]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
25 years agoFixed printf format errors on alpha.
dt [Sat, 24 Apr 1999 18:50:48 +0000 (18:50 +0000)]
Fixed printf format errors on alpha.

25 years agoAdd missing strings.
phk [Sat, 24 Apr 1999 18:28:24 +0000 (18:28 +0000)]
Add missing strings.

PR: 11285
Submitted by: Chris Costello <chris@calldei.com>
Reviewed by: phk

25 years agotypo in comment
phk [Sat, 24 Apr 1999 18:26:54 +0000 (18:26 +0000)]
typo in comment

PR: 11285
Submitted by: Chris Costello <chris@calldei.com>
Reviewed by: phk

25 years agoso_linger is in seconds, not in 1/HZ
ache [Sat, 24 Apr 1999 18:25:35 +0000 (18:25 +0000)]
so_linger is in seconds, not in 1/HZ

PR: 11252
Submitted by: Martin Kammerhofer <dada@sbox.tu-graz.ac.at>

25 years agoChanged the type of id_port from short into int to avoid wrong
kato [Sat, 24 Apr 1999 18:24:43 +0000 (18:24 +0000)]
Changed the type of id_port from short into int to avoid wrong
conversion from short to unsigned long which is an argument of
bus_alloc_resource.  Since the value -1 is used to indicate no port
reousece, id_port need to be signed (suggested by Doug Rabson and
Peter Wemm.)

25 years agoLite2 bugfixes merge:
ache [Sat, 24 Apr 1999 18:22:34 +0000 (18:22 +0000)]
Lite2 bugfixes merge:
so_linger is in seconds, not in 1/HZ
range checking in SO_*TIMEO was wrong

PR: 11252

25 years agoMinor nit. It is probably still incorrect, but at least don't
dcs [Sat, 24 Apr 1999 18:05:45 +0000 (18:05 +0000)]
Minor nit. It is probably still incorrect, but at least don't
claim to use i386-specific files on alpha...

25 years agoFix a braino in the v_id wraparound code. Give more (current) details
phk [Sat, 24 Apr 1999 17:58:14 +0000 (17:58 +0000)]
Fix a braino in the v_id wraparound code.  Give more (current) details
in comment.

PR: 11307
Spotted by: Ville-Pertti Keinonen <will@iki.fi>

25 years agoMove loader.conf.5, while it is still in it's first revision.
dcs [Sat, 24 Apr 1999 17:49:47 +0000 (17:49 +0000)]
Move loader.conf.5, while it is still in it's first revision.

25 years agoremove uid switching before login_getpwclass, now done inside libutil
ache [Sat, 24 Apr 1999 17:26:32 +0000 (17:26 +0000)]
remove uid switching before login_getpwclass, now done inside libutil
add gid switching before chdir and comment why it needed

25 years agoAdd loader.4th man page (worst man page I ever wrote -- reviews
dcs [Sat, 24 Apr 1999 17:25:35 +0000 (17:25 +0000)]
Add loader.4th man page (worst man page I ever wrote -- reviews
are welcomed). Correct a nasty bug in loader.4th before anyone
notices, add a couple of features.

25 years agooops, add pwd!=NULL check to previous fix
ache [Sat, 24 Apr 1999 17:17:42 +0000 (17:17 +0000)]
oops, add pwd!=NULL check to previous fix

25 years agoSwitch to user UID/GID before checking/reading its ~/.login_conf
ache [Sat, 24 Apr 1999 17:01:58 +0000 (17:01 +0000)]
Switch to user UID/GID before checking/reading its ~/.login_conf
- some NFSes have root read access disabled

25 years agoFix spelling of '#if 0'.
dt [Sat, 24 Apr 1999 16:39:52 +0000 (16:39 +0000)]
Fix spelling of '#if 0'.

25 years agoadd MLINKS for two functions used from login_auth.c
ache [Sat, 24 Apr 1999 16:30:30 +0000 (16:30 +0000)]
add MLINKS for two functions used from login_auth.c
comment out unused functions from login_auth.3

25 years agoDrop tty/net/bio/cam interrupt class labels, it's meaningless here now.
peter [Sat, 24 Apr 1999 16:07:51 +0000 (16:07 +0000)]
Drop tty/net/bio/cam interrupt class labels, it's meaningless here now.

25 years agoDrop the tty/net/bio/cam interrupt class labels, they are meaninless here
peter [Sat, 24 Apr 1999 16:04:16 +0000 (16:04 +0000)]
Drop the tty/net/bio/cam interrupt class labels, they are meaninless here
now.

25 years agoUse pointer arithmetic as appropriate.
dt [Sat, 24 Apr 1999 13:23:48 +0000 (13:23 +0000)]
Use pointer arithmetic as appropriate.

25 years agopostpone the sending of IGMP LEAVE msg to after deleting the
luigi [Sat, 24 Apr 1999 12:28:51 +0000 (12:28 +0000)]
postpone the sending of IGMP LEAVE msg to after deleting the
mc address from the address list. The latter operation on some
hardware resets the card, potentially canceling the pending LEAVE
pkt.

25 years agoFixed printf format errors on alpha.
dt [Sat, 24 Apr 1999 11:29:48 +0000 (11:29 +0000)]
Fixed printf format errors on alpha.

25 years agoUse pointer arithmetic to do pointer arithmetic.
dt [Sat, 24 Apr 1999 11:25:01 +0000 (11:25 +0000)]
Use pointer arithmetic to do pointer arithmetic.

25 years agoModify the non-i386 version of disable_intr() and enable_intr() so that they
dt [Sat, 24 Apr 1999 10:41:21 +0000 (10:41 +0000)]
Modify the non-i386 version of disable_intr() and enable_intr() so that they
don't produce a warning on every use.

25 years agoRevert part of the previous commit. Keep the example that shows how the
nik [Sat, 24 Apr 1999 10:08:09 +0000 (10:08 +0000)]
Revert part of the previous commit.  Keep the example that shows how the
"-h" flag is used, but use "chown" in the example instead of "file".

Prompted by: bde

25 years agoAdd an option to disable responses to SAP_GETNEAREST_SERVER requests.
jhay [Sat, 24 Apr 1999 09:18:49 +0000 (09:18 +0000)]
Add an option to disable responses to SAP_GETNEAREST_SERVER requests.

Submitted by: Boris Popov <bp@butya.kz>

25 years agoBack out parts of rev 1.4; making isa.c preserve the hints across a
peter [Sat, 24 Apr 1999 07:04:51 +0000 (07:04 +0000)]
Back out parts of rev 1.4; making isa.c preserve the hints across a
release made it unneeded.

25 years agoRevert part of 1.9; we don't need to reset the port after release now that
peter [Sat, 24 Apr 1999 06:48:27 +0000 (06:48 +0000)]
Revert part of 1.9; we don't need to reset the port after release now that
the isa bus doesn't clear the hints at that point.

25 years agoDon't clear the hints on release, just the resource containers.
peter [Sat, 24 Apr 1999 06:47:24 +0000 (06:47 +0000)]
Don't clear the hints on release, just the resource containers.

25 years agoTypo, use SYS_RES_IRQ when releasing the irq resource, not SYS_RES_IOPORT.
peter [Sat, 24 Apr 1999 06:46:10 +0000 (06:46 +0000)]
Typo, use SYS_RES_IRQ when releasing the irq resource, not SYS_RES_IOPORT.

25 years agoOptimize better for space.
kato [Sat, 24 Apr 1999 04:29:13 +0000 (04:29 +0000)]
Optimize better for space.

Submitted by: Takahashi Yoshihiro <nyan@wyvern.cc.kogakuin.ac.jp>

25 years agoCommented out adv_isa.c.
kato [Sat, 24 Apr 1999 04:22:57 +0000 (04:22 +0000)]
Commented out adv_isa.c.

25 years ago- PC98 doesn't support 16bits bus I/F of adv deriver.
kato [Sat, 24 Apr 1999 04:21:42 +0000 (04:21 +0000)]
- PC98 doesn't support 16bits bus I/F of adv deriver.
- PC98 has the bs driver.

Submitted by: Takahashi Yoshihiro <nyan@dd.catv.ne.jp>

25 years ago1MB is not 1024 * 1024 * 1024 but 1024 * 1024.
kato [Sat, 24 Apr 1999 04:16:22 +0000 (04:16 +0000)]
1MB is not 1024 * 1024 * 1024 but 1024 * 1024.

25 years agoRemove some left-over stuff from NetBSD that we don't need. This
jdp [Sat, 24 Apr 1999 02:07:17 +0000 (02:07 +0000)]
Remove some left-over stuff from NetBSD that we don't need.  This
eliminates the need to include the dynamic linker's private header
file, as well as two other headers from <sys>.

25 years agoRemove the registration stuff; we're redoing this and the emailed
jkh [Sat, 24 Apr 1999 01:53:55 +0000 (01:53 +0000)]
Remove the registration stuff; we're redoing this and the emailed
registrations have a high failure rate for various reasons.

25 years agoTypo of `same_ports' directive.
sada [Sat, 24 Apr 1999 01:42:44 +0000 (01:42 +0000)]
Typo of `same_ports' directive.
Submitted by: Masaki Nohtomi <noutomi@jbm-net.or.jp>

25 years agoahc_pci.c:
gibbs [Fri, 23 Apr 1999 23:30:22 +0000 (23:30 +0000)]
ahc_pci.c:
Convert to new bus and bus dma.
Use latest PCI API.

bt_pci.c:
Fix a few bugs in how resourses are released left over from
when this driver was converted to new bus.

25 years agostaticize.
gibbs [Fri, 23 Apr 1999 23:29:23 +0000 (23:29 +0000)]
staticize.

25 years agoahb.c:
gibbs [Fri, 23 Apr 1999 23:29:00 +0000 (23:29 +0000)]
ahb.c:
Synchronize malloc types for our softc.

ahc_eisa.c:
Sync to aic7xxx driver bus dma and new bus changes.

25 years agobt.c:
gibbs [Fri, 23 Apr 1999 23:28:20 +0000 (23:28 +0000)]
bt.c:
Simplify some inline functions.
btreg.h:
Correct a comment.

25 years agoComplete conversion to bus dma. This driver now works on the alpha.
gibbs [Fri, 23 Apr 1999 23:27:31 +0000 (23:27 +0000)]
Complete conversion to bus dma.  This driver now works on the alpha.

aicasm_symbol.c:
Correct an unaligned access problem.  You can't rely on DB to
store your data in an aligned fashion.

25 years agoKill a compiler warning.
gibbs [Fri, 23 Apr 1999 23:26:10 +0000 (23:26 +0000)]
Kill a compiler warning.

25 years agoRelease bus entries in the EDT when a path is released.
gibbs [Fri, 23 Apr 1999 23:25:48 +0000 (23:25 +0000)]
Release bus entries in the EDT when a path is released.
Null out path entries when the path is released to catch bugs.
Free our configuration hook when we're done with it.

25 years agoMake the register_intr() glue actually have a chance of working...
peter [Fri, 23 Apr 1999 21:01:19 +0000 (21:01 +0000)]
Make the register_intr() glue actually have a chance of working...

25 years agos/lkm(4)/kld(4)/
peter [Fri, 23 Apr 1999 20:53:34 +0000 (20:53 +0000)]
s/lkm(4)/kld(4)/

25 years agoMake pmap_collect() an official pmap interface.
dt [Fri, 23 Apr 1999 20:29:58 +0000 (20:29 +0000)]
Make pmap_collect() an official pmap interface.

25 years agoMoved cpu_set_fork_handler's prototype from <machine/cpu.h> to <sys/proc.h>.
dt [Fri, 23 Apr 1999 20:22:44 +0000 (20:22 +0000)]
Moved cpu_set_fork_handler's prototype from <machine/cpu.h> to <sys/proc.h>.

Suggested by: bde

25 years agoFixed several (not all) warnings.
dt [Fri, 23 Apr 1999 19:53:38 +0000 (19:53 +0000)]
Fixed several (not all) warnings.

25 years agolocal_cron -> local_periodic
ghelmer [Fri, 23 Apr 1999 18:26:55 +0000 (18:26 +0000)]
local_cron -> local_periodic

PR: docs/11253

25 years agoReference kld(4).
ghelmer [Fri, 23 Apr 1999 16:12:44 +0000 (16:12 +0000)]
Reference kld(4).

Submitted by: Nathan Ahlstrom <nrahlstr@winternet.com>

25 years agoSpelling police
brian [Fri, 23 Apr 1999 13:45:50 +0000 (13:45 +0000)]
Spelling police

25 years agoUpdate VM86 comment - it's used for VESA too.
peter [Fri, 23 Apr 1999 06:30:10 +0000 (06:30 +0000)]
Update VM86 comment - it's used for VESA too.

PR: 7976
Submitted by: Stefan Eggers <seggers@semyam.dinoco.de>

25 years agofix -n with 16-chars hostnames, modified patch from PR
ache [Thu, 22 Apr 1999 23:40:56 +0000 (23:40 +0000)]
fix -n with 16-chars hostnames, modified patch from PR

PR: 11286
Submitted by:  Hans Huebner <hans@chaosradio.berlin.ccc.de>

25 years agoUpdate #include statements to reflect the new location of "sioreg.h".
jdp [Thu, 22 Apr 1999 21:02:44 +0000 (21:02 +0000)]
Update #include statements to reflect the new location of "sioreg.h".
With these changes plus the egcs fix I committed a few minutes ago,
"make -DWANT_AOUT world" works again.  Most likely, "make upgrade"
is fixed too, though I haven't tested that.

25 years agoSorry didn't get this in the last commit: MACHINE_ARCH changes courtesy
obrien [Thu, 22 Apr 1999 20:12:21 +0000 (20:12 +0000)]
Sorry didn't get this in the last commit: MACHINE_ARCH changes courtesy
of Warner Losh <imp@village.org>.  (he's got plans for FreeBSD mips :-))

25 years agoUse `GNU_ARCH' in most places where `MACHINE_ARCH' was used. Building for
obrien [Thu, 22 Apr 1999 19:35:16 +0000 (19:35 +0000)]
Use `GNU_ARCH' in most places where `MACHINE_ARCH' was used.  Building for
mips has some particularlies.

25 years agoAdd libf2c.so.2 for our FORTRAN friends.
obrien [Thu, 22 Apr 1999 19:10:48 +0000 (19:10 +0000)]
Add libf2c.so.2 for our FORTRAN friends.

25 years agoUse new `FBSD_SWITCH_TAKES_ARG' and `FBSD_WORD_SWITCH_TAKES_ARG'.
obrien [Thu, 22 Apr 1999 17:45:24 +0000 (17:45 +0000)]
Use new `FBSD_SWITCH_TAKES_ARG' and `FBSD_WORD_SWITCH_TAKES_ARG'.

25 years agoMove `SWITCH_TAKES_ARG' to `FBSD_SWITCH_TAKES_ARG' and `WORD_SWITCH_TAKES_ARG'
obrien [Thu, 22 Apr 1999 17:45:01 +0000 (17:45 +0000)]
Move `SWITCH_TAKES_ARG' to `FBSD_SWITCH_TAKES_ARG' and `WORD_SWITCH_TAKES_ARG'
to `FBSD_WORD_SWITCH_TAKES_ARG' which allows architectures to tweak these
values if needed.

25 years agoDon't build with -g. The a.out egcs can't handle it at the moment.
jdp [Thu, 22 Apr 1999 15:47:21 +0000 (15:47 +0000)]
Don't build with -g.  The a.out egcs can't handle it at the moment.
This seems to have snuck back into the Makefile in an unrelated
commit (rev. 1.8).

25 years agoFix so that this driver works again when compiled with EGCS (which
davidn [Thu, 22 Apr 1999 15:19:37 +0000 (15:19 +0000)]
Fix so that this driver works again when compiled with EGCS (which
apparently has a much better optimiser, requiring 'volatile' memory
to be declared as such).

25 years agoFix the display of the "nice" value of processes like ntpd that use the
jhay [Thu, 22 Apr 1999 14:34:53 +0000 (14:34 +0000)]
Fix the display of the "nice" value of processes like ntpd that use the
posix sched_setscheduler() to set their priority.

Noticed by: Mark Allwright <mallwri@orion.didata.co.za>

25 years agoReturn the port size from the probe.
peter [Thu, 22 Apr 1999 13:10:43 +0000 (13:10 +0000)]
Return the port size from the probe.
"ppc0 at port 0x378 irq 7 drq 3 on isa0" becomes
"ppc0 at port 0x378-0x37f irq 7 drq 3 on isa0"

25 years agoFix the promise_intr function, it should use 'lun' not 'unit'
sos [Thu, 22 Apr 1999 08:07:44 +0000 (08:07 +0000)]
Fix the promise_intr function, it should use 'lun' not 'unit'
to get the offset into ata_devices.

25 years agoFor mips assembler, we need to define TARGET_BYTES_BIG_ENDIAN to be
imp [Thu, 22 Apr 1999 07:19:20 +0000 (07:19 +0000)]
For mips assembler, we need to define TARGET_BYTES_BIG_ENDIAN to be
zero when building for little endian machines.

Correct the target names for mips.  We just use the generic targets
for mips elf, so the mipse[lb]-unknown-freebsd emulation types don't
exist.

25 years agoFix all the mipseb Makefiles. They were broken when I checked them in
imp [Thu, 22 Apr 1999 07:16:11 +0000 (07:16 +0000)]
Fix all the mipseb Makefiles.  They were broken when I checked them in
before.

Added SYS.h for mipseb and mipsel.

I now get part way through building libc in the cross environment that
I have (along with pending mipse[bl] changes to the intree egcs) with
these changes.

25 years agoFix poor heading format.
grog [Thu, 22 Apr 1999 04:05:56 +0000 (04:05 +0000)]
Fix poor heading format.

Submitted-by: Matthew Fuller <fullermd@over-yonder.net>
PR:        docs/11271

25 years agoBack out my change from 6 April PDT that added a new dlversion()
jdp [Thu, 22 Apr 1999 01:54:38 +0000 (01:54 +0000)]
Back out my change from 6 April PDT that added a new dlversion()
function.  It was an ill-considered feature.  It didn't solve the
problem I wanted it to solve.   And it added Yet Another Version
Number that would have to be maintained at every release point.
I'm nuking it now before anybody grows too fond of it.

25 years agoHave config/i386/freebsd.h directly include att.h. I have no idea why stock
obrien [Thu, 22 Apr 1999 00:20:36 +0000 (00:20 +0000)]
Have config/i386/freebsd.h directly include att.h.  I have no idea why stock
EGCS doesn't just do this.  None of the rest of our other platforms will need
this file.

25 years agoTypo fix in diagnostic: -alldir --> -alldirs
alex [Wed, 21 Apr 1999 22:42:36 +0000 (22:42 +0000)]
Typo fix in diagnostic: -alldir --> -alldirs

PR: 11049
Submitted by: Gerhard Gonter <gonter@whisky.wu-wien.ac.at>

25 years agoWork around an egcs optimizer bug (i386). This should fix the active ftp
luoqi [Wed, 21 Apr 1999 21:28:01 +0000 (21:28 +0000)]
Work around an egcs optimizer bug (i386). This should fix the active ftp
hang problem. A bug report has been sent to cygnus.

25 years agoMake usage() reflect reality (add the -aout and -elf options).
billf [Wed, 21 Apr 1999 20:06:42 +0000 (20:06 +0000)]
Make usage() reflect reality (add the -aout and -elf options).

OK'd by: bde

25 years agoUpdate VERSREQ.
dfr [Wed, 21 Apr 1999 19:50:27 +0000 (19:50 +0000)]
Update VERSREQ.

25 years agoFix my breakage of BRIDGE compiling option without IPFIREWALL..
peter [Wed, 21 Apr 1999 18:23:00 +0000 (18:23 +0000)]
Fix my breakage of BRIDGE compiling option without IPFIREWALL..
(Note that if you have bridge compiled in and then kldload ipfw, bridge
 won't automatically use it - knowledge of ipfw/dummynet is compiled in)

25 years agoRemove the old libf2c. libg2c is now serving us well.
obrien [Wed, 21 Apr 1999 16:43:03 +0000 (16:43 +0000)]
Remove the old libf2c.  libg2c is now serving us well.

25 years agoThe function msgrcv() could copy larger data than it should do
sada [Wed, 21 Apr 1999 13:30:01 +0000 (13:30 +0000)]
The function msgrcv() could copy larger data than it should do
under some circumstances.
PR: kern/10765
Submitted by: Yasuhito FUTATSUKI <futatuki@fureai.or.jp>

25 years agoSync with sys/i386/isa/isa_dma.c revision 1.2.
kato [Wed, 21 Apr 1999 12:17:00 +0000 (12:17 +0000)]
Sync with sys/i386/isa/isa_dma.c revision 1.2.

25 years agoSync with sys/i386/isa/clock.c revision 1.131.
kato [Wed, 21 Apr 1999 12:14:37 +0000 (12:14 +0000)]
Sync with sys/i386/isa/clock.c revision 1.131.

25 years agoForgot one.
n_hibma [Wed, 21 Apr 1999 12:08:07 +0000 (12:08 +0000)]
Forgot one.

Original log message:

Remove feature creep: STAILQ_REMOVE_HEAD_UNTIL added it for convenience
but we can do without it.

Obtained from: Peter Wemm

25 years agoNext approach to make loops in interactive interruptable.
cracauer [Wed, 21 Apr 1999 11:52:39 +0000 (11:52 +0000)]
Next approach to make loops in interactive interruptable.

PR: bin/9173

25 years agoUse ${.CURDIR} in `exists()' constructs.
jkoshy [Wed, 21 Apr 1999 11:09:40 +0000 (11:09 +0000)]
Use ${.CURDIR} in `exists()' constructs.

PR: misc/4395
Submitted by: J Wunsch <j@ida.interface-business.de>

25 years agoMerge a diff that Soren sent me to resolve some lun / unit problems.
peter [Wed, 21 Apr 1999 10:58:07 +0000 (10:58 +0000)]
Merge a diff that Soren sent me to resolve some lun / unit problems.
While here, also fix my additions to use naming that's more consistant
with Sorens.  (ie: s/softc/scp/)

25 years agopmap_ts_referenced(): set fault on {read|execute|write} bits on the page on,
dt [Wed, 21 Apr 1999 10:51:04 +0000 (10:51 +0000)]
pmap_ts_referenced(): set fault on {read|execute|write} bits on the page on,
so we will know when the page is accessed again.

(Also, make it return an integer, not "boolean" TRUE/FALSE).

25 years agoChange type of a variable from u_int to size_t, so that pointer to it may be
dt [Wed, 21 Apr 1999 09:41:07 +0000 (09:41 +0000)]
Change type of a variable from u_int to size_t, so that pointer to it may be
used as a last argument to copyinstr().

25 years agoint -> size_t
brian [Wed, 21 Apr 1999 08:13:25 +0000 (08:13 +0000)]
int -> size_t

25 years agoSplit the recorded chap challenge into two - one for the
brian [Wed, 21 Apr 1999 08:03:51 +0000 (08:03 +0000)]
Split the recorded chap challenge into two - one for the
receiver and one for the sender.  This allows two simultaneous
chap conversations - something that I *thought* I was already
doing on a daily basis myself until the existence of the
problem was
Beaten into me by: sos

25 years agoIf ioctl TIOCMGET fails, continue. The device may be
brian [Wed, 21 Apr 1999 08:03:35 +0000 (08:03 +0000)]
If ioctl TIOCMGET fails, continue.  The device may be
a pseudo device created by an interactive version of
rlogin/telnet/ssh & friends

25 years agoDEVICE_SUSPEND was always returning ENXIO for reasons unknown. For
imp [Wed, 21 Apr 1999 07:57:55 +0000 (07:57 +0000)]
DEVICE_SUSPEND was always returning ENXIO for reasons unknown.  For
now we noisily ignore this (and all errors).  DEVICE_SUSPEND should be
corrected, but I wanted to unbreak suspend until that happens.

25 years agoTARGET was being defined wrong, causing ld to always fail. It is the
imp [Wed, 21 Apr 1999 07:46:06 +0000 (07:46 +0000)]
TARGET was being defined wrong, causing ld to always fail.  It is the
name of the bfd target, not the gnu-standard target name.  Corrected
to be elf32-{big,little}mips from mipse[bl]-unknown-freebsd.

DEFAULT_EMULATION was bogusly defined, causing ld to always fail (this
was masked by the TARGET bogosity).  Define correctly as elf32bmip and
elf32lmip.  Mips doesn't follow the same conventions as i386 and alpha
do in this area.

ld now appears to work correctly for the uncommitted mips changes to
egcs.

25 years agoAdd Id.
obrien [Wed, 21 Apr 1999 07:42:02 +0000 (07:42 +0000)]
Add Id.

25 years agooops, SMP was missing includes for a typedef.
peter [Wed, 21 Apr 1999 07:41:40 +0000 (07:41 +0000)]
oops, SMP was missing includes for a typedef.

25 years agoStage 1 of a cleanup of the i386 interrupt registration mechanism.
peter [Wed, 21 Apr 1999 07:26:30 +0000 (07:26 +0000)]
Stage 1 of a cleanup of the i386 interrupt registration mechanism.
Interrupts under the new scheme are managed by the i386 nexus with the
awareness of the resource manager.  There is further room for optimizing
the interfaces still.  All the users of register_intr()/intr_create()
should be gone, with the exception of pcic and i386/isa/clock.c.

25 years agoAdd compat3x to the mix.
obrien [Wed, 21 Apr 1999 07:22:37 +0000 (07:22 +0000)]
Add compat3x to the mix.

Reviewed by: jkh

25 years agoAs I understand it, these register_intr()'s shouldn't be here. The isa
peter [Wed, 21 Apr 1999 07:02:19 +0000 (07:02 +0000)]
As I understand it, these register_intr()'s shouldn't be here.  The isa
driver attaches the interupt itself.

25 years agoAdd entry for newbus. It is rather long, but lots of nits can shoot
imp [Wed, 21 Apr 1999 06:57:35 +0000 (06:57 +0000)]
Add entry for newbus.  It is rather long, but lots of nits can shoot
people with this upgrade.  Lemme know what you think, or any
corrections that should be made to it.

Corrected a few typos based on submissions from various people.

Add note that make aout-to-elf is currently broken and what to do to
work around it.  This is implied by make -DWANT_AOUT being broken, but
that isn't obvious to many people.

25 years agoAddress several problems in vn_read and vn_write:
alc [Wed, 21 Apr 1999 05:56:45 +0000 (05:56 +0000)]
Address several problems in vn_read and vn_write:

1. Make read-ahead work for pread and aio_read.

2. Fix one place where a comparison of uio_offset with -1
   wasn't updated to use FOF_OFFSET.

3. Honor O_APPEND in the FOF_OFFSET case.

In addition, use the variable name "ioflag" in both vn_read and
vn_write to avoid possible confusion between the variable "flag"
and the parameter "flags".

Submitted by: Bruce Evans <bde@zeta.org.au> and me

25 years agoSpelling error fixed.
nsayer [Wed, 21 Apr 1999 05:40:12 +0000 (05:40 +0000)]
Spelling error fixed.

25 years agoAfter relocating the main program, but before calling any of the
jdp [Wed, 21 Apr 1999 04:06:57 +0000 (04:06 +0000)]
After relocating the main program, but before calling any of the
_init() functions, initialize the global variables "__progname" and
"environ".  This makes it possible for the _init() functions to call
things like getenv() and err().

25 years agoAdded consts to cpu_set_fork_handler prototype. (Follow i386 version.)
dt [Tue, 20 Apr 1999 22:53:54 +0000 (22:53 +0000)]
Added consts to cpu_set_fork_handler prototype. (Follow i386 version.)

25 years agoRemove feature creep: STAILQ_REMOVE_HEAD_UNTIL added it for convenience
n_hibma [Tue, 20 Apr 1999 22:37:18 +0000 (22:37 +0000)]
Remove feature creep: STAILQ_REMOVE_HEAD_UNTIL added it for convenience
but we can do without it.

25 years agoalpha/include/param.h: #define NCPUS 1
dt [Tue, 20 Apr 1999 22:20:57 +0000 (22:20 +0000)]
alpha/include/param.h: #define NCPUS 1
alpha/include/lock.h: remove nop simplelock macros, which are defined
in <sys/lock.h> if NCPUS == 1.

As a result, NULL_SIMPLELOCK is defined, and a few warnings removed.