22:38:22 +00:00
is
5e6f97d4ef A first hack at DraCo real-time clock support; only reads clock, currently.
Also only whole seconds are used (while the hardware provides 1/256 seconds).
1997-07-06 22:27:19 +00:00
pk
f02db5d4f6 Turn off microsparc II compatibility bit in Fujitsu turbosparc processors. 1997-07-06 22:23:37 +00:00
pk
0ab67aca5d Define the turbosparc's processor configuration register. 1997-07-06 22:21:11 +00:00
pk
c4e35e2f8b Have a patchable variable `ignore_bogus_traps' (0 by default) that
controls whether or we panic when receiving an unhandled trap.
1997-07-06 21:34:45 +00:00
pk
bbb57cadb4 Use the module info fields cputype' and vactype' to hold a processor
family's type and cache information. This can be optionally modified
by the associated `match' routine.
1997-07-06 21:18:27 +00:00
pk
55a6f97322 Add VAC_UNKNOWN to `enum vactype'. 1997-07-06 21:15:14 +00:00
pk
81469e565e Add two more fields (cputype' and vactype') to the module info. 1997-07-06 21:14:25 +00:00
christos
3de323b091 Fix printf formats. 1997-07-06 19:32:39 +00:00
pk
48ad000318 Slight optimization when pv_unlink'ing last entry (suggested by Eric Fair).
Use MR4() macro where appropriate.
1997-07-06 12:29:54 +00:00
pk
c3af302842 Use setpgt4m() in setptesw4m(). 1997-07-06 12:22:39 +00:00
thorpej
3a3d0fcb0f Add new DDB and KGDB files. 1997-07-06 04:09:45 +00:00
thorpej
55fdd1f6be Add a new line in a printf, for purely cosmetic reasons. 1997-07-05 20:58:30 +00:00
leo
cd3ee4b310 If booted with RB_ASKNAME, don't go looking for a root device. 1997-07-05 20:57:40 +00:00
leo
a323500b0f Pull in latest Amiga bus/address error handling. 1997-07-05 20:50:41 +00:00
thorpej
15de91e26d Add a breakpoint() inline, used by DDB and KGDB. 1997-07-05 20:49:46 +00:00
thorpej
bbf08442c7 Define the ALTENTRY() macro here. 1997-07-05 20:49:19 +00:00
thorpej
bc831d9a96 Add glue for remote KGDB. 1997-07-05 20:48:12 +00:00
thorpej
9ae7db0945 Garbage-collect old KGDB glue. 1997-07-05 20:47:35 +00:00
thorpej
46222f0d29 Split the memory access functions out of db_interface.c, so that they can
be shared by DDB and KGDB.
1997-07-05 20:46:37 +00:00
thorpej
f796580395 Machine-dependent portions of remote serial KGDB, for NetBSD/i386. Based
on Matthias Pfaller's NetBSD/pc532 kgdb_machdep.c.
1997-07-05 20:44:57 +00:00
is
5171612d6c Multichannel output still doesn't work, but while working on it, I repaired
the double buffering code, and setting a couple of other parameters works now.
1997-07-04 21:00:15 +00:00
is
e03fce55ed addql does set condition codes, so we are not allowed to put it in between
the andl and the jeq.
Mea maxima culpa.
1997-07-04 20:52:50 +00:00
ragge
59501e1879 Fix bug causing all uda's to be recognized as ctlr 1 from boot. 1997-07-04 13:26:02 +00:00
ragge
d7d40e44f6 Bugfix to mscp tape code; so that weof/eot works correctly. 1997-07-04 11:58:20 +00:00
drochner
41c3159f16 Emulate old bootblocks better: Use type of boot harddisk, extracted
from disklabel, as default.
(Pushed by Darren Reed <darrenr@cyber.com.au>:-)
1997-07-04 10:52:44 +00:00
scottr
9fc7759ee2 Optimize the bus error handlers a bit; save a couple cycles by dropping
through a branch rather than taking it in the case of an MMU fault,
which also has the desirable side effect of possibly avoiding an
I-cache miss.
1997-07-04 04:49:15 +00:00
pk
aff11374d7 Defer setting the CACHEPAGETABLES flag until cache_enable(). 1997-07-02 14:52:59 +00:00
pk
94113eb121 Remove a bogon. 1997-07-02 14:39:24 +00:00
jonathan
eca6630217 Fix NSCC tests. Use NSCC > 0 (needs-flag) rather than NSCC > 1 (needs-count).
Fix prototypes previously hidden by same.
1997-07-02 13:59:15 +00:00
drochner
0b58e59709 Correct the error message printing for failed boot attempts
("filename" must be processed by parsebootfile() first).
Being here:
- remove the special "#ifdef MATTHIAS" mainloop
- make some "booting..." output match reality
- print '\n' and '\b' more controlled
1997-07-02 13:20:36 +00:00
mikel
7b93ae0fcb if exec_netbsd() fails, indicate what file we were trying to boot.
suggested by Havard Eidnes in PR port-i386/3771.
1997-07-02 04:07:43 +00:00
jeremy
bbf76eb63e Implement pmap and va caching within C table managers to help speed
up PV->(pmap,va) operations.
1997-07-02 03:23:57 +00:00
cjs
3b3cb5bcff Update comments to show that ed driver on PCI works with 3c90x cards. 1997-07-02 02:44:09 +00:00
scottr
d00173ae5a Add support for Portrait Video Card. From SUNAGAWA Keiki
<kei_sun@ba2.so-net.or.jp> in PR 3821.
1997-07-01 19:04:18 +00:00
jonathan
da53d70f23 Move generic mips functions setregs(), sendsig(), sys_signal()
to sys/arch/mips/mips/mips_machdep.c.  Delete from pica, pmax machdep.c.

Delint pica machdep.c.
1997-07-01 09:32:13 +00:00
jonathan
dfa33cb24e Remove "untested" comments and messages. 1997-07-01 07:21:14 +00:00
jonathan
e0bb7d9370 Establish handler for vertical-retrace interrupts.
Use establish-time argument in handlers, to work with multiple framebuffers.
1997-06-30 22:08:58 +00:00
jonathan
8586e62e14 Enable stack tracebacks if MDB is configured. 1997-06-30 14:42:32 +00:00
scottr
f2bdde942a Due to reliability problems on some models, back out the more radical of
the pc532/dev/ncr.c-inspired changes; in particular, wait for previous
transfer to complete before starting another, as we used to do.  Retain
splbio() protection and simplified sbc_pdma_in().
1997-06-30 05:24:35 +00:00
ragge
9f009387fc Must grok includes. Also; prettier compile-time printout. 1997-06-29 21:30:09 +00:00
leo
d4c5e4e6fe Add code to set pitch and duration of the keyboard bell. (Thomas Gerner). 1997-06-29 20:30:49 +00:00
scottr
ea99cc2796 Fix tyop in last; will link 040-enabled kernels now. 1997-06-29 19:54:57 +00:00
scottr
4428f246a1 Split up CPU-dependent bus error/address error handlers. Mostly from
the Amiga port, but with an extension to save the actual fault
address (in m68k_fault_addr) when we catch a bus error.
1997-06-29 19:02:07 +00:00
scottr
6eb0e7c503 Prototype DCIS(). 1997-06-29 18:56:11 +00:00
pk
363c28ce88 Fix typo in last commit. 1997-06-29 18:44:21 +00:00
is
86f86c044c Use function types for vectab, instead of u_int8_t *. 1997-06-29 13:49:45 +00:00
scottr
0b7e64bbfa Update for mac68k_buserr_addr -> m68k_fault_addr change. 1997-06-29 06:10:35 +00:00
scottr
7044c8fd8a Get several cleanup chores out of the way. The code is functionally
identical to the previous incarnation.

 - Update using m68k asm.h macros
 - Move initialization towards the front of the file
 - Rename mac68k_buserr_addr to m68k_fault_addr
 - Reorganize trap 15 handler, similar in structure to -- though not as
   complete as -- the hp300 version
 - Reorganize doboot() for easier integration of external cache, and
   make room for the latter (#ifdef __notyet__)
 - General garbage collection of unused code/data
1997-06-29 06:07:39 +00:00
scottr
a3bfbcf80a Remove unused (#ifdef'ed) vector 0 initalizer 1997-06-29 05:52:11 +00:00
scottr
be164c61ed Garbage collect bad{,b,w,l}addr() 1997-06-29 00:26:12 +00:00
scottr
494cda5d27 Use <m68k/include/asm.h>. 1997-06-28 23:35:02 +00:00
leo
1f92b460a0 Add 'libsa/atari_stand.h' that defines atari specific things. This allows
for a removal of our own copy of stand.h which was out-of-date by definition.
1997-06-28 21:36:28 +00:00
pk
adc8e8d8b0 Accommodate non-VM_PMAP_HACK code. 1997-06-28 19:59:03 +00:00
pk
a9b31740a2 Remove __VM_PMAP_HACK. 1997-06-28 19:56:12 +00:00
thorpej
b1d65b1d3f Adjust start address here, now that exec() no longer does it for us. 1997-06-28 07:20:25 +00:00
mhitch
d6b6efec34 Moved the mini-debug routines out of trap.c into their own file, like the
original pica port.
1997-06-28 03:59:46 +00:00
mhitch
a503f4436c Mini-debuuger is now included by options MDB.
Move mini-debugger routines to separate file, minidebug.c.
1997-06-28 03:57:55 +00:00
mhitch
566b174c13 Mini-debugger now included by options MDB.
Cpu_regs() is included by options DEBUG, as are the stacktrace routines,
so move it inside the #ifdef DEBUG along with stacktrace().
1997-06-28 03:55:05 +00:00
mhitch
8c12914cdb Fix typo.
Include minidebug.c with options MDB.
1997-06-28 03:43:21 +00:00
is
de025de65f Add support for T-racing of emulated instructions. (Yes, gdb uses this.) 1997-06-27 23:32:09 +00:00
jeremy
63c23f2337 Add reference to DMA selection variable used in dev/ic/ncr53c9x.c and sample
code to enable it.  Currently commented out.
1997-06-27 02:07:32 +00:00
is
f3efbb406a The recent buserr handler cleanup in the Amiga locore.s made the label
_buserr point to the 68020/030 buserr code _only_. This has broken access
error handling in the 060 support code.
This is repaired by jumping to _buserr60 from the 060SP, and by providing
a _buserr60 label identical to the _buserr in the unchanged m68k ports
using the 68060.
1997-06-26 22:28:42 +00:00
scottr
b1f4ee5c2a Identify and use the Apple Twisted Pair NB card properly, from PR 3789
submitted by Denny Gentry <denny1@home.com>; investigation on the hardware
done by Bob Nestor.
1997-06-26 21:08:10 +00:00
drochner
7f15295bde Define _STANDALONE for standalone programs. 1997-06-26 19:03:32 +00:00
perry
73de3e34b8 "fake" disklabels should make the whole thing the 'c' partition, not
the 'a' partition. Sanity checked by thorpej.

This, incidently, may have been causing user errors in initializing
new disks, as described in (now closed) pr-2729 from Scott Reynolds,
because users typically don't try to edit their c partitions to be
"correct".
1997-06-26 05:40:30 +00:00
jeremy
335b7a2d00 Duplicate changes recently made to the SPARC version of this file:
Do not print warnings if selection of a target via DMA was interrupted
by another target.
1997-06-26 02:47:09 +00:00
thorpej
0404c01c46 #define DB_AOUT_SYMBOLS 1997-06-26 01:26:56 +00:00