345d748512
PDB_BOOT and PDB_BOOT1 should not be the same value.
2000-08-02 22:25:45 +00:00
7ec33e899f
Get kernel profiling to work. Now if only gprof worked....
2000-08-02 22:24:38 +00:00
dfe3516515
Include NONPLUS and just set the extras we need for 64-bit kernels.
2000-08-02 02:23:58 +00:00
a8cd48f9e6
`mem' is now a pointer, not an array. Fixes kernel coredumps.
2000-08-01 16:49:47 +00:00
57d2ec0b98
Overhaul cache flush code and coredump code.
2000-08-01 00:40:15 +00:00
568dca001c
Cleanup trapstate handling and garbage collect old stuff.
2000-08-01 00:36:05 +00:00
e5a87444c7
Cleanup trapframe and TSTATE handling.
2000-08-01 00:34:13 +00:00
6f9bfddd60
Add UVM_PAGEIDLE_ZERO.
2000-08-01 00:31:15 +00:00
cacbbd1117
Garbage collection.
2000-08-01 00:28:55 +00:00
d09c8ea0f1
Revamp cache flush routines.
2000-08-01 00:28:02 +00:00
07fc19421d
Pull in <machine/frame.h> to define BIAS and CC64FSZ.
2000-08-01 00:25:02 +00:00
aeb88321db
Call the parent bus dvmamap_unload function to clear the cache rather
...
than do it ourselves.
2000-08-01 00:22:41 +00:00
ede765839e
Look for fpu files in the sparc subdirectory where they now reside.
2000-07-31 22:17:05 +00:00
978acd5367
sync with GENERIC32; enable rnd
2000-07-31 13:21:45 +00:00
ec14782383
hardcode onboard "hme" to "hme0"
2000-07-31 13:16:24 +00:00
b5ce82b4a0
define BDPRINTF in the !DEBUG case to nothing. PR#10688.
2000-07-31 05:40:22 +00:00
e25636b5da
g/c RB_DFLTROOT
...
I've tried hard to find also various usage() messages and remove the
appropriate flag from there as well, hopefully all occurences are covered.
2000-07-29 20:06:27 +00:00
fe3f6176b4
Make 64-bit stack traces look prettier.
2000-07-28 19:10:33 +00:00
652a74ef51
Don't dump if there's no address space reserved for it.
2000-07-28 19:08:25 +00:00
256fc99a68
include the 32-bit kernel, not myself
2000-07-27 14:39:50 +00:00
8b5a1a1dea
do the 32 bit / 64 bit name swap, as per GENERIC.
2000-07-27 14:38:45 +00:00
cc41b6af8f
remove COMPAT_AOUT option.
2000-07-27 14:34:06 +00:00
a254dae933
ebus sc_is is a pointer now, already.
2000-07-27 14:17:10 +00:00
dea4c479f5
intr_list_handler() should only take one argument. it ignored the 2nd anyway.
2000-07-27 14:13:50 +00:00
556a915306
add commented LOCKDEBUG
2000-07-27 14:10:25 +00:00
2726e45533
fix warnings.
2000-07-27 13:59:26 +00:00
270fd5cc55
delete unused variable.
2000-07-27 13:58:54 +00:00
eafc0ade2d
fix warnings.
2000-07-27 13:57:23 +00:00
2079f86890
On-board devices on psycho machines seem to have the `interrupt' property
...
in the parent bus format (i.e. an INO) rather than being represented as
an PCI interrupt line. Provide a hack to work around this in pci_attach_hook().
2000-07-26 17:46:56 +00:00
e668bc0310
Use the mk48txx driver for time-of-day clock access.
2000-07-26 13:39:36 +00:00
785d8d91b9
Remove unused <sys/dev/ic/intersil7170.h>
...
Remove unused `extern struct idprom idprom'
Pull in <dev/clock_subr.h>
2000-07-26 12:39:20 +00:00
11142830f9
Switch back to newvers_stand.sh. Dunno why it bombed the first time.
2000-07-25 21:53:55 +00:00
714a495a4a
Duh, hp300 -> sparc64.
2000-07-25 20:48:19 +00:00
234b20f27c
Put back missing newvers.sh.
2000-07-25 20:47:44 +00:00
219910d2e6
add disabled LOCKDEBUG; re-enable all the phy drivers
2000-07-25 17:26:36 +00:00
320361c474
Re-instate a couple of delay slot NOPs marked `XXX spitfire bug?'
2000-07-25 15:15:32 +00:00
b77782d785
Optimization to pmap_zero_page(): we only need to use the first 16 FP regs,
...
as in bzero().
2000-07-24 16:21:02 +00:00
fe9795a1e5
Fix a whitespace glitch.
2000-07-24 16:08:18 +00:00
a0ae8e1ecb
Pull in the bzero() changes from libc for real, with a couple of kernel-
...
specific hacks:
* Make the bias handling depend on STKB, so it switches correctly for 32-bit
kernels.
* Use the 32-bit stw/ld for 32-bit kernels.
* Add a few `-STKB's that were missed in the previous change.
2000-07-24 15:57:07 +00:00
98127a7ee3
Cosmetic & readability:
...
- local label cleanup in interrupt handling code;
- in debugging sections when testing `intrdebug': use a unique
local label (97) to enhanced clarity;
- comment-out unused code;
- kill a bug as the satisfying result of doing all of the above.
2000-07-24 14:55:56 +00:00
1045aba51d
Define UVM_PAGE_IDLE_ZERO.
2000-07-24 09:14:56 +00:00
2933116823
Fix stack issues in bzero.
2000-07-24 07:40:40 +00:00
2f13e00500
Revert the previous change for now. I don't have time to debug it tonight.
2000-07-24 06:30:34 +00:00
ac182d9c19
Er, eliminate the -p link option as well, which just outputs huge amounts
...
of spooge.
2000-07-24 04:54:16 +00:00
711ffefeac
Nuke lderr.
2000-07-24 04:45:49 +00:00
0c2ad746ea
Slight rearrangement.
2000-07-24 04:39:01 +00:00
b65d3e93c9
Make this a bit safer against errors by using some .tmp files.
2000-07-24 02:00:39 +00:00
62fe935649
Give this a prayer of working with obj dirs.
2000-07-24 01:55:59 +00:00
8de9ab874b
Make this at least pretend to compile.
2000-07-24 01:51:44 +00:00
51b68e1fdd
Fix compiler warnings.
2000-07-24 01:49:37 +00:00
2f39009bda
Fix multiple problems with floating point and structure handling.
2000-07-23 21:36:56 +00:00
acbf46fbf4
Copy bzero change from libc.
2000-07-23 20:39:14 +00:00
999ba2b285
Give 64-bit processes access to a full 64-bit address space (90% of which
...
they can't use anyway, but that's another issue.)
2000-07-23 07:02:19 +00:00
526bc1de59
Include <machine/frame.h> to get the definition of BIAS and CC64FSZ.
2000-07-23 06:57:55 +00:00
46a7cab70a
make this usable for for userland asm, which needs CC64FSZ
2000-07-23 06:15:33 +00:00
01d7296690
switch to sys/conf/newvers_stand.sh for generation of version file
...
slighly adjust the Makefile WRT vers.c and SRCS & CLEANFILES
sync the message in boot/version with other ports using newvers_stand.sh
2000-07-21 22:06:11 +00:00
263f73d7f4
Now we're running 64-bit userland don't panic the system if a process steps
...
into the address space hole.
2000-07-20 23:29:50 +00:00
0a25b480c6
Clean up disassembly of sparc v9 insns and handling of negative values.
2000-07-20 20:43:21 +00:00
4badffe2c1
For now, alias `network' to `hme'.
2000-07-20 16:29:03 +00:00
71169064c4
Since `intrcnt' points to an array of long ints, reserve the appropriate
...
amount of space depending on _LP64. Provide macros to load/store signed
and unsigned long ints.
Also, make the `intrcnt' and `intrpending' arrays 16 entries long, as on
sparc/v9 interrupt #15 is just a regular interrupt.
2000-07-20 13:28:39 +00:00
6996dd8691
Fix merge problem.
2000-07-19 15:48:25 +00:00
25f8633edd
Add an option to poll the ROM console from hardclock you can break into DDB
...
even if callouts are disabled.
2000-07-19 15:41:53 +00:00
e3fc09812a
make myetheraddr() work even if the clock hasn't been setup yet
2000-07-19 10:20:09 +00:00
422e1b7236
Disable interrupt interlock using ih->ih_pending field. Seems to cause
...
more problems than it solves.
2000-07-19 04:36:42 +00:00
625dc5189f
Miscellaneous cleanup.
2000-07-19 03:24:07 +00:00
067b7521b5
Fix PIC bugs.
2000-07-18 22:38:11 +00:00
59a7148fe9
address casts: long -> u_long
2000-07-18 20:19:56 +00:00
db76bf893e
Cast pointer to `u_long' (not `long') for correct results.
2000-07-18 20:15:36 +00:00
602e908589
Correct args usage comment in probeget.
...
Whitespace cleanup.
2000-07-18 20:14:45 +00:00
1d14ce3842
probeget: 9 -> 0x9
2000-07-18 19:27:57 +00:00
c0e91fb184
If `ih_pending' is set simply return from the trap, don't schedule another
...
soft interrupt.
2000-07-18 19:25:26 +00:00
a49fddc9b3
openfirmware: don't lower %pil
2000-07-18 12:52:56 +00:00
34270e85bd
Use probeget() in pci_config_read() for the psycho, to avoid bus
...
faults when probing PCI space.
2000-07-18 11:37:31 +00:00
073e24cd1b
Set up register bank sharing bwteen psycho A & B based on matching
...
physical addresses. Allocate IOMMU state separately and refer to it
from both psycho_softc's.
2000-07-18 11:35:03 +00:00
faaa100d6e
#if 0 some dumpsys() debugging messages
2000-07-18 05:18:35 +00:00
d2626038cd
kill dead files.
2000-07-18 04:21:17 +00:00
040268c10e
s/auxreg/auxio/
2000-07-18 04:17:23 +00:00
1d1b0968f1
IF we're changing hz, change tick accordingly too.
2000-07-17 18:32:25 +00:00
ec726f07ca
Kill typos in interrupt code.
2000-07-17 17:06:00 +00:00
966115ff9e
probeget: do the same song and dance as ld*a() macros in ctlreg.h do.
2000-07-17 14:08:43 +00:00
eb5863f522
#ifdef BOOT_DEBUG --> BDPRINTF(PDB_BOOT, (...))
2000-07-15 14:26:39 +00:00
4eacbabe76
Make 64-bit stack tracebacks look decent.
2000-07-14 21:07:10 +00:00
6667752114
Move some macro definitions out of genassym and into locore.s
2000-07-14 21:04:16 +00:00
3143adecfb
Tell gcc that stores clobber "memory". Probably makes no difference since
...
they should not be used to access normal memory anyway.
2000-07-14 21:02:11 +00:00
41da794863
Make all the inline functions static.
2000-07-14 21:00:17 +00:00
7b6085ff22
Add support for 64-bit PIC and dynamic linking.
2000-07-14 20:59:07 +00:00
9de310aebe
Use per-PBM control/status register addresses provided by the PROM.
...
Print version/implementation info.
2000-07-14 15:13:35 +00:00
04fe9636ba
Provide details on the `psy_csr' register.
2000-07-14 15:09:51 +00:00
8258c0e780
Fix padding in struct pci_ctl.
2000-07-14 13:17:12 +00:00
e5a9cce870
Cast physmem to u_int64_t before applying ctob().
2000-07-14 13:15:44 +00:00
68edef1bf6
Back out previous for now; it requires a system wide prototype change.
2000-07-14 09:58:52 +00:00
c32ffb8447
Make `physmem' an `unsigned int', so it can properly represent large memory
...
configurations.
2000-07-14 00:02:35 +00:00
9da8631e2b
printf() format for 64-bit integers `tf_pc' and `tf_npc'.
2000-07-13 16:44:03 +00:00
c0f74e1bda
remove memreg.h.
2000-07-13 06:30:16 +00:00
22820972ac
remove vaddrs.h
2000-07-13 06:25:55 +00:00
81ec8bef47
remove old (bogus) comment.
2000-07-13 06:23:07 +00:00
d5a1bec37e
bump some system values.
2000-07-13 05:40:51 +00:00
7e8f67ad97
No need to fetch the psycho's own registers as thet are already passed
...
down from the parent bus. Also account for the difference in `reg'
properties in psycho vs sabre chips.
2000-07-12 21:49:44 +00:00
f56554a578
If there is no "interrupt-map" property on the ebus node, assume the
...
child "interrupt" properties are already in a format suitable for
the parent bus.
2000-07-12 21:07:36 +00:00
1f45545899
Add UltraSPARC III specific flag bit. Currently ignored.
2000-07-11 21:55:11 +00:00
f10d7699fc
Break into debugger by typing `+++++',
2000-07-10 20:24:23 +00:00
7403ef3d00
Need to include <sys/exec_elf.h> to make things happy. So much for adding
...
dead code.
2000-07-10 01:15:17 +00:00
048f1387db
Function signature for bus_intr_establish() changed.
2000-07-10 01:11:14 +00:00
1cfb5e6456
Grab and use the memory model info from the flags in the ELF header.
2000-07-09 22:39:17 +00:00
ec71379542
Use TSO memory model for 32-bit userland.
2000-07-09 22:02:32 +00:00
e5cd515b89
Upgrate interrupt handling:
...
1) use interrupt vectors for specific zs chips rather than polling
all of them.
2) use softintr_schedule() and schedule it for a particular device
rather than poll all devices.
2000-07-09 21:58:43 +00:00
f43ab74a43
Shuffle around some fields to waste less space on LP64.
2000-07-09 21:55:24 +00:00
406e0f779f
Add a `device class' interrupt level argument (from machine/intr.h)
...
to bus_interrupt_establish().
It's currently only used in sparc64/dev/psycho.c to assign a CPU interrupt
level to devices in PCI slots.
2000-07-09 20:57:41 +00:00
152f5c5f28
remove unused variable.
2000-07-09 13:35:54 +00:00
5a7cb20576
gdb requires a `struct fpstate' so provide one.
2000-07-08 19:25:14 +00:00
065b56c096
backout part of previous: turn off intrdebug, and another debug message
2000-07-08 05:04:01 +00:00
2c1dc518bd
Changes for readability:
...
- #define LP64 dependent STACKFRAME macro once
- use a LOAD_ASCIZ macro for setting up the format string
for debug messages.
2000-07-07 23:28:28 +00:00
468c727a28
Temporary fix to prevent the clock from stopping on UIIi machines.
2000-07-07 19:59:36 +00:00
14b1b05f10
Rename GENERIC to GENERIC32 and GENERIC64 to GENERIC to allow snapshots to
...
build properly.
2000-07-07 19:11:56 +00:00
2b69bc2115
BUFCACHE=20
2000-07-07 14:22:57 +00:00
b91804af6c
split up debug flags so i can turn off some of them
2000-07-07 13:10:34 +00:00
cacf690261
Try both the BOOTPARAM and DHCP protocol (in that order as is the custom
...
with Sun machines).
2000-07-07 12:59:16 +00:00
4b21f66d57
work around a bug passing arguments to iommu_dvma_load_raw()
2000-07-07 12:53:29 +00:00
c0f4ea2a12
check that this *entry* is BUSCLASS_NONE (not the current device) in
...
bus_compatible(). fixes boot device detection on sbus. while i'm here,
sprinkle DPRINTF()s in the boot device detection code.
2000-07-07 12:44:47 +00:00
9a6c45e283
options BUFCACHE=20, plus a 64 bit config.
2000-07-07 08:44:16 +00:00
3f2289a3d7
Handle bus_dma aligment properly.
2000-07-07 02:50:19 +00:00
6a1f10a7d5
Added two new options to the bootblock: -V turns on verbose mode and -D
...
makes it load but not execute. If fcode-debug? is enabled in the firmware
an arbitrary file can then be loaded with:
ok " <path>" do-boot
2000-07-06 17:15:07 +00:00
63d57aa16a
Fixup counter-timer interrupt lossage.
2000-07-06 14:13:05 +00:00
f6d8b8e963
We want to index the PCI `interrupt clear' registers with the Interrupt
...
Number Offset. (this stuff needs a serious cleanup).
2000-07-05 12:11:59 +00:00
502fc00210
Comment out EON ISO-over-ip tunneling pseudo-driver, and document as
...
broken/fragile. Unlikely to be of much use, and confuses new users
when their system crashes when they, or their dhclient stumble over
it. See kern/10500, kern/8994 for the gory details.
2000-07-05 04:07:22 +00:00
968524445a
Prevent interrupts from being posted if they are active and deprecate polled
...
interrupts. This seems to eliminate the spurious interrupts.
2000-07-03 17:56:06 +00:00
58c36586cd
Use the same IPLs as everyone else.
2000-07-03 17:42:37 +00:00
16fd30ec1d
- add a bus class check to the conversion of prom names to NetBSD driver
...
names, as we need to convert `pci' to both `psycho' and `simba'. add
a member to dev_compat_table[] for this, using BUSCLASS_NONE for when
we don't care (all other cases, currently). add `ide' -> `pciide' and
`disk' -> `wd' mappings here, also.
- "psycho" is a BUSCLASS_MAINBUS device; it has UPA attachment semantics.
this fixes bootpath detection for IDE disks.
2000-07-03 14:38:05 +00:00
2310f6d755
Add `ahc' (Adaptec 2940, et.al. SCSI controller); it is known to work.
2000-07-02 21:11:26 +00:00
f504adf782
Clean up interrupt handling some more.
2000-07-02 16:13:21 +00:00
f3e6e48cb8
update to reality:
...
- TRAPWIN is defined in locore.s
- don't config netbsd_wd0
- enable MSDOSFS and FFS_IE
- enable INET6 & IPSEC*
- enable ipfilter
- comment sbus & zs at sbus
- enable scsi devices seeing we have scsi bus support
- hardwire the onboard hme as 'hme0'
- disable audiocs for now
- disable rnd for now
2000-07-02 15:02:29 +00:00
585f7645bf
iommu_dvmamap_load*(), pass an alignment constraint, as well as
...
EX_BOUNDZERO, to extent_alloc(). this fixes problems with it
(extent_alloc()) returning incorrect values, when the start of
the extent is not aligned as strongly as the requested area..
raw access to ide disks now works.
2000-07-02 14:00:38 +00:00
15c79c03c2
use bitmask_snprintf().
2000-07-02 13:35:35 +00:00
9ec4bcf95b
whoops; turn off trapdebug again.
2000-07-02 12:39:01 +00:00
a5c13f9ad4
Kwality control:
...
* put #includes of opt headers and headers to get protos used by
net/netisr_dispatch.h in net/netisr.h (if !defined(_LOCORE)) (rather than
in netisr_dispatch.h itself, and potentially nowhere, respectively).
* require netisr.h to be included before netisr_dispatch.h.
* minor additional cleanup of both netisr.h and netisr_dispatch.h.
* clean up uses to remove now-unnecessary header file inclusions, and
local prototypes of the fns.
* convert netisr dispatch implementations which didn't use
netisr_dispatch.h (pc532) to use it.
2000-07-02 04:40:33 +00:00
707e41f21e
Fix interrupt delivery on UltraSPARC IIi machines.
2000-06-30 22:58:01 +00:00
3250ce2976
Reserve a page for dumping the kernel.
2000-06-30 22:17:53 +00:00
d738f90fd7
raise MSIZE from 128 to 256.
...
- for sizeof(void *) == 8 arch, this is mandatory. MHLEN is too small
already (less than 80) and there are chances for unwanted packet loss due
to m_pullup restriction.
- for other cases, the change should avoid allocating clusters in most cases
(even when you have IPv4 IPsec tunnel, or IPv6 with moderate amount of
extension header)
portmasters: if your arch chokes with the change (high memory usage or
whatever), please backout the change for your arch.
2000-06-30 17:55:11 +00:00
d76ae83df8
add PULLDOWN_TEST for all the platforms.
...
XXX should be moved to somewhere else when stabilized
2000-06-30 17:10:15 +00:00
ff9280f4ee
This seems to fix the problem in __semctl13.c without breaking anyghing else.
2000-06-29 20:15:24 +00:00
8906c1b836
Fix bus_space_read_region_{2,4,8}().
2000-06-29 14:10:16 +00:00
4d14e81c0b
remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
2000-06-29 07:37:53 +00:00
d82c31e07e
Typo.
2000-06-27 08:43:02 +00:00
bb2ed0f487
G/c _BSD_INTPTR_T_ and _BSD_UINTPTR_T_.
2000-06-27 05:53:22 +00:00
47b5c5e3b1
Resolve some formatting nits; add __intptr_t and __uintptr_t.
2000-06-27 04:58:51 +00:00
7d89e9904d
We want prom_claim_phys() to return the address passed into it.
2000-06-26 19:45:54 +00:00
9b1e1aeb66
prom_claim_phys: it appears one result argument too many was specified.
...
Now we can return the actual result of the call rather than merely
reflecting the input.
2000-06-26 19:41:20 +00:00
ff54ba628e
prom_claim_phys: set `align' parameter to 0 (meaning: get the chunk at the
...
specified address).
2000-06-26 19:36:07 +00:00
e695f72a2e
Add <machine/int_types.h>, which provides namespace-pure definitions
...
of exact-width integer types.
2000-06-26 15:42:16 +00:00
7e7c0311e4
Advertise we have __GENERIC_SOFT_INTERRUPTS.
2000-06-26 15:13:26 +00:00
2f159a1bac
remove/move more mach vm header files:
...
<vm/pglist.h> -> <uvm/uvm_pglist.h>
<vm/vm_inherit.h> -> <uvm/uvm_inherit.h>
<vm/vm_kern.h> -> into <uvm/uvm_extern.h>
<vm/vm_object.h> -> nothing
<vm/vm_pager.h> -> into <uvm/uvm_pager.h>
also includes a bunch of <vm/vm_page.h> include removals (due to redudancy
with <vm/vm.h>), and a scattering of other similar headers.
2000-06-26 14:20:25 +00:00
889c658b5b
Change the kernel mmap interface so that the offset to map is an
...
"off_t" and the return value is a "paddr_t" to allow mappings
at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which
only changed the offset to a "vm_offset_t".
Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.
2000-06-26 04:55:19 +00:00
53be5b215c
<vm/vm_pageout.h> is already empty; kill it totally.
2000-06-25 13:49:33 +00:00
2e1e35d58e
Enable interrupts.
2000-06-24 23:27:23 +00:00
6c5d2dbbac
With these changes the kernel seems almost stable again.
2000-06-24 20:48:36 +00:00
8e1fd2e3c6
Remove a couple of references to vaddrs.h that slipped through.
2000-06-24 16:51:34 +00:00
49bf46caf7
Add PROM symbols to 32-bit kernels too.
2000-06-24 04:42:55 +00:00
85cab8fe27
More VM cleanup:
...
Move the cpu_info structure above the interrupt stack so it won't
get corrupted if the stack overflows.
Flush the D$ before and after all MMU bypass accesses since the
D$ latches all of those.
2000-06-24 04:38:20 +00:00
25bb978ba8
Printing out %pc values should now be safe.
2000-06-24 04:26:45 +00:00
0db2dfd53f
Make these interrupt levels reflect reality a bit better.
2000-06-24 04:25:08 +00:00
cc02c0d142
masterpaddr is gone.
2000-06-24 04:24:10 +00:00
be308420d4
Get rid of vaddrs.h.
2000-06-24 04:20:58 +00:00
3bc65c6d9e
Use probeget to read kernel data to prevent the machine from falling over if the PC goes bad.
2000-06-23 19:52:57 +00:00
e9b3925f8b
Make probeget() and probeset() really work for a change.
2000-06-23 19:50:54 +00:00
133ea38323
Add a WEAK_ALIAS() macro.
2000-06-23 12:18:45 +00:00
1ee7d22c0b
Remove prefix construction for softdeps. Add SOFTDEP option to GENERIC.
2000-06-22 20:27:49 +00:00
d24721788a
Correct calculation of length for dma segments in iommu_dmamap_load().
2000-06-22 18:18:19 +00:00
75eaa6ba70
lda is unsigned, ldswa is signed. Fixes a 32-bit sign extension bug.
2000-06-20 18:06:12 +00:00
245619bad4
Have separate data and text segments, make text read-only.
2000-06-19 23:30:32 +00:00
4ffba8be2a
Fix uninitialized pointer bug.
2000-06-19 22:29:15 +00:00
cafc122317
catch up with fpu changes.
2000-06-18 08:13:45 +00:00
f44ba0a4dc
back out part of previous.
2000-06-18 08:13:26 +00:00
fab0adad15
iommu_flush becomes iommu_strbuf_flush cuz that's what it does.
2000-06-18 07:17:40 +00:00
70a9c29b9b
give BSDB_MAP a value. add a shushing cast.
2000-06-18 07:13:41 +00:00
b6026e7c35
kill dead code.
2000-06-18 07:12:39 +00:00
fde87ed5d6
add a (u_long) cast for quietness
2000-06-18 07:10:23 +00:00
0bd5307bf8
remove an unused variable.
2000-06-18 07:05:09 +00:00
9105157268
duplicate "file" should only include "compile-with"
2000-06-18 07:03:35 +00:00
4f9416205c
update to reality
2000-06-18 07:02:04 +00:00
e36a32c98f
obsolete these; use the merge sparc version.
2000-06-18 06:58:30 +00:00
41d83cead7
use arch/sparc/fpu
2000-06-18 06:57:15 +00:00
f007a974bb
turn off softdep for now; we have enough other things to worry about.
2000-06-15 18:23:52 +00:00
eccc09d9de
fix a couple of errors in the previous. from eeh.
2000-06-15 15:51:07 +00:00
e26f7acd77
my Ultra-1
2000-06-15 13:50:20 +00:00
223d7455c7
Remove the obsolete config fragments for kernel crypto, because the IPsec
...
crypto code is now in the kernel source tree.
2000-06-14 22:02:13 +00:00
3517559eb5
Missing paren.
2000-06-13 05:04:05 +00:00
fb0404cc75
Start reorganizing the kernel for MULTIPROCESSOR support.
2000-06-12 23:32:46 +00:00
65bca8bde2
Fix OF_val2sym and OF_sym2val.
2000-06-12 23:26:38 +00:00
3dfb9ee7ac
More debugging support.
2000-06-12 23:20:54 +00:00
0146a569d2
Comment describing the IOMMU DVMA address space layout.
2000-06-12 23:19:05 +00:00
cc29e9b22c
ebus_ca needs to specify enough memory to hold an ebus_softc, not a struct device.
2000-06-12 22:36:59 +00:00
69d1f5db9d
clean up cruft.
2000-06-12 05:31:30 +00:00
fdfb59b4b9
ih_map & ih_clr are volatile.
2000-06-12 05:29:43 +00:00
9918c00a61
formatting nit.
2000-06-12 05:27:27 +00:00
d26027dd5b
Only use 32-bit addresses for all accesses that don't bypass the MMU.
...
Makes SBus work again.
2000-06-10 20:51:43 +00:00
c69e701de8
Separate COPTS from CFLAGS to fix problems w/unoptimized files, and add
...
lfs_loop.c to the list that breaks optimization.
2000-06-09 15:28:21 +00:00
ed15212269
myetheraddr(): don't lose badly when the clock hasn't been attached yet.
2000-06-09 05:27:40 +00:00
7539c8d1ce
Turn on PCI MEM and DMA.
2000-06-08 23:03:17 +00:00
6c144faa4c
Allow multiple drivers to share one interrupt vector.
2000-06-08 23:01:22 +00:00
1b31fdcbb7
Fix uninitialized variable bug.
2000-06-08 22:58:42 +00:00
1dfab2070a
Override optimization for files that break gcc.
2000-06-08 21:25:18 +00:00
94ffeedb49
Allow for cacheable device maps (in case there's some RAM out there).
2000-06-08 17:59:32 +00:00
00e45ca869
Print all 64-bits of TTE data in db_pmap_kernel().
2000-06-08 17:45:29 +00:00
6baca0c1ad
Fixup signed/unsigned issues so 32-bit addresses are not sign extended.
2000-06-08 17:43:24 +00:00
860feb4e74
Fix ordering of sync operations depending on whether it's PRE- or POST-.
2000-06-08 17:41:46 +00:00
a273b333a3
Support large (>1 page) DVMA maps so UltraSPARC IIi will work.
2000-06-08 16:17:29 +00:00
08af61208e
Fix DVMA base address calculations.
2000-06-08 15:23:44 +00:00
ee41ef7a04
ensure an ebus really does exist in ebus_match; fixes spurious attachments with the ebus on a hme pci card.
2000-06-08 06:52:25 +00:00
9a8d03d6bd
Add a couple of missing semi-colons.
2000-06-07 09:16:41 +00:00
a9aa2abf94
defopt SYSCALL_DEBUG.
2000-06-06 18:52:30 +00:00
cffb580806
Implement the more flexiable `evcnt' interface as discussed (briefly) on
...
tech-kern and now documented in evcnt(9).
2000-06-04 19:14:14 +00:00
03228642f7
Use all 64 address bits in ld*a()/st*a() macros so physical device addresses
...
work. (Also add some bus_space_*() debug hooks).
2000-06-02 22:56:32 +00:00
f2ab1eb70c
db_stack_trace_cmd -> db_stack_trace_print
2000-06-02 15:38:32 +00:00
94084a33f8
New softintr interface.
2000-06-02 15:36:53 +00:00
47449a63d2
kill __P in these files. (I had to look at them anyway with an eye for
...
adding some protos... and adding them with __P seems wrong, but mixing
__P and not __P in the same file seems wrong too, so...)
2000-06-01 00:04:50 +00:00
8c2d00aaeb
Add a comment about needing to initialize p_cpu when multiple
...
processors are supported.
2000-05-31 05:09:14 +00:00
e03e9e8086
Rather than starting init and creating kthreads by forking and then
...
doing a cpu_set_kpc(), just pass the entry point and argument all
the way down the fork path starting with fork1(). In order to
avoid special-casing the normal fork in every cpu_fork(), MI code
passes down child_return() and the child process pointer explicitly.
This fixes a race condition on multiprocessor systems; a CPU could
grab the newly created processes (which has been placed on a run queue)
before cpu_set_kpc() would be performed.
2000-05-28 05:48:59 +00:00
40339b39f9
Reduce use of curproc in several places:
...
- Change ktrace interface to pass in the current process, rather than
p->p_tracep, since the various ktr* function need curproc anyway.
- Add curproc as a parameter to mi_switch() since all callers had it
handy anyway.
- Add a second proc argument for inferior() since callers all had
curproc handy.
Also, miscellaneous cleanups in ktrace:
- ktrace now always uses file-based, rather than vnode-based I/O
(simplifies, increases type safety); eliminate KTRFLAG_FD & KTRFAC_FD.
Do non-blocking I/O, and yield a finite number of times when receiving
EWOULDBLOCK before giving up.
- move code duplicated between sys_fktrace and sys_ktrace into ktrace_common.
- simplify interface to ktrwrite()
2000-05-27 00:40:29 +00:00
907ed5fa81
Add `-D' flag to drop to the PROM before jumping to the kernel.
2000-05-26 22:59:51 +00:00
a7d0570e67
First sweep at scheduler state cleanup. Collect MI scheduler
...
state into global and per-CPU scheduler state:
- Global state: sched_qs (run queues), sched_whichqs (bitmap
of non-empty run queues), sched_slpque (sleep queues).
NOTE: These may collectively move into a struct schedstate
at some point in the future.
- Per-CPU state, struct schedstate_percpu: spc_runtime
(time process on this CPU started running), spc_flags
(replaces struct proc's p_schedflags), and
spc_curpriority (usrpri of processes on this CPU).
- Every platform must now supply a struct cpu_info and
a curcpu() macro. Simplify existing cpu_info declarations
where appropriate.
- All references to per-CPU scheduler state now made through
curcpu(). NOTE: this will likely be adjusted in the future
after further changes to struct proc are made.
Tested on i386 and Alpha. Changes are mostly mechanical, but apologies
in advance if it doesn't compile on a particular platform.
2000-05-26 21:19:19 +00:00
7d2fdbcb55
Don't set `runtime' here; it's done at the end of mi_switch().
...
cvS: ----------------------------------------------------------------------
2000-05-26 05:27:29 +00:00
8e44b27348
Rename the machine-specific stack trace printing functions
...
from db_stack_trace_cmd() to db_stack_trace_print(),
and add an additional argument, a function pointer for an
output routine (i.e. printf() or db_printf()).
Add db_stack_trace_cmd() in db_command.[ch], calling
db_stack_trace_print() with db_printf() as the printer.
Move count==-1 special handling from db_stack_trace_print() [nee
db_stack_trace_cmd()] to db_stack_trace_cmd() [nascent here].
Again, I'm unable to test compilation on all affected platforms,
so advance apologies for potential brokenness.
2000-05-26 03:34:24 +00:00
8964c35eca
Introduce a new process state distinct from SRUN called SONPROC
...
which indicates that the process is actually running on a
processor. Test against SONPROC as appropriate rather than
combinations of SRUN and curproc. Update all context switch code
to properly set SONPROC when the process becomes the current
process on the CPU.
2000-05-26 00:36:42 +00:00
f3528d725d
Interface change: db_printsym() takes a third argument, pr, a function
...
pointer indicating how to print the symbol. This allows db_printsym()
to called in places where db_printf() is not an appropriate output
function.
While straightforward, apologies in advance if I've introduced any minor
syntax errors; I was unable to test compilation this on all the affected
platforms.
2000-05-25 19:57:30 +00:00
ca743bac9a
Use the PCI function code to distinguish simba bus A from bus B instead
...
of the current hack.
2000-05-24 20:27:52 +00:00
c965d1deec
If we have sparse PCI spaces we can run out of IO map space when mapping
...
PCI config space. Since PCI config space is mostly used by PCI bus drivers,
we won't actually map it in. Instead we use MMU bypass ASI accesses to read
and write PCI config space.
2000-05-24 20:24:58 +00:00
1140468205
Use preempt(), not an open-coded equivalent (which won't be
...
equivalent for long).
2000-05-24 16:48:33 +00:00
8b4fa50a4d
Need to be able to turn off pcons.
2000-05-23 05:11:28 +00:00
62d33e59ff
Align things pessimistically so a 64-bit kernel will accept symbols from a 32-bit loader.
2000-05-22 19:15:33 +00:00
296eaaa36d
add two new CPU_ISXXX versions: CPU_ISSUN4U, CPU_ISSUN4MOR4U
2000-05-22 02:35:23 +00:00
17c567f9de
Oops. Seems like we do need this file to make pcons work.
2000-05-20 14:23:12 +00:00
071aed40ac
A foolish consistency; most parts of the kernel use bp->b_data, so
...
change these from bp->b_un.b_addr to bp->b_data, as well. This also
allows us more flexibility to experiment with other data buffer types
hung off of struct buf.
2000-05-19 18:54:22 +00:00
dd5143017a
Use the new emergency fallback PROM console driver.
2000-05-19 05:28:47 +00:00
424619ca1a
Fix the sparc64 console.
...
Unlike the other Sun machines, UltraSPARCs can have consoles run on different
chips than zs, so we need to support them. So, here we go:
Add a new PROM console driver with a major number and everything.
This is the default driver if nothing else attaches. It does not
use the keyboard driver since the PROM translates keystrokes itself.
(Unfortunately it also swallows L1-A).
Have the keyboard driver take over the console when it attaches on a
serial port. When a serial port detects a keyboard and attaches the
keyboard driver, it needs to provide a set of consdev vectors. They
keyboard driver will use those to send I/O to the keyboard and mouse.
2000-05-19 05:26:16 +00:00
4bd0bb352b
this file has not been used for a long time.
2000-05-19 01:09:21 +00:00
0cd44feb0e
fix if/ifdef mistake
2000-05-17 10:28:14 +00:00
20bc66f592
- add some psycho interrupt debugging stuff (#if 0'ed out)
...
- fix arguments passed in psycho_dmamap_load_raw()'s declaration, and
what it passes to it's parent bus.
2000-05-17 10:17:01 +00:00
eee5e33e8a
- in iommu_remove(), move len adjustment outside if statement, leaving
...
just one copy of this.
- remove duplicate setting of sgsize in iommu_dvmamap_load().
- fix DIAGNOSTIC check in iommu_dvmamap_load_raw() to panic() in unexpected
conditions only (not normal ones), and also add an extra case.
2000-05-17 09:53:53 +00:00
53fa9da4df
move zs_attach() to after the setting of the console device, to avoid
...
having cn_dev being 0 and thus cnopen() being called recursively. this
fixes this problem for zs-based console's only.
2000-05-17 09:28:22 +00:00
651712a140
add a note about US IIi and PCI_INTERRUPT_LINE register
2000-05-17 09:25:58 +00:00
6dad37e26a
expand a comment slightly.
2000-05-17 09:25:27 +00:00
f2a19f58b5
fix || vs. && bug, and clean up some cruft.
2000-05-17 09:16:44 +00:00
0eeab9735f
fix another warning.
2000-05-17 09:12:10 +00:00
3b90b5c5c4
Add support for bus_dma_load_raw() so the le driver will work again.
2000-05-17 02:31:12 +00:00
463931b3ba
Nuke dk_establish() from orbit except from those ports which still use
...
it to determine the boot device: mvme68k, pc532, macppc, ofppc. Those
platforms should be changed to use device_register(). In the mean time,
those ports defined __BROKEN_DK_ESTABLISH.
2000-05-16 05:45:44 +00:00
9f48bba3ae
Add "install" target, so "make install" after building the kernel
...
does something useful. The target can be redefined by putting a
install-kernel-${MACHINE_NAME} target that fits your needs into
/etc/mk.conf.
2000-05-09 00:56:21 +00:00
3a6e3a4670
Add (sometimes commented out) MIIVERBOSE option.
2000-05-08 13:49:44 +00:00
0ce52a2a08
implement U2P PCI support in psycho_init. this is completely untested.
2000-05-06 04:15:35 +00:00
0b2854eb6c
Fix for lint during build of libc.
2000-05-05 20:12:00 +00:00
855b79db92
Let each platform typedef the new __cpu_simple_lock_t, which should
...
be the most efficient type used for the atomic operations in the
simplelock structure, and should also be __volatile.
2000-05-02 04:41:04 +00:00
a2edf1690f
Fix a declaration inside __ldstub().
2000-05-01 00:46:29 +00:00
dc8f3d3b3f
Oops, install <machine/lock.h> on these systems, too.
2000-04-29 03:45:42 +00:00
f51470a514
Require that each each MACHINE/MACHINE_ARCH supply a lock.h. This file
...
contains the values __SIMPLELOCK_LOCKED and __SIMPLELOCK_UNLOCKED, which
replace the old SIMPLELOCK_LOCKED and SIMPLELOCK_UNLOCKED. These files
are also required to supply inline functions __cpu_simple_lock(),
__cpu_simple_lock_try(), and __cpu_simple_unlock() if locking is to be
supported on that platform (i.e. if MULTIPROCESSOR is defined in the
_KERNEL case). Change these functions to take an int * (&alp->lock_data)
rather than the struct simplelock * itself.
These changes make it possible for userland to use the locking primitives
by including <machine/lock.h>.
2000-04-29 03:31:45 +00:00
dab52c26ad
- add a note about streaming buffers and US IIi not having them.
...
- make some debugging messages in iommu_remove() saner and add some more.
- decrement 'len' in the no streaming buffer case, also.
- in iommu_dvmamem_map(), do not enter these mappings into the IOMMU,
only into the CPU (the former is done at _load time).
- make a panic that shouldn't happen a DIAGNOSTIC.
2000-04-25 14:59:38 +00:00
5099c60132
If we trap due to a bad kernel stack, try to switch to the interrupt strack
...
before breaking into the debugger.
2000-04-22 22:06:06 +00:00
1013298c2a
whole bunch of changes:
...
- merge IOMMU DVMA code from sbus/psycho into iommu.c. this code was
identical and a few minor inconsistencies had crept in. this way
keeps them all in sync.
- with this code gone from psycho, merge the psycho.c and psycho_bus.c
files. same with ebus/ebus_bus.c. delete the _bus.c files.
- add a _ds_boundary member to the dma segment structure, so that later
dma mappings can find this value.
- set _ds_boundary in machdep.c:_bus_dmamem_alloc().
- kill much dead code.
2000-04-22 17:05:59 +00:00
db956d0701
s#\b#\n# so an error message appears correctly.
2000-04-22 16:42:25 +00:00
5aa4575432
- convert to DPRINTF() style debug messages.
...
- kill dead code.
- use the sbus dv_xname to construct a dvma map map, to pass to
extent_alloc(), to help support multiple sbusses
2000-04-22 12:36:29 +00:00
6e576e3c03
clean up mdallocsys().
2000-04-22 12:03:33 +00:00
e5ab0feccc
Have prom_cngetc() wait until input has been typed.
2000-04-19 08:15:06 +00:00
3be41bae18
add stf pseudo interface (commented out due to possible security risks)
2000-04-19 06:50:27 +00:00
19a31e7b5e
add a newline to this fake autoconfig line.
2000-04-19 02:26:08 +00:00
c58f0ee9cd
Prevent the system from getting watchdog reset by returning directly to user
...
mode if restoring the user window traps. Also reduce the time we run with
%tl>0.
2000-04-18 02:12:25 +00:00
2aec15d3d3
Turn off traptrace before printing a panic message so we don't flush the useful
...
traces out with the printf guts, and wrap some lines at 80 columns.
2000-04-18 02:07:57 +00:00
2aba26f773
- auxio cleanup and sbus support.
...
- make GENERIC64 include GENERIC and set the 3 optoins it needs. suggested
by hubert feyrer.
- add a comment that we maybe should use the `bpp' driver, not the lpt, on
the ebus because the `bpp' driver does DMA already.
- ebus_attach_args got a member renamed
2000-04-15 03:08:12 +00:00
e4ca29ac00
kill dead code
2000-04-15 03:07:38 +00:00
7865e586d3
Typo in comments. (Zilog 8350 -> 8530)
2000-04-14 13:29:57 +00:00
01ff053e58
add bpt and lpt. fixes final gcc warnings for GENERIC.
2000-04-14 08:36:39 +00:00
af6a8541e4
add some casts to (u_long) to shut up gcc. cleanup unused variables
2000-04-14 08:29:03 +00:00
726f5f3169
two more compiler warnings killed.
2000-04-14 04:08:26 +00:00
9b019404f8
Apparently gas 2.9.5 (don't ask) does not like to add values to %lo()
...
directives so move the additions inside %lo() to make it happy.
Hopefully this won't break older versions.
2000-04-13 18:40:27 +00:00
92aaaad1e1
Make /etc/mk.conf optional for kernel builds.
2000-04-13 16:57:51 +00:00
fd85c1c4a2
avoid a potential memory leak.
2000-04-13 14:39:34 +00:00
5792df1cf9
- do the "options<space><tab>" thing everywhere.
...
- disable COMPAT_AOUT.
- enable hme at sbus/pci
- enable bpp at sbus
- clean up esp attachments
- enable all xxphy's
- sync GENERIC & GENERIC64.
2000-04-13 11:23:14 +00:00
fa13d97855
this does not need auxreg.h
2000-04-13 09:54:43 +00:00
b60fe7f45a
store all the physical address bits.
2000-04-13 09:53:49 +00:00
e7f3d69787
this does not need auxreg.h
2000-04-13 09:52:46 +00:00
542f08a422
add a version file.
2000-04-13 09:38:12 +00:00
f2ab01bf42
copy the "sparc" clean/cleandir target code, so that "cleandir" actually does a "clean"
2000-04-13 09:23:38 +00:00
a388797ed5
Ignore unknown devices on mainbus.
2000-04-12 11:40:12 +00:00
1f7a988bd2
unconditionalize "ifdef sparc".
2000-04-12 04:06:41 +00:00
1fdef1b204
deja vu: add "struct device *booted_device;" so raidframe links.
2000-04-12 00:47:10 +00:00
b2d31a5cc9
- fix a speeling error
...
- rename some numbered labels to be a bit less confusing.
- remove dead code
- fix a debug error message (argh!)
2000-04-10 16:28:56 +00:00
cb2b15a484
- UltraSPARC IIi does not have streaming buffers; don't try to find it's
...
diag regsiter to work out why the (non-existant) strbufs don't work.
- check for malloc failure in _all_ places.
- setup the PCI control register as recommended in the IIi users manual.
2000-04-10 16:11:23 +00:00
79f22c2c51
UltraSPARC IIi CPU has bits 6-10 of the INO hardwired to 1. when we
...
install an interrupt handler, make sure we set these bits to 1. now,
interrupt_vector in locore can find our registered interrupt handlers
and at least try to setup a call to them.
2000-04-10 16:07:50 +00:00
304778e6c2
- IRSR_BUSY is bit 5 not bit 4.
...
- IRSR_MID(x) is &'d with 0x1f (5 bits) not 0xf.
2000-04-10 16:05:13 +00:00
f7618dcb04
add 4 more bits for the PCI control register
2000-04-10 16:04:02 +00:00
cfadc0991a
Get a kernel without DDB to build.
2000-04-10 13:34:18 +00:00
e2ed27832e
with these two, POWEROFSEVEN now compiles cleanly.
2000-04-08 15:15:41 +00:00
171d3a1576
don't use -Wuninitialized
2000-04-08 04:46:15 +00:00
d0be8351d9
add casts to stop gcc complaining
2000-04-08 04:45:41 +00:00
4583d39d24
correct the pci ino->ipl table. shut up gcc once more.
2000-04-08 04:35:01 +00:00
5b5f175f38
- store the ebus's parent so we can find the IOMMU
...
- use generic iommu routines in the ebus code (uses the above)
- add some more comments.
2000-04-08 04:33:09 +00:00
d360a37afd
- remove unused variable.
...
- include <sparc64/sparc64/cache.h> to get cache_flush() prototype.
- remove an unnecessary cast.
2000-04-08 04:17:31 +00:00
34d04ff6db
include <ddb/db_output.h> to get the db_printf() prototype.
2000-04-08 03:09:00 +00:00
748cfc6b99
rearrange a little to accommodate real SUNW,psycho support.
2000-04-08 03:08:20 +00:00
70e38e4034
actually, i wrote these in 2000
2000-04-08 03:07:07 +00:00
4044b5a41c
Use separate callouts for motor-on and motor-off. Fixes a condition
...
where the floppy driver would wedge because a motor-on timeout would
be cancelled by another I/O operation cancelling a motor-off timeout.
From enami tsugutomo <enami@sm.sony.co.jp>.
2000-04-07 16:58:53 +00:00
580ccd66ad
make this work with years-old new-style mdsetimage, etc.
2000-04-06 13:39:26 +00:00
966add799c
avoid some more compiler warnings, and fix a couple of errors.
2000-04-06 13:32:49 +00:00
eeh