67fc8c19be
if MULTIPROCESSOR, make the registers virtual, so that we can access other
...
processors registers. rework enter/exit ddb code to look more like the i386
code. make "mach cpu" and "mach cpu N" work.
2001-06-08 09:49:28 +00:00
7666800896
need a per-cpu ddb_regs pointer.
2001-06-08 09:45:50 +00:00
1ae283fcdd
use raise_ipi_wait_and_unlock().
2001-06-08 09:40:31 +00:00
e3b20fd18d
options PROG32 for now.
2001-06-08 06:16:09 +00:00
85656841a3
Change _ARM32* to _NETWINDER* Include <machine/psl.h> to get psl
...
definitions.
2001-06-08 06:14:25 +00:00
1f4156858b
add an include subdir rule in the Makefile to make netwinder use the
...
"new" include file structure to be used on arm ports.
2001-06-08 06:13:02 +00:00
7c77092fcf
Change options<tab> to options<space>
2001-06-08 06:10:40 +00:00
68e8e6ff9c
make this compile again.
2001-06-08 05:43:00 +00:00
7bb4cfc2db
split {sunos,svr4}_sigcode out for LKMs
2001-06-08 04:49:45 +00:00
e5bd00e48d
For ports that wire up pciide in compatibility mode, have
...
them define __HAVE_PCIIDE_MACHDEP_COMPAT_INTR_ESTABLISH
in pci_machdep.h and pciide_map_compat_intr() only calls
pciide_machdep_compat_intr_establish() if that preprocessor
define exists.
Ports that don't need to do this no longer need to supply a
dummy function.
2001-06-08 04:48:54 +00:00
9a3de71872
add CPUINFO_FLAGS
2001-06-08 04:21:41 +00:00
9634e929d1
Oops, missed in the previous commit: bump PIL_SER to 13 as well.
...
It seems to be unused...
2001-06-08 01:33:32 +00:00
d2f0c68d47
Make macppc compile by default with -Wall -Wmissing-prototype in addition
...
to its currently enabled warnings. For the lack of a better place, most
global function prototype wound up in <machine/autoconf.h> unless there
was a better place for them. ofb_* structs renamed to offb_* to avoid
conflict with ofb_softc in <dev/ofw/openfirm.h>
2001-06-08 00:32:01 +00:00
f6d422beb8
Fix warning about braces & if. Include libkern.h for strncpy prototype
2001-06-08 00:21:45 +00:00
75a27eccff
Rename pte_spill to pmap_pte_spill. Fix pmap_clear_{referenced,modify}
...
to return the previous state of the bit. Make it compile under
-Wmissing-prototypes -Wall. Change around some debug stuff.
2001-06-08 00:16:24 +00:00
a8bb9b127a
All soft interrupts are currently handled by soft interrupt 1, so make
...
all splsoftxxx() calls block soft interrupt 1 and make spllowersoftclock()
re-enable soft interrupt 1.
XXX: This needs to be reworked so that spllowersoftclock() only turns on
softclock interrupts, not all soft interrupt. That change is coming soon.
2001-06-08 00:09:28 +00:00
86d8f950ef
Fix initial interrupt mask settings for `hard' interupts to mask out soft
...
interrupts as well. The system is quite stable now that soft interrupts
don't pop up at random places where they're unexpected 8-)
Also, while here, fix CPU clock speed calculation for the Indy.
2001-06-08 00:02:41 +00:00
0859a7601d
- Differentiate beteen the 8003 and 80c03 (the 80c03 has some additional
...
features).
- Implement crude multicast support -- enable reception of all multicast
frames if we have any addresses on the multicast list. TODO: use the
64-bit hash table on the 80c03.
- Implement promiscuous mode.
IPv6 doesn't quite work yet -- DAD finds a duplicate for the
link-local. The 8003 hears its own transmissions, but IFF_SIMPLEX
is not set, so I'm a little confused.
2001-06-07 23:05:51 +00:00
4e6f32c9d7
Fix the broken stray irq handling for footbridge, this needs to be
...
propegated to other arm32 irq handlers, until there's a common core part
for doing irq handling.
Untested as I've no cards that generate stray irq's.
2001-06-07 21:47:36 +00:00
75d8b9d463
Checkin fix to ARM7TDMI support. As per patch from John Fremlin to
...
port-arm32.
2001-06-07 21:07:22 +00:00
b4e0748bf3
Don't use PCLK as the baud rate generator, as the Indy doens't
...
like that, and it has no effect on the Indigo2.
From Christopher SEKIYA <wileyc@rezrov.net>.
Also, query the "ConsoleOut" ARCS variable (i.e. "what the console
really is") rather than "console" ("what the console is configured
as").
2001-06-07 19:23:03 +00:00
d3b1cc167d
more SMP work:
...
make IPI's work. modify boot_secondary_processors() to clear the startup flag
in each cpu. new raise_ipi_wait_and_unlock() that calls raise_ipi(), waits
for the cpu to acknowledge it got the message, and then unlocks the msglock.
use the new framework in mp_{pause,resume}_cpus(). nmi_soft() takes a
`struct trapframe *', to be used by ddb.
2001-06-07 17:59:47 +00:00
ba991d978b
move ev_out32() here.
2001-06-07 17:54:35 +00:00
51b7188429
make lkm friendly.
2001-06-07 17:49:51 +00:00
5d512c674c
GENERIC + memory disk hooks.
2001-06-07 17:08:13 +00:00
af0929e6ce
Enable a few more devices (adw, dpt, siop, sip, sd). Do not compile DEBUG.
2001-06-07 17:05:11 +00:00
6246dceb77
print p_pid & p_comm for alignment errors.
2001-06-07 15:31:16 +00:00
a7df60e8e9
Pass -N to the linker, so that it will generate a single PT_LOAD
...
section (a'la OMAGIC).
2001-06-07 15:26:31 +00:00
a1830df1bb
Fix a deadlock in the tx code -- the code checked the DMA status before
...
sync'ing the descriptor, leading to a race where the packet could complete
after the DMA check but before the sync of the descriptor, sending the code
into unknown territory and the DMA engine into space. Also, fix a bug with
spurious watchdog resets going off due to the if_timer not always being set
on transmit.
Finally, clean up use of hardcoded magic numbers, get them from hpcreg.h
when possible.
2001-06-07 12:20:42 +00:00
63f856cb2e
Add a bunch of defines for bits withing the ethernet-related registers so
...
the Seeq driver can use them instead of hardcoded constants.
2001-06-07 12:10:34 +00:00
3aa99cbeac
We really need a wsdisplay on the Milan.
2001-06-07 08:41:11 +00:00
e371e8a5ae
Try to search for an isa bus during console configuration too... Otherwise,
...
we miss the console keyboard on the Milan.
2001-06-07 08:35:29 +00:00
3700ec551e
The newly generated config files.
2001-06-07 08:01:19 +00:00
d355dc319f
Oops, we don't want 'MILAN' anymore....
2001-06-07 08:00:24 +00:00
3bc0965e01
Everything you always wanted to know about Atari,Milan,IDE and byteswaps...
...
The Milan configuration is split up in 2 parts:
- MILAN_ISAIDE
Supports the traditionally formatted IDE disks on the Milan
- MILAN_PCIIDE
Supports the byte swapped formatted IDE disks on the Milan
If you want to read more, checkout the comment in MILAN.in...
2001-06-07 07:56:57 +00:00
92a2e67c31
Remove the atari specific incarnation of wdc_isa.c. We don't need it (in
...
fact it didn't work that well either), all could be done by simply adding the
NOSTREAM options to the flags of the generic ISA frontend.
There was probably to much byteswapping in my brain when I invented this.
2001-06-07 07:23:02 +00:00
762bab2606
implement pmap_clear_modify() correctly.
2001-06-07 05:29:13 +00:00
178aeb74a0
work around lint issue (inspired by similar work in sparc port)
2001-06-07 04:44:05 +00:00
5429ee5e02
add iha(4); it works in my pc164 after thorpej fixed a minor issue
2001-06-07 03:51:53 +00:00
d69d2974a4
Sprinkle some volatile keywords around inline asm so that the optimizer
...
doesn't reorder critical section codes.
2001-06-07 02:38:59 +00:00
9b27fa3ff8
set the coredump32 hook. make LKM friendly.
2001-06-06 21:39:50 +00:00
059d4936ab
extract the netbsd32, sunos, 32 bit svr4 and 64 bit svr4 sigcode out into
...
separate files, so that compat lkms can link with them as well.
2001-06-06 21:19:48 +00:00
ea39a5c3b4
make a NETBSD32 stack check LKM friendly.
2001-06-06 21:06:52 +00:00
695e5e7ab5
Changes new pmap, common param.h, vmparam.h, and moved includes.
2001-06-06 17:50:14 +00:00
f294982771
Enable tlp for 21x4x chips.
2001-06-06 17:47:54 +00:00
daf8279e4b
Change bebox port to use the include MPC6XX files and the new common bus.h
...
framework.
2001-06-06 17:42:29 +00:00
e38a49d835
Include a common bus.h and bus_dma.c that powerpc ports can use
...
if they so choose.
2001-06-06 17:37:37 +00:00
938edd5b75
Introduce a new & faster pmap for the MPC6xx (60x, 7xx, 7xxx) PPC CPUs.
...
Move MPC6xx dependent header files to powerpc/include/mpc6xx/
2001-06-06 17:36:01 +00:00
92359cee67
Bump splserial to 13 to support com port on Tadpole and JavaStations.
...
While I'm here - remove redefinition of PIL_TTY, move splnet
definition to keep the list sorted and sync its comment with reality.
2001-06-06 13:45:18 +00:00
ec218acedd
make this compile without MULTIPROCESSOR.
2001-06-06 06:29:36 +00:00
8688ff47c8
make this compile without DEBUG.
2001-06-06 06:23:13 +00:00
2f94a06e98
HD64461 frame buffer is big-endian. set HPCFB_SWAP_BYTE.
2001-06-05 17:24:48 +00:00
5d3a461f45
use _KERNEL_OPT
2001-06-05 14:43:04 +00:00
32874c14c8
Detect MicroVAX 3100/m80 correctly. From Michael Kukat.
2001-06-05 11:25:11 +00:00
95fe4db7e3
Replace arm/arm32/db_trace.c and arm26/arm26/db_trace.c with a unified version
...
in arm/arm. This version is based on the arm26 version, and includes dumping
the contents of stack frames, with automatic determination of the save code
pointer offset.
2001-06-05 09:25:05 +00:00
261bd8f8ac
Add get_pc_str_offset(), which returns the offset between the address of an
...
instruction that stores the program counter and the value of PC that's stored.
This can vary between ARM implementations, but is guaranteed to be constant on
a given one.
2001-06-05 09:19:32 +00:00
eadf88cc61
(locally defined) HZ -> hz
2001-06-05 05:20:20 +00:00
5523a584c5
HZ -> hz
2001-06-05 05:12:18 +00:00
ec291b503a
Allow the text address to be set with makeoptions.
2001-06-05 05:05:46 +00:00
95eaadac34
Use symbolic names for the CTB terminal type field, rather than
...
hard-coded constants.
2001-06-05 04:53:11 +00:00
f81f19e7b3
Add rudimentary multiprocessor support for DDB.
2001-06-04 21:37:11 +00:00
46762d3560
fix lint exposed by GCC 3.0 20010604 (prerelease).
2001-06-04 20:56:51 +00:00
842b6023c3
HD64461 video module.
2001-06-04 17:08:36 +00:00
93ece82137
Check for console printf's in spinlocks, per discussion on tech-smp.
2001-06-04 15:37:05 +00:00
30e020fbaa
The beginning of pmap locks. While here, some cleaning and KNF.
2001-06-04 15:36:00 +00:00
81d993ac3b
Add splipi().
2001-06-04 15:34:57 +00:00
da2e5c43e1
Add IPI_TBIA to flush the translation buffer.
2001-06-04 15:34:15 +00:00
d35f5e00dc
Move some locks closer to the important point.
2001-06-04 15:33:07 +00:00
878db7cfb8
Add support for ARM7TDMI, as provided in a patch from John Fremlin to port-arm32.
...
Shouldn't effect any currently in tree ports.
2001-06-03 18:32:33 +00:00
e2f864bf95
Break out the MP-dependent calls. Add definitions for the IPI functions.
2001-06-03 15:12:57 +00:00
d0a8785b5e
Add define for IPI vector.
2001-06-03 15:10:34 +00:00
8d7b76581f
Implement spinlocks as subroutines instead of inlines.
...
Add SPINLOCK_SPIN_HOOK; the VAX has low-priority IPIs like Alpha.
2001-06-03 15:10:11 +00:00
6e219aa5a7
Add bbssi/bbcci (bit set/clear and branch interlocked) as inline functions.
2001-06-03 15:08:32 +00:00
1da9a91480
A bunch of fixes:
...
- Make generic console routines not relying on running on master cpu.
- Add routine to start console transmitter (after IPI).
- Use real IPIs instead of the "console doorbell".
- Add routines cpu_send_ipi()/cpu_handle_ipi().
2001-06-03 15:07:20 +00:00
ad811093a3
Add iha at pci .
2001-06-03 13:46:57 +00:00
e1fdb3abbf
Rather than duplicating the LDM/STM/LDC/STC fixup code between
...
early_abort_fixup() and late_abort_fixup(), have the latter tail-call the
former. This saves another 200 bytes, and I've found my ARM710a card now, so
I've even tested it.
2001-06-03 13:38:14 +00:00
c9d2bd8419
Sort some entries.
2001-06-03 13:21:28 +00:00
28bd948dba
finish moving context management back to being global. add a new ctx_lock
...
for context administration.
2001-06-03 04:03:28 +00:00
88905c8ca1
clear and restore pcb_onfault around calling uvm_fault(),
...
so that bugs in the fault-handling code will panic sooner.
2001-06-03 03:12:31 +00:00
f0f999ee47
Skip the pseudo-header checksum if nxt == 0.
2001-06-03 01:37:28 +00:00
6a32761143
Now that cpu_dataabt_fixup() can fail, print slightly more useful information
...
(address and disassembly of the instruction that aborted) when it does so.
2001-06-02 22:48:40 +00:00
3c88c46567
In {early,late}_abort_fixup(), return ABORT_FIXUP_FAILED, rather than
...
panic'ing, if we hit an instruction we can't fix up. This saves 250-odd bytes
of code, and should allow the caller to print a more useful message.
2001-06-02 22:30:07 +00:00
f594a52dae
late_abort_fixup: Get the indentation right!
2001-06-02 22:08:11 +00:00
0c870e10e0
Convert bus_space_{read,write}_multi_1() into macros, since they're just
...
wrappers round {read,write}_multi_1() anyway.
2001-06-02 21:31:02 +00:00
54a986d7fc
Replace arm6_dataabt_fixup() and arm7_dataabt_fixup() with early_abort_fixup()
...
and late_abort_fixup(), based on the abort model in use, rather than the CPU
type. This cleans up the code and makes it smaller. Only tested on an
ARM6 -- I can't find my ARM710a card right now.
2001-06-02 21:03:32 +00:00
7089595665
Initial cpufunc operations for ARM3. Not actually used yet.
2001-06-02 19:01:03 +00:00
821ec03ed9
replace vm_map{,_entry}_t with struct vm_map{,_entry} *.
2001-06-02 18:09:08 +00:00
0319a3d7bf
Add a missing line from i386/i386/pmap.c to make this work again.
2001-06-02 16:47:17 +00:00
4db9600c94
use hd0a instead sd0a in 'help' command example
2001-06-02 14:56:13 +00:00
af667cdd42
The io offsets in arm32/isa/timerreg.h included the IO_TIMER1 offset.
...
With dev/ic/i8253reg.h this is not the case.
Add IO_TIMER1 to all offsets used in inb and outb. This fixes port-arm32/13009.
2001-06-02 12:51:28 +00:00
68035f55f1
In DEBUG kernels, catch page faults in interrupt handlers early.
2001-06-02 11:17:32 +00:00
c81d426dcf
Save R14 (in a proper stack frame) around {read,write}_multi_1, since page
...
faults can corrupt it on arm26.
2001-06-02 11:15:56 +00:00
447793851c
Add hcsc to every configuration that has SCSI podules.
2001-06-02 10:48:54 +00:00
972ef526d2
Add hcsc, scsibus and some SCSI devices.
...
Also, add some useful compiler options to FOURMEG.
2001-06-02 10:47:55 +00:00
2d35661e6a
Use write_munti_1() to implement bus_space_write_multi_1().
2001-06-02 10:45:43 +00:00
ac0d39dfe8
Add write_multi_1().
2001-06-02 10:44:56 +00:00
3df8e425e7
Correct permissions on vmapbuf'ed pages. A read from a device is a write to
...
memory, and vice versa.
2001-06-02 10:44:02 +00:00
40033835d8
Comment DIAGNOSTIC, DEBUG out
...
Use the 'root on ? type ?' config by default, leave the nfs one commented out,
add line for root on ed0a
2001-06-02 10:37:27 +00:00
9bde6cff5c
remove bogus curly braces from a structure forward-declaration.
2001-06-02 01:19:29 +00:00
f6af05f407
Back the last revision off, upon Bill Sommerfeld's request
2001-06-02 00:46:00 +00:00
5dbe622e77
When polling, raise priority level to tty, to block keyboard interrupts
...
when the system is "warm", i.e. interrupts are not blocked anymore.
This seems to be necessary on my PS/2 Model 70 keyboard - without this,
system ends up in endless loop calling the keyboard intr routine if a key
is pressed when polling. This _may_ be just specific to level-triggered
interrupts PS/2 MCA uses, though it's more likely it's just the way the
particular keyboard controller works.
Discussed on tech-kern@.
2001-06-02 00:01:04 +00:00
b001ce9b90
g/c COMPAT_OLDBOOT from bootblocks, sprinkle some const
...
bump bootblock version to 2.10
2001-06-01 23:26:30 +00:00
7e716e6849
Memory map and register definitions for the Algorithmics BONITO
...
MIPS memory and PCI controller. This file is provided by Algorithmics.
2001-06-01 20:29:33 +00:00
2b5df8b7ff
vm_page_t -> struct vm_page *
2001-06-01 19:52:54 +00:00
5fbaedc40d
sync with GENERIC 1.405
2001-06-01 16:50:31 +00:00
71cb790fb5
Add support for the Algorithmics P-4032 board. This is totally
...
untested, since I have no P-4032 board, but it's no worse than
the current situation, which is "totally non-working P-4032
support in the ARC port, of all places".
2001-06-01 16:00:03 +00:00
23a3dc1508
If a bus doens't want to use an extent, don't force it to.
2001-06-01 15:57:31 +00:00
356699e86b
Fat-trimming.
2001-06-01 15:30:11 +00:00
c702830c42
The P-4032 has no ISA bridge/bus, so remove all P-4032 conditionals.
2001-06-01 15:20:06 +00:00
07187f8a43
Nuke FRAMEBUF_HW_BASE. This value is provided by the bootloader.
2001-06-01 14:08:35 +00:00
6349d0ae81
Nuke FRAMEBUF_{HW_,}BASE. This hack does no good.
2001-06-01 14:06:43 +00:00
ea77ec5725
Move some low-level code (SVC mode switching and PID address translation
...
disabling) which was written using inlined asm in initarm() to just after
the kernel entry point.
2001-06-01 14:04:29 +00:00
78aad8f79e
Mark evcnt(9) item as done.
2001-06-01 08:31:07 +00:00
37326d639b
Rename assembly files that are pre-processed by cpp to use an S suffix.
2001-06-01 04:20:04 +00:00
9f159a8918
Remove 4096-byte gap between .reginfo and .data, suggested by
...
Ian Taylor <ian@zembu.com>.
2001-06-01 03:55:30 +00:00
f4f6c1dd1c
Enable SCSI.
2001-06-01 03:53:29 +00:00
76b3b49bb8
Add support for jornada720's framebuffer.
...
The functionality is almost same as bivideo, and hardware rasops and
power saving are not yet supported.
2001-06-01 02:51:03 +00:00
f57a0847d2
fix off by one error. extra one page is needed for the case where
...
buffer is not page aligned (e.g. raw i/o case).
2001-05-31 20:56:29 +00:00
9bac50748f
"opt_ddb.h" should be included at the beginning of source file,
...
because some headers (in this case <systm.h>) refers symbols (e.g. DDB)
defined in the opt_ddb.h
2001-05-31 19:41:57 +00:00
59ba4788ce
Deprecate intrcnt/intrnames in favour of the generic evcnt(9) interface.
2001-05-31 18:46:07 +00:00
1a891f65d4
One more gcc-2.95 issue.
2001-05-31 09:19:25 +00:00
283e406cb0
drop to prom on "halt" as well.
2001-05-31 08:55:19 +00:00
da8f8f1e72
define _KERNEL_OPT as well as _KERNEL. we will use this in the future to
...
get kernel "opt_foo.h" headers, rather than _KERNEL && !_LKM.
2001-05-31 07:37:01 +00:00
1f4f9ae5c2
Fix minor typos.
2001-05-31 07:24:23 +00:00
f998b62e15
Fix commnet; hpcmips isn't x86 architecture.
2001-05-31 02:30:04 +00:00
299159546d
s/Alpha/MIPS/ in comment.
2001-05-31 02:20:55 +00:00
c227148511
PRiD 0x18 is shared by RC32334, 332 and 355. These SoCs are
...
distinguished by SYSID register in the system controller. Note
that PRiD 0x20 is for a standalone RC32364 processor which has the
same 32300 core inside. Rather better to name them MIPS32 ISA.
2001-05-31 02:06:26 +00:00
ba87a6d444
Add gcc-2.95 version.
2001-05-30 20:37:48 +00:00
d84d2c6c85
add missing #include "opt_kgdb.h"
2001-05-30 15:24:23 +00:00
e2c5f5866f
Use sys/arch/arm/arm/blockio.S.
...
Actually, these functions aren't used by hpcarm but is necessary to be
able to link sys/arch/arm/mainbus thing.
2001-05-30 14:39:19 +00:00
49927dd845
Add itepoll() prototype as suggested by 'maximum entropy' (pr#13047).
2001-05-30 14:25:07 +00:00
2eae10adb5
Inline bswap{16,32} when the argument is constant.
2001-05-30 13:08:34 +00:00
72943f1165
Pasto.
2001-05-30 12:52:06 +00:00
67afbd6270
use _KERNEL_OPT
2001-05-30 11:57:16 +00:00
f32430d518
Add a case clause for IDT RC32332/RC32334 processor personality
...
inside a commented-out block.
2001-05-30 09:06:28 +00:00
16a60efd2c
Add PRiD 0x18 for IDT RC32332/RC32334 processors.
2001-05-30 07:21:51 +00:00
2d632efc89
Start using the blockio functions. This makes raw reads from hcsc 20% faster
...
(is that all?).
2001-05-30 00:19:43 +00:00
f15491e9d3
Move the declarations of blockio.S functions from katelib.h to a new, exciting
...
<arm/blockio.h>. katelib.h includes it for compatibility.
2001-05-30 00:16:36 +00:00
d51c27dfa8
Add read_multi_1() which does what it says fairly fast.
2001-05-30 00:14:09 +00:00
966a1b86f8
initialize the b_dep field in malloc()'d buffers. fixes PR 13065.
2001-05-29 23:08:06 +00:00
7d9a4dbad7
Move blockio.S from sys/arch/arm32/arm32 to sys/arch/arm/arm, since I want to
...
start using it on arm26.
hpcarm gets to keep its own version, since it's not identical and I haven't
got an hpcarm box to test.
2001-05-29 23:03:20 +00:00
d674a69f7a
And now we're running!
2001-05-29 21:55:00 +00:00
2de1ab0486
Add ci_cpuid.
...
Change the CLKF_ macros because the idle loop is now running at IPL1.
2001-05-29 21:29:32 +00:00
3b4d7a2948
Must lock kernel when softclock is called from outside hardclock.
2001-05-29 21:28:14 +00:00
1cf7a84a88
Add a bunch of locking code for MP systems.
2001-05-29 21:27:25 +00:00
c20a10d57d
Call proc_trampoline_mp() after fork in multiprocessor environment.
2001-05-29 21:26:44 +00:00
6492868388
Change Swtch so that the old process context is stored directly and the
...
idle loop is on the interrupt stack instead of the last running process'
stack. This is needed to work in a multiprocessor environment.
2001-05-29 21:25:11 +00:00
f0c1fb1bb2
Initialize DDB at boot time and break into it if the "d" argument
...
is specified to the kernel.
XXX PMON doen't load symbols for us -- need a dbsym(1) for ELF.
2001-05-29 18:40:25 +00:00
5331656107
The QED RM7000 can use the same idle routine as the QED RM52xx.
2001-05-29 18:19:20 +00:00
74fa4349ae
Install power-saving idle routines at the end of cpu_identify(). We
...
currently handle the QED RM52xx here.
2001-05-29 17:54:56 +00:00
mrg