96381cfdb1
4.99.49: struct vnode changed.
2008-01-09 16:16:27 +00:00
0b52913dee
Go back to freeing on disk inodes in the inactive routine. It would be
...
better not to do this, but it rules out potential side effects with softdep.
2008-01-09 16:15:22 +00:00
be408dd2a7
Add options SYMTAB_SPACE for ksyms(4) and ddb(4).
2008-01-09 15:45:27 +00:00
9b1f3b007e
Remove the unused mutex from softc.
2008-01-09 15:44:33 +00:00
fce49684a2
Add GPIO_PIN_PULLDOWN, GPIO_PIN_INVIN and GPIO_PIN_INVOUT for gcscpcib(4),
...
from OpenBSD.
2008-01-09 15:11:30 +00:00
53fd152fd5
Remove useless returns at the end of void functions.
2008-01-09 14:52:52 +00:00
e430816fa1
emo-style(9).
2008-01-09 14:49:48 +00:00
774fc1f9f9
Fix another botch on migration to MI mips3_clock.c for algor in 2006:
...
- in algor_p????_cal_timer() functions, initialize ci_divisor_delay
and ci_cycles_per_hz values in curcpu() for MI mips3_clock.c:mips3_delay()
rather than obsolete and homegrown delay_divisor
- remove old _delay() stuff from locore_machdep.S
- make MI mips3_delay() work even before frequency values in curcpu()
are initialized
Now NetBSD/algor P5064 kernel works on gxemul.
While here, add a hack for emulators:
- ignore a measured CPU frequency if the value looks unlikely
Pullup request for netbsd-4 will be sent later.
2008-01-09 14:46:48 +00:00
cbfd19ee15
+ gcscpcib(4)
2008-01-09 14:44:51 +00:00
f5d88c37cb
Fix up some comments, from Yojiro UO in private email.
2008-01-09 14:32:36 +00:00
63cd4d7992
Add gcscpcib(4).
2008-01-09 14:28:54 +00:00
810aa8579e
Add gcscpcib(4): a driver for the AMD CS5535 and CS5536 Companion Device
...
that acts as PCI-ISA bridge and supports a Timecounter, Watchdog Timer
and GPIO.
Adapted by Yojiro UO and minor tweaks by me from OpenBSD. Tested on
CS5535 and CS5536. This closes PR kern/37577.
2008-01-09 14:23:46 +00:00
81ce14a30e
Fix botch on MI todr(9) migration of algor on September 2006:
...
- year0 should be 1980, not 80
- put a newline after MI mc146818_attach()
- call todr_attach(9)
2008-01-09 14:16:35 +00:00
7d83a06dfc
Declare nbpg if we're building for sparc, and initialize to the
...
smallest sparc pagesize. This should deal with the IOCPARM_MAX
build problem, which occurs on sparc because various models have
different page size, so it's run-time determined instead.
pooka says that rump doesn't use ioctl(), so the value is apparently
of no consequence.
2008-01-09 13:59:43 +00:00
e6599f15b0
Fix headers.
2008-01-09 13:52:33 +00:00
28302167c1
Remove unused variables
2008-01-09 13:24:57 +00:00
d8de5b102b
Remove support for NetBSD/pc532.
2008-01-09 11:25:58 +00:00
1d97c261e0
remove now unused variable.
2008-01-09 11:13:16 +00:00
4b3ae01c7d
Refactor part of the sys_revoke() code so that it can be used in the
...
compat code. Allows for the removal of two redundant kauth(9) calls.
okay christos@.
2008-01-09 08:18:12 +00:00
7e7e1a0aef
Kill two KAUTH_GENERIC_ISSUSER requests and replace them with something
...
more meaningful.
2008-01-09 08:15:53 +00:00
58b3d9db6c
extend the grossness that is pegasospci_indirect_attach_hook() a little
...
further and enable the vr(4) located at pci 0 dev 13 fn 0.
now my vr(4) works from disk boot.
2008-01-09 07:35:29 +00:00
a191b55bff
Bump uvmexp.intrs when we get a clock or statclock interrupt.
2008-01-09 06:50:36 +00:00
d9a61c14e9
fix SPINLOCK_BACKOFF_HOOK.
2008-01-09 00:23:18 +00:00
343d0e5699
kill unused PMF_USER_RELOAD.
2008-01-08 22:24:09 +00:00
b7ce913912
pmap_map_ptes: fix a access-after-free bug.
...
add a reference on ci->ci_pmap before acquiring locks, so that it won't
disappear while we are blocking on the locks.
2008-01-08 22:22:30 +00:00
9798c4cd61
In svr4_32_sys_systeminfo(), remove two redundant kauth(9) calls for
...
SVR4_SI_SET_HOSTNAME and SVR4_SI_SET_SRPC_DOMAIN, as permission is already
checked by old_sysctl().
okay christos@.
2008-01-08 22:13:07 +00:00
2395a8f2d2
In linux_sys_sysmips(), remove redundant kauth(9) call for LINUX_SETNAME
...
as the permission check is done by old_sysctl().
okay christos@.
2008-01-08 22:09:48 +00:00
3d902dfb3e
Make compat_43_sys_sethostid() use old_sysctl(), allowing for the removal
...
of a kauth(9) call.
okay christos@.
2008-01-08 22:08:45 +00:00
145b26e5d2
In svr4_sys_sysarch(), remove redundant kauth(9) call. Permission is
...
checked by old_sysctl().
okay christos@.
2008-01-08 22:07:29 +00:00
8e4891367e
cpu_intr_p() for sparc, from Andrew Doran.
2008-01-08 21:32:10 +00:00
b2351ba0df
pacify lint with casts.
2008-01-08 20:56:22 +00:00
34dc4c6501
Switch Xen to generic TODR. Tested by Manuel Bouyer.
2008-01-08 20:37:34 +00:00
98b295d02d
uobj->vmobjlock is a mutex these days.
2008-01-08 19:26:44 +00:00
1ad062e684
Allocating up to 8k on stack isn't a good idea ...
2008-01-08 19:26:10 +00:00
7d4e4f0e9a
The test for the 0xaa55 at the end of an extended partition pbr has always
...
been incorrect (and checked the end of the mbr - which always suceeds).
Rather than fixing the text, just delete it (saving 8 bytes from mbr_ext).
2008-01-08 18:31:56 +00:00
50b7998db1
After rtw_attach() is called, the hardware has been powered down,
...
so don't fiddle with the Function Event Registers.
2008-01-08 18:26:09 +00:00
732cd3c8a0
Convert Atari to generic TODR and timecounter.
2008-01-08 18:04:15 +00:00
30cd7dc199
assym.h dependency.
2008-01-08 17:46:31 +00:00
03b0d74e1e
Set proper mask values for ipl_sr_bits[] and
...
use C99 initializer to avoid confusion.
Tested on gxemul.
2008-01-08 16:15:04 +00:00
808b08ff6f
Short-cut interrupt processing if device has been suspended.
...
uvm_fault reported by xtraeme.
2008-01-08 15:50:35 +00:00
58facc2685
Remove __HAVE_TIMECOUNTER conditionals.
2008-01-08 14:46:28 +00:00
c4bf54683c
Add timecounter support for sbmips machines and Broadcom SiByte CPUs.
...
Based on patch from joerg@, with tweaks by me to work with sbmips
non-use of the CP0 count/compare registers for clock interrupts.
2008-01-08 14:38:47 +00:00
585198d4ab
I'm not positive that the mips_ipl_si_to_sr assignments are correct - but
...
at least the thing compiles now.
2008-01-08 14:28:35 +00:00
da6c04f088
explicitly include <sys/device.h>
2008-01-08 14:26:58 +00:00
a280711321
Mask is unsigned, add missing u.
2008-01-08 13:52:00 +00:00
18a5a8d13d
Finish conversion of ppc and evbppc to timecounter.
...
Tested by simonb@ on WALNUT.
2008-01-08 13:47:49 +00:00
ca9c510ab2
explicitly include simplelock.h
2008-01-08 13:28:22 +00:00
b24bad5798
add include for incomplete struct blah blah blah
2008-01-08 13:26:26 +00:00
42f654a26b
change the layout in u-area and reduce UPAGES.
2008-01-08 13:15:01 +00:00
74e0872eca
simplify locking and remove vm_map_upgrade/downgrade.
...
this fixes a deadlock due to read-lock recursion of map->lock.
2008-01-08 13:09:55 +00:00
05379725b5
lock_stubs.S needs assym.h.
2008-01-08 12:17:32 +00:00
71990d8d21
ANSIfy.
2008-01-08 12:05:49 +00:00
99cad24ec6
GC this file - it should have been removed during the ppcoea-renovation
...
merge.
2008-01-08 11:54:47 +00:00
943b531346
Remove a no-longer-used local variable.
2008-01-08 07:59:23 +00:00
e1adc2881b
Add IPL_SOFT* defined at IPL_NONE.
...
Change _ARM32_INTR_H to _ACORN32_INTR_H
2008-01-08 06:30:22 +00:00
306d43c5bf
Remove unused variable.
2008-01-08 06:29:39 +00:00
6d2503c7a9
Don't do arithmetic on void pointers.
2008-01-08 06:28:29 +00:00
e0b21313b6
In device_foreach_child(), use a safe idiom for walking a list
...
whose elements we might be deleting. This stops us from crashing
in config_detach_children().
2008-01-08 06:27:46 +00:00
2414d5854f
Make sys_swapctl match syscallargs.h
2008-01-08 06:25:55 +00:00
b202d18acd
Use aprint_*_dev(), device_xname(), and device_t. Join a line.
2008-01-08 04:47:44 +00:00
f384d52f35
Delete an unnecessary cast.
2008-01-08 03:37:45 +00:00
6bc0c58216
As of this commit, all arm32 kernel now build.
2008-01-08 02:07:50 +00:00
f4f1c6c970
More kernels SYMTAB_SPACE increased.
2008-01-08 02:06:43 +00:00
04a4a202ed
include "conf/std"
2008-01-08 02:05:30 +00:00
71fc6881ec
In general_exception() in do_panic case print error message *before*
...
we drop into ddb.
2008-01-08 01:25:13 +00:00
1424adf410
Support FIONBIO.
...
Support non-blocking reads from USB printers. With this, escputil can
query the printer model, and I can print a test page with gutenprint
5.1 and cups. Add many comments explaining why the support is not
quite right. (This change is quite a bit less invasive than a
"correct" fix, and thus more reasonable to pull up to netbsd-4.)
Rototill debugging, so that all DPRINTFN have a level, and so that the
message is printed if the debug level is >= the number in the source
code. Document debug level plan.
Fix bug where only one status change would be printed (by inspection,
not tested).
Document scheme for reading/discarding data when ulpt(4) is opened for
write only.
Delete dead code.
Plan and patch posted on tech-kern without objections.
2008-01-08 00:58:09 +00:00
c0187a5323
pmap_destroy: cast return of atomic_dec_uint_nv to (int) to avoid leaking
...
pmaps. The refcount can be zero due to pmap_remove_all. Problem found by
tnn@.
2008-01-08 00:45:23 +00:00
72b3924d8b
Set stacksize to minimum acceptable (two pages, one for cc, one
...
for actual stack) to mimic kernel stack size better.
2008-01-08 00:23:15 +00:00
964f16c2b9
Make fork use kauth.
...
Been running in my tree for over a month at least.
Reviewed and okay yamt@, and special thanks to him as well as rittera@
for making this possible through fixing NDIS to not call fork1() with
l1 != curlwp.
2008-01-07 23:51:06 +00:00
0b5d0f6802
Actually copy the ras list onto the child process during fork() - hi ad.
...
Might fix the 'ras' part of kern/37650.
For some reason the itimers aren't inherited either.
2008-01-07 20:56:32 +00:00
574389c2a0
Fix build if !MODULAR. Reported by elad@.
2008-01-07 20:42:48 +00:00
c22be27a3d
Added elad's changes to ndis_kthread_create. Setting the stack size in fork1 apparently has no effect in the kernel. Note: needs testing
2008-01-07 19:10:58 +00:00
b7dafc15be
Don't both about module name / ksyms until the name is known for sure.
2008-01-07 18:25:56 +00:00
e47a5502a7
Use gettimeofday as timecounter.
2008-01-07 17:27:12 +00:00
0a53bb0b5b
x86 always has timecounter support.
2008-01-07 16:58:09 +00:00
52603a7d6d
Fix 'panic: softdep_update_inodeblock: update failed'.
2008-01-07 16:56:27 +00:00
ebd253a526
Garbage collect !__HAVE_TIMECOUNTER code.
2008-01-07 16:55:15 +00:00
2937b71b26
Add PaX ASLR for testing
2008-01-07 16:46:59 +00:00
7e82e3084b
Timecounter support for VAX. Basic testing in simh done on a backported
...
version to netbsd-4.
2008-01-07 16:40:16 +00:00
72506c7508
Regen.
2008-01-07 16:16:38 +00:00
fe8aceeb5e
More MPSAFE syscalls.
2008-01-07 16:15:36 +00:00
8b52263cd1
Hold kernel_lock across IFNET_FOREACH().
2008-01-07 16:13:49 +00:00
77e0b2120f
Patch up sysctl locking:
...
- Lock processes, credentials, filehead etc correctly.
- Acquire a read hold on sysctl_treelock if only doing a query.
- Don't wire down the output buffer. It doesn't work correctly and the code
regularly does long term sleeps with it held - it's not worth it.
- Don't hold locks other than sysctl_lock while doing copyout().
- Drop sysctl_lock while doing copyout / allocating memory in a few places.
- Don't take kernel_lock for sysctl.
- Fix a number of bugs spotted along the way
2008-01-07 16:12:52 +00:00
eb4920235a
Don't drop vnode refs until we are done with the mount.
2008-01-07 16:08:46 +00:00
2b341ae01b
Add timecounter(9) support for mvme68k.
...
Based on a patch from Garrett D'Amore and several tweaks by me.
Compile tested only, but reviewed by joerg@ and
no objection from scw@ on port-mvme68k.
2008-01-07 14:39:28 +00:00
7f4f83c80c
Shared clock.c with amiga, so it supports generic time counter.
2008-01-07 13:55:36 +00:00
a32939fca4
bwrite, bdwrite: bufcache_lock must be held for reassignbuf.
2008-01-07 12:50:38 +00:00
297c776822
Make VFS_SYNC() and friends functions, and have them acquire kernel_lock
...
if the FS is not marked MPSAFE.
2008-01-07 12:34:12 +00:00
e5e170b198
Regen.
2008-01-07 12:12:12 +00:00
0dbd478087
Remove wrong linux32_sys_mmap2_args define.
2008-01-07 12:11:52 +00:00
2e8a5bee68
lwp_ctl_alloc: fix error handling.
2008-01-07 11:41:29 +00:00
cfde7482cd
Actually return the computed uptime in microuptime(). Fixes many
...
wierd timing issues on non-timecounter ports (tested on sbmips, and
omap ARM by Marty Fouts).
Also move the microtime() inside the splclock() block in the same
function - it may be possible for the "time" variable to be updated
after the call to microtime() but before we use it to calculate the
offset from mono_time.
2008-01-07 10:28:07 +00:00
7e40665ceb
Fixed indentation, it should be
...
option<space><tab>
makeoption<space><tab>
2008-01-07 08:39:03 +00:00
a678f921c8
Fixed option indentation, it should be
...
option<space><tab>
makeoption<space><tab>
2008-01-07 08:30:03 +00:00
73b30d3cfe
Define some "special" values of register MMCR_WPVMAP.
2008-01-07 08:01:45 +00:00
40fec83596
Delete unused mips_ipl_si_to_sr[].
2008-01-07 07:28:14 +00:00
cb1ed32e14
Update AR5312 interrupt masks along the same lines as the AR5315.
2008-01-07 07:14:37 +00:00
b82c6f8442
Cosmetic: re-number the comments on elements of ipl_sr_bits[].
2008-01-07 07:12:06 +00:00
189256ad32
Make this compile again by making changes similar to the changes
...
that ad@ made to arch/mips/alchemy/au_icu.c. Compiles and runs,
but this probably deserves a looksie by someone with more MIPS
clue.
2008-01-07 06:55:32 +00:00
b82e361f77
Use pmf(9). Use device_t and device_private().
2008-01-07 06:53:08 +00:00
303f9ebb5b
Make the ACPI_MADT_DEBUG code compile again.
2008-01-07 06:04:07 +00:00
51f964a289
softdep_freefile: don't acquire ufsmount lock twice.
2008-01-07 05:20:25 +00:00
e97e9077cc
Note landisk import date.
2008-01-07 05:02:38 +00:00
f38b1e35de
Now that sh3 has didicated interrupt stack we can check for cpu_intr_p
...
with simple sp comparison, so we can drop ci_idepth.
2008-01-07 05:00:12 +00:00
b786c078f9
Add endif comment for __HAVE_FAST_SOFTINTS.
2008-01-07 04:43:54 +00:00
1e0e088d60
cpu_intr_p - we can simply check if we are running on the interrupt stack.
2008-01-07 04:41:40 +00:00
d534bbfa2e
Really, remove last vestiges of simplelock debugging.
2008-01-07 01:55:15 +00:00
907a6a54db
Fix dodgy assertions. From Anders Hjalmarsson.
2008-01-07 00:43:27 +00:00
b439191858
Remove vestiges of simplelock debugging.
2008-01-07 00:27:35 +00:00
de5c0b29ed
Build kobj_machdep.c if MODULAR.
2008-01-07 00:26:32 +00:00
521d96c676
The stamp mutex is not unsafe, so remove this comment so people don't get the wrong idea. ok ad@
2008-01-07 00:25:19 +00:00
9202b67eca
Overzealous; pull in machine/lock.h again.
2008-01-07 00:20:18 +00:00
d8f0d4be74
Catch up with new option to dev/pckbport/wskbdmap_mfii.c
2008-01-07 00:19:08 +00:00
80f35bef49
#include <sys/mutex.h>
2008-01-06 20:53:06 +00:00
d879c8360a
defflag one moe variable.
2008-01-06 19:42:33 +00:00
d1b79165fa
clamp min/max values properly.
2008-01-06 19:42:03 +00:00
c04c9a55d0
Add timecounter support and borrow counter-based delay from i386.
...
From joerg; testing and final tweaks by me.
2008-01-06 18:50:29 +00:00
890508dbe9
argh, I did not intend to remove all declarations of physmem.
2008-01-06 18:35:24 +00:00
ae4518d415
Remove superflouse extern decls.
2008-01-06 18:24:06 +00:00
fb8f17133f
Move the mutex_init of pmap_lock even later into pmap_init().
...
This makes LOCKDEBUG kernels work again.
2008-01-06 18:07:16 +00:00
0805a9416a
Add MODULAR option. Include kernel linker stuff if MODULAR.
2008-01-06 18:03:58 +00:00
9a2bbccfc7
Load the next cpu_info pointer with LDPTR instead of ld.
2008-01-06 17:22:15 +00:00
bf49138bfc
Fix off-by-one in kobj_read range checking for case KT_MEMORY
2008-01-06 15:13:07 +00:00
b9e6d93835
- Allow loading ELF objects from memory. Based on a patch from jmcneill.
...
- Allow for changing object name later, once known.
- Split open/load of the objects.
2008-01-06 14:47:26 +00:00
dff5d8444c
Only do the xcall if !cold.
2008-01-06 14:33:27 +00:00
55cdd729c2
Fix a typo: s/Kgdb_dev/kgdb_dev/
2008-01-06 13:54:01 +00:00
2a8b08575b
We can not mutex_init() before setting up the per-CPU mappings needed
...
by curcpu() - a LOCKDEBUG kernel will die otherwise.
2008-01-06 13:41:28 +00:00
9602a50eb5
recognized the names of Sigmatel STAC9227-9230
2008-01-06 13:33:21 +00:00
67cc6c4e0d
Update curcpu()->ci_idepth, not the (left over) local.
2008-01-06 13:28:17 +00:00
b6e6554ace
include sys/simplelock.h
2008-01-06 13:25:22 +00:00
f3df4c2289
include sys/simplock.h
2008-01-06 13:18:19 +00:00
86e51d72d1
fix a bug of S/PDIF audio format selection
...
PR/36493 by Arcady Antipin
2008-01-06 12:56:20 +00:00
4507f75dd0
Fix fallout from vmlocking2: in xbdback_co_io_loop(), return the right object.
2008-01-06 12:43:42 +00:00
941409f5f4
curcpu()->ci_depth is now updated in irq_dispatch.S, remove duplication
...
from footbridge.
2008-01-06 12:17:13 +00:00
eb14906cb2
Include <sys/device.h> explicitly to pull in needed definitions.
2008-01-06 10:33:24 +00:00
d9322458cc
Truly kill current_intr_depth once and for all.
2008-01-06 03:45:26 +00:00
940cdbb355
current_intr_depth is dead. Make sure we don't use it anymore.
2008-01-06 03:11:42 +00:00
ace1262f40
current_intr_depth is dead. kill it off for good.
2008-01-06 03:01:59 +00:00
258a4b78cf
With the new IPL world, things are easier for us: if we can get the
...
nell hardware interrupt handler run at IPL_VM, we can call the socket
drivers interrupt handler directly.
This is always possible on sparc64, but on sparc we might have to fall
back to the old softint bounce. Since this uses arbitrary IPLs, we
can not use the new softint_* for this - we'll have to use the old
sparc_softintr_* functions.
2008-01-06 02:29:58 +00:00
0c0de8072e
Cleanup softint lossage.
2008-01-06 01:37:53 +00:00
3ef8cd67cc
Increase SYMTAB_SPACE because of more kernel bloat.
2008-01-06 01:35:34 +00:00
e7574fcb84
Make sure interrupt handler is updating curcpu()->ci_depth
2008-01-06 01:19:42 +00:00
503fb6483e
A quick hack to make cpu_intr_p() work on shark. Needs to be fixed
...
properly by someone who does arm assembly.
2008-01-06 00:56:18 +00:00
e5942a85c1
Missing lock acquire.
2008-01-05 23:54:24 +00:00
58dc3540b0
Add fgetdummy/fputdummy: allocate and free dummy 'struct file' entries
...
to be used when traversing filehead.
2008-01-05 23:53:21 +00:00
798130646c
Add missing include
2008-01-05 22:50:12 +00:00
bb81964965
allocate dedicated stacks for NMI. PR/37662.
2008-01-05 22:48:31 +00:00
d5681ba54e
Add missing include.
2008-01-05 22:45:22 +00:00
55ac8a439e
remove a stale comment.
2008-01-05 21:59:33 +00:00
f98ad43750
remove no longer necessary cpu_maxproc.
2008-01-05 21:52:01 +00:00
ad