7401 Commits

Author SHA1 Message Date
peter
b99f0b0dfb Add missing $Id$
Please note, I have reconstructed the RCS file for this file so that it
has a CSRG vendor branch again.
1997-03-25 17:09:18 +00:00
peter
57b19c0bca Replace original rev 1.3; Author: bde; Date: 1997/02/25 17:24:43;
Fix counting of simplelocks in SIMPLELOCK_DEBUG
Fix style regression
1997-03-25 16:38:01 +00:00
peter
aef498c3be Replace original rev 1.2; Author: mpp; Date: 1997/02/12 06:52:30
Add missing #include <sys/systm.h>
1997-03-25 16:36:35 +00:00
peter
458e027ff4 Replace original rev 1.2; Author: mpp; Date: 1997/02/12 06:52:47
- add missing prototype for lockmgr_printinfo()
1997-03-25 16:35:07 +00:00
peter
65d2feb26c Replace original revision 1.1; Author dyson; Date: 1997/02/10 02:28:15
Changes from Lite2:
- DEBUG -> SIMPLELOCK_DEBUG
- cosmetic fixes
- bzero of lock at init time -> explicit init of members.
1997-03-25 16:32:46 +00:00
peter
b872711565 This commit was generated by cvs2svn to compensate for changes in r24269,
which included commits to RCS files with non-trunk default branches.
1997-03-25 16:27:20 +00:00
peter
7edf64c788 Import 4.4BSD-Lite2 onto CSRG branch 1997-03-25 16:27:20 +00:00
peter
2ea77a1af1 Import 4.4BSD-Lite2 onto CSRG branch 1997-03-25 16:27:20 +00:00
dg
61a2f5437b Made a couple of minor optimizations that improve performance of the
common case of the interrupt routine by about 20%.
1997-03-25 14:54:38 +00:00
kato
782ae9e282 Fixed reset port address for PC-9801-108 card. 1997-03-25 10:06:41 +00:00
kato
893a4745d3 Synchronize with sys/i386/isa/isa.c revision 1.79. 1997-03-25 10:04:50 +00:00
peter
5dc5a09bf1 Use the correct (relative to the implementation) ordering of args in
the VOP_LINK() calls, Closes PR#3064

Submitted by: bde
1997-03-25 05:13:40 +00:00
peter
6c63b328f9 The local fs interface does not allow link()/unlink() of directories,
do not allow a remote nfs client to cause local fs corruption either.
1997-03-25 05:08:28 +00:00
fsmp
1d259aacb1 "Louis A. Mamakos" <louie@TransSys.COM> new bt848 struct
Randall Hopper <rhh@ct.picker.com> GHUE/GBRIGHT bug
Louis Mamakos made a new bt848 struct, including massive changes to the entire
body of code, substituting array offsets with struct members.

Randall Hopper aadded fixes of BT848_GHUE & BT848_GBRIG.

I (fsmp):
  added polled hardware i2c routines,
  removed all existing software i2c routines.
  added  eeprom support.
1997-03-25 04:18:24 +00:00
ache
48fd2f3385 Use FreeBSD-style attach message 1997-03-25 04:16:25 +00:00
fsmp
9187e88957 added support for EEProms writing/reading. 1997-03-25 04:11:13 +00:00
ache
a2fd5d5125 Use port? for sbxvi0 since it is autoconfigured from sb0
Change sb0 irq from 7 to 5 since 5 is Creative default now.
1997-03-25 03:38:33 +00:00
ache
ada777da33 Copy io_base configuration address from sb config.
This change is cosmetique.
1997-03-25 03:32:31 +00:00
ache
5e588bcaa3 Replace more verbose "at <not configured>" with less verbose "at ?",
we don't need much attention here, because this diagnostic printed first
and then card will be configured.
1997-03-25 03:29:40 +00:00
ache
0b0ae0cc74 Follow config intention for iobase:
print "at <not configured>" for iobase == -1 (autodetect not happens)
and not print anything for iobase == -2 (none)
Old code treat this two special config numbers as big port numbers.
1997-03-25 03:13:05 +00:00
davidn
db2614fa00 Remove unneeded prototype (and warning) for stlintr(). 1997-03-24 21:38:51 +00:00
gibbs
e81c5f4477 Fix a fencepost error in ahc_find_scb that could cause us to wrongfully
find an SCB still down on the card that was paged out.  This only affects
error recovery.

Submitted by: Daniel M. Eischen <deischen@iworks.InterWorks.org>
1997-03-24 17:42:25 +00:00
bde
51c8cf75db Don't include <sys/ioctl.h> in the kernel. Stage 10 (final). Warn if
<sys/ioctl.h> is included in the kernel.  It still compiles.

This set of changes reduces the number of dependencies in LINT/.depend
from about 31000 to about 30000.  This should make LINT kernels
compile a whole 1% faster.  Further reductions to less than 20000
dependencies can easily be made.  E.g., 20 headers of nested spam
from <sys/param.h> for 500 object files gives 10000 dependencies;
less than half of these are real.
1997-03-24 12:48:46 +00:00
bde
3c9a4992bc Don't include <sys/ioctl.h> in the kernel. Stage 9: same changes
in pc98 as in isa.
1997-03-24 12:29:48 +00:00
bde
8f145614c6 Don't include <sys/ioctl.h> in the kernel. Stage 6: include
<sys/filio.h>, <sys/sockio.h> and <sys/ttycom.h> instead of
<sys/ioctl.h> in a couple of files.  This is still only 1/3
as spammish as <sys/ioctl.h> - 5 or 6 old tty ioctl headers
aren't needed.
1997-03-24 12:12:36 +00:00
bde
000a68175f Don't include <sys/ioctl.h> in the kernel. Stage 5: include
<sys/ioctl_compat.h> and sometimes <sys/filio.h> instead of
<sys/ioctl.h> in tty-related files.  <sys/ttycom.h> is still
usually imported bogusly via <sys/termios.h>.
1997-03-24 12:03:06 +00:00
bde
a0d9474a34 Don't include <sys/ioctl.h> in the kernel. Stage 4: include
<sys/ttycom.h> and sometimes <sys/filio.h> instead of <sys/ioctl.h>
in miscellaneous files.  Most of these files have nothing to do
with ttys but need to include <sys/ttycom.h> to get the definitions
of TIOC[SG]PGRP which are (ab)used to convert F[SG]ETOWN fcntls into
ioctls.
1997-03-24 11:52:29 +00:00
bde
f62eb55e94 Don't include <sys/ioctl.h> in the kernel. Stage 3: include
<sys/filio.h> instead of <sys/ioctl.h> in non-network non-tty files.
1997-03-24 11:37:53 +00:00
bde
e79c0a4ca0 Don't include <sys/ioctl.h> in the kernel. Stage 2: include
<sys/sockio.h> instead of <sys/ioctl.h> in network files.
1997-03-24 11:33:46 +00:00
bde
117209856b Don't include <sys/ioctl.h> in the kernel. Stage 1: don't include
it when it is not used.  In most cases, the reasons for including it
went away when the special ioctl headers became self-sufficient.
1997-03-24 11:25:10 +00:00
bde
2652de9da3 Don't default to (wrong) SCO definitions for _IOR() etc. if <sys/ioctl.h>
doesn't happen to be included before this header.

This header was missed in previous cleanups because it didn't include
<sys/ioctl.h> or <sys/ioccom.h>.  Clean it now:
- #include <sys/types.h> since it is necessary to make the header self-
  sufficient (there are a couple of u_char's).
- uniformized idempotency ifdef.  Copied the style in the 4.4Lite
  ioctl headers.
1997-03-24 11:07:52 +00:00
davidn
b7b1dda54d Make stlintr() public to allow correct linkage. 1997-03-24 07:25:31 +00:00
kato
17d7dec351 Fix typo.
Submitted by:	Bruce Evans <bde@zeta.org.au>
1997-03-24 07:23:05 +00:00
kato
c4421409ce Merge PC-98 code int i386/isa/ft.c. 1997-03-24 07:20:35 +00:00
gibbs
4626bcf70b Fix a nasty bug that meant a QUEUE_FULL status would result in a lost
SCB.  This is probably a main reason for the recent reports of timeouts.
1997-03-24 05:05:18 +00:00
gibbs
bc569a7a8f Re-arrange the code to upload an SCB to ensure that there is at least
on instruction between the last load into the DFIFO and enabling the DMA.
During back to back SCB DMA, it was possible to hang the card.
1997-03-24 05:02:55 +00:00
bde
bf850fb228 Removed nested #includes of <scsi/scsi_debug.h> and <scsi/scsi_driver.h>
from <scsi/scsiconf.h> and fixed everything that depended on them.
(Missed this one.)
1997-03-24 04:26:46 +00:00
bde
554ea3437c Fixed corrupted newline and corrupted tab in previous commit. 1997-03-24 03:19:37 +00:00
kato
f7ccd15f7a Fixed corrupted CFLAGS definition. (Deleted harmful backslash.)
Submitted by:	H. Nokubi <h-nokubi@nmit.mt.nec.co.jp>
1997-03-24 02:38:48 +00:00
gibbs
9f4b5d5d62 free_xs must be called at splbio(). This is usually the case since the main
caller is scsi_done which the controller interrupt handlers call.  In the
case of a non-buffer based transaction, the xs structure is freed by the
process that initiated the transfer in scsi_scsi_cmd.  In this case, an
explicit splbio/splx pair around the call to free_xs is required.  Without
the splbio protection, the xs free list could be corrupted, and the type
driver's start routine might run without spl protection.

Submitted by:	Tor Egge <Tor.Egge@idt.ntnu.no>
Obtained from: PR kern/2891
1997-03-24 01:46:15 +00:00
guido
c337c37259 Add generation number randomization. Newly created filesystems wil now
automatically have random generation numbers. The kenel way of handling those
also changed. Further it is advised to run fsirand on all your nfs exported
filesystems. the code is mostly copied from OpenBSD, with the randomization
chanegd to use /dev/urandom
Reviewed by:	Garrett
Obtained from: OpenBSD
1997-03-23 20:08:22 +00:00
kato
11933497bf Created new section `CPU OPTIONS'. CPU classes and math emulator are
moved to there.  Options for CPU feature is also described there.
1997-03-23 11:08:28 +00:00
bde
7803d65f32 Removed nested #includes of <scsi/scsi_debug.h> and <scsi/scsi_driver.h>
from <scsi/scsiconf.h> and fixed everything that depended on them.
1997-03-23 06:33:55 +00:00
bde
2143997eab Removed nested #includes of <scsi/scsi_debug.h> and <scsi/scsi_driver.h>
from <scsi/scsiconf.h> and fixed everything that depended on them.
1997-03-23 06:26:47 +00:00
kato
022268abf5 Added Accton EN1207-TX support. 1997-03-23 05:10:14 +00:00
bde
d12a47b40b Removed nested #include of <sys/conf.h> from <scsi/scsi_driver.h>
and fixed everything that depended on getting it from the wrong
place.  Most of the broken things actually only depended on getting
the declaration of their interrupt handler from "ioconf.h".
1997-03-23 04:40:00 +00:00
bde
3676c95c1e Don't #include <sys/fcntl.h> in <sys/file.h> if KERNEL is defined.
Fixed everything that depended on getting fcntl.h stuff from the wrong
place.  Most things don't depend on file.h stuff at all.
1997-03-23 03:49:00 +00:00
bde
0d3591bdbd Don't #include <sys/fcntl.h> in <sys/file.h> if KERNEL is defined.
Fixed everything that depended on getting fcntl.h stuff from the wrong
place.  Most things don't depend on file.h stuff at all.
1997-03-23 03:37:54 +00:00
dyson
9c1cce114f Fix a significant error in the accounting for pre-zeroed pages. This
is a candidate for RELENG_2_2...
1997-03-23 02:44:54 +00:00
bde
3dc0739b4b Merged the rest of lfs from Lite2. It compiles (uncleanly) but is as
unlikely to work as before.
1997-03-23 00:45:27 +00:00