Commit Graph

79216 Commits

Author SHA1 Message Date
thorpej
1bf3f87777 Fix a problem with kcopy()/copyin()/copyout() on MP systems
that became apparent when UBC was added: store a pointer to
the process itself, not a pointer to ci->ci_curproc.

This gets us back to where we were before UBC went in: MP
kernels get to single-user mode, and can run processes on
both CPUs, but things go south when we try to come into
multi-user mode.
2001-04-18 00:38:10 +00:00
bjh21
6fe345e833 "asc" doesn't exist on arm26 any more. 2001-04-17 23:51:07 +00:00
kleink
2c30ea4c58 Add missing quotes to test -z argument; from Tim Preston in PR bin/12687. 2001-04-17 23:25:38 +00:00
thorpej
a51a1d8cdd - Get rid of the prot bits in the mem_clusters[] array when
reserving RAM in the bus_mem extent map.  Problem pointed
  out by Artur Grabowski.
- Work around a slightly annoying bit of behavior exhibited by
  the UP1000 firmware.  The UP1000 firmware reports the space
  consumed by the "ISA hole" in the same MDDT entry as two
  chunks of RAM (on either side of the hole) used by the PALcode,
  all as one "reserved for PALcode" chunk.  We must take this
  into account when reserving RAM in the bus_mem extent map.
2001-04-17 21:52:00 +00:00
pk
4485b7f720 pmap_enter():
* turn ad-hoc `pmap resources allocated during sleep' into a panic condition
  (this should all be prevented by higher level VM locking).
* arrange for ENOMEM to be returned if PMAP_CANFAIL is on, and malloc()
  and/or pool_get() fail.
2001-04-17 20:14:46 +00:00
kleink
08c6392504 Need <limits.h> in _CTYPE_PRIVATE environments. 2001-04-17 20:12:31 +00:00
kleink
7aa4bf8f8f Replace (sizeof (char) << 3) with a less silly-looking CHAR_BIT;
from Peter Seebach in PR standards/9333.
2001-04-17 20:09:50 +00:00
thorpej
13e5178e3d Don't hold vp->v_interlock when calling vcount(); vcount() calls
vgone(), which may sleep.
2001-04-17 18:49:26 +00:00
thorpej
513dcb38be Descend into man8.macppc. 2001-04-17 17:51:39 +00:00
toshii
58165a53a2 Make masked interrupts don't bring CPU out of idle mode. 2001-04-17 17:12:40 +00:00
toshii
cfa894dce0 Use idle mode in process waiting idle loop. 2001-04-17 16:14:57 +00:00
toshii
de8e20334a Map a page needed for entering idle mode.
Also, use zero memory at physaddr 0xe0000000 to clean D cache.
2001-04-17 16:10:47 +00:00
toshii
ffe7f4da8a Fix a bug crept in the previous revision. Call dosoftints() after
reenabling interrupts.
2001-04-17 16:07:13 +00:00
toshii
a78b59337f Overhaul interrupt handlers.
- Set the match register before calling hardclock() to prevent
 clockintr lossage.
- Better handling of match register overrun case.
2001-04-17 15:58:51 +00:00
toshii
0c1210c6b3 Make the debugging hack needed for DEBUG_BEFOREMMU functional again. 2001-04-17 15:53:04 +00:00
toshii
5373b5ab86 Special care must be taken in order to run statclock at 128Hz.
For now, change statclock rate to 64Hz for simplicity.

Thanks to Ross Harvey for pointing out this issue.
2001-04-17 15:49:02 +00:00
sato
cdc83120a7 fix SINGLE_VRIP_BASE condition. 2001-04-17 15:45:20 +00:00
toshii
af9740eeee Initialize platid in hpc_machdep(). 2001-04-17 15:33:39 +00:00
takemura
a8106b2b59 Fix typo. kbdparam.h -> kdbparam.h.
RP12685
2001-04-17 15:19:41 +00:00
jdolecek
53a980f666 Sync set lists. From Andrew Brown. 2001-04-17 14:43:07 +00:00
ad
a7a9126e0e Regen. 2001-04-17 14:30:01 +00:00
ad
6a59be7d6b Add Bt760 Ethernet. 2001-04-17 14:29:47 +00:00
kleink
40c8d08347 Oops, intptrs are always long integers. 2001-04-17 13:46:02 +00:00
ad
fae451430d Nits. 2001-04-17 13:32:39 +00:00
ad
bd8e0b62fb Let '\377' be whitespace. 2001-04-17 13:31:00 +00:00
itojun
b7ab24621c do not attempt to pass null pointer to krb5 library. PR 12683 2001-04-17 12:27:37 +00:00
sato
97683d27ae - definition for vr cpu identify.
- definition for vr cpu group identify.
2001-04-17 11:13:16 +00:00
hubertf
dcdc94af10 Don't warn about @src being deprecated any longer - we add it per default
to each binary pkg again (via pkg_create -L) to assist DESTDIR builds.

Noted by Dan McMahill in private mail.
2001-04-17 10:42:11 +00:00
hubertf
152216c60d on pkgdb_remove, don't retain old errors - clear errno before.
This prevents some warnings from being printed over and over again
after one error was found.
2001-04-17 10:33:35 +00:00
reinoud
23f20b8af6 Update DEVNAMES for the new wscons devices for arm32 wich i had forgotten. 2001-04-17 10:33:21 +00:00
augustss
6497b0a623 Regen. 2001-04-17 08:38:23 +00:00
augustss
7dfc3d8076 Add a Xirlink camera. From PR 11570. 2001-04-17 08:38:01 +00:00
dbj
6456e9d8a9 increase maxsegs of esp main dma map by one to avoid error
tweak esp glue to read xfer_len from dma map
add some debugging checks for dma driver invariants
read DD_LIMIT instead of DD_SAVED_LIMIT on scsi dma shutdowns
2001-04-17 03:42:24 +00:00
lukem
7ee956fdcc use own code instead of bother with glob() to do ~ expansion in pathname;
there's no need to support glob wildcards in this case when it's not expanded
here in the non-~ case
2001-04-17 01:37:04 +00:00
ross
24ba541b4e Follow up with style commit to put the isdn nodes in major number order. 2001-04-17 01:30:30 +00:00
ross
02e8f1f2e5 First do minimal commit to correct botch in isdn major numbers. 2001-04-17 01:28:55 +00:00
lukem
6443de4b8e limit the number of matches in a ~ pathname glob, and complain if more
than one path is matched.
2001-04-17 00:59:58 +00:00
augustss
264c404593 Clean up formatting a little.
Split protocol variable into wire and command protocols.
2001-04-17 00:50:13 +00:00
augustss
f14db62d03 Add a DIAGNOSTIC. 2001-04-17 00:05:33 +00:00
thorpej
5b35dc8136 When unmounting a file system, acquire the syncer_lock before
vfs_busy'ing just before the dounmount() call.  This is to avoid
sleeping with the mountlist_slock held -- but we must acquire
syncer_lock before vfs_busy because the syncer itself uses
syncer_lock -> vfs_busy locking order.
2001-04-16 22:41:09 +00:00
dbj
fa84b1d943 make sure that dm_segs is last field of next68k_bus_dmamap,
so that variable length allocations work
2001-04-16 21:36:58 +00:00
mbw
3b5568e369 By popular demand.
I've organized the topics by Open Firmware version, since many users found
the old scheme way too confusing (clause out Open Firmware version-specific
content).

It's still really confusing to read (particularly the .txt version which
doesn't adequately distinguish headings from the body)

Also added a bit more info on partitioning vs. OF version.
2001-04-16 20:42:43 +00:00
bjh21
d71d51409d "acemidi" podulebus device. 2001-04-16 20:14:04 +00:00
chris
7b97ef04ca Enable clockswitching on sa110 (with it disabled the core clock runs at the speed of the memory bus) 2001-04-16 18:09:56 +00:00
bjh21
074987f4a3 Ensure that we have some idea what a label_t is, even in userland. 2001-04-16 17:30:32 +00:00
thorpej
18c1bea288 Oops, properly mark a memory operand as both an input and an
output as necessary.
2001-04-16 17:05:22 +00:00
itojun
8799a9c64b give a default value to net.inet.ip.maxfragpackets, to protect us from
"lots of fragmented packets" DoS attack.

the current default value is derived from ipv6 counterpart, which is
a magical value "200".  it should be enough for normal systems, not sure
if it is enough when you take hundreds of thousands of tcp connections on
your system.  if you have proposal for a better value with concrete reasons,
let me know.
2001-04-16 17:03:33 +00:00
chs
88c05e77fe reads at or after EOF should "succeed". 2001-04-16 14:37:43 +00:00
bjh21
020a050984 Randomize the statclock period properly. We use the same algorithm as the
sparc port.
2001-04-16 14:12:38 +00:00
dbj
b2bff5fc7f add receive ethernet bpf tap
keep receive ethernet crc and set M_HASFCS
change bus_dma MD fields to keep xfer_len for entire map
  instead of per segment
turn off automatic dma restarts in preparation for changing
  scsi driver to handle them.
add dma debugging routines to keep history of dma states
tweak checking for unusual dma limit register
2001-04-16 14:12:12 +00:00