rwatson [Thu, 22 Mar 2001 04:27:46 +0000 (04:27 +0000)]
o Remove BINMODE=0550 -- no need to restrict execution rights as the
authorization check is on /dev/{,k}mem.
o Update man page to reflect requirement that gid kmem privileges must
be held, not root. (submitted by: peter)
rwatson [Thu, 22 Mar 2001 04:10:51 +0000 (04:10 +0000)]
o Strip BINMODE from Makefile. Installing watch with 0500 provides no
added security benefit, as it doesn't provide extra privilege.
Protection against inappropriate snooping is done by protecting
/dev/snp* properly.
bmah [Thu, 22 Mar 2001 02:40:50 +0000 (02:40 +0000)]
New release notes: OpenSSH resistance to traffic analysis, make(1)
bugs fixed, fsck_ffs in background on mounted filesystems, TI-RPC,
portmap(8)->rpcbind(8), IPv6 support in NFS, lockd, rc(8) dependency
handling.
Synch up hardware devices with etc/defaults/pccard.conf 1.98.2.19
(yes, from RELENG_4). This change needs to be propagated to RELNOTESng.
jhb [Thu, 22 Mar 2001 01:47:02 +0000 (01:47 +0000)]
- XF86.patch is no longer needed as it is all checked into the XFree86
port conditional on the BUILD_XDIST variable.
- Define BUILD_XDIST=yes when compiling XFree86.
alfred [Wed, 21 Mar 2001 23:53:40 +0000 (23:53 +0000)]
This adds /etc/netconfig to the list of files that get copied to
/stand/etc on the mfsroot, and hence to /etc on the install target
disk for use during the install. It also links netconfig and services
into /etc on the mfsroot so that they are available to a fixit
shell.
wpaul [Wed, 21 Mar 2001 23:47:18 +0000 (23:47 +0000)]
ypserv has intimate knowledge of the server transport handle which is
needed to make the asynchronous DNS lookup mechanism work. (It needs to
be able to get/set the transaction ID in the trasport handle so it can
deliver a delayed UDP response when a reply is received from a DNS
server.) With TI-TPC, the transport handle has changed slightly (what
used to be an int is now a size_t) so we need to account for this.
rwatson [Wed, 21 Mar 2001 23:00:39 +0000 (23:00 +0000)]
o Remove an unnecessary debugging printf from ufs_extattr_lookup(),
which resulted in the output of warning messages at boot if
UFS_EXTATTR_AUTOSTART was enabled but ".attribute" and possible
sub-directories weren't in a mounted MFS or UFS file systems.
Pointed out by: dcs
Obtained from: TrustedBSD Project
scottl [Wed, 21 Mar 2001 21:53:09 +0000 (21:53 +0000)]
Back out rev 1.4. The advertised 64K limit on transfers only applies when
using 64bit S/G entries. With this reverted, we are seeing >92MB/sec reads
and >42MB/sec writes on a RAID-5 container.
jhb [Wed, 21 Mar 2001 20:08:07 +0000 (20:08 +0000)]
Always disable paging when exiting back to real mode after receiving a
fatal trap. Also, reload the GDT register to point to BTX's GDT before
playing around with the segment registers to return to real mode. This is
helpful if the kernel causes a fatal exception before it has setup its own
IDT and fault handlers. For example, if one happens to break mtx_init().
Without these changes BTX would recursively page fault (if paging was not
disabled) or triple fault and reset the CPU (without the GDT reload)
instead of providing a potentially useful register dump.
orion [Wed, 21 Mar 2001 14:10:51 +0000 (14:10 +0000)]
Change type of channel speed variable from "int" to "u_int32_t" to
avoid overflow when scaling playback rate. Fixes 44.1k playback from
being 48k (I am deaf to this difference...).
orion [Wed, 21 Mar 2001 12:51:37 +0000 (12:51 +0000)]
Change ordering of SPDIF register pokes. SPDIF enable needs to be the
last poke in sequence. Enabling SPDIF was coercing output rate to
48K, not good for 44.1K tracks.
mckusick [Wed, 21 Mar 2001 09:48:03 +0000 (09:48 +0000)]
Additions to run checks on live filesystems. This change will not
affect current systems until fsck is modified to use these new
facilities. To try out this change, set the fsck passno to zero
in /etc/fstab to cause the filesystem to be mounted without running
fsck, then run `fsck_ffs -p -B <filesystem>' after the system has
been brought up multiuser to run a background cleanup on <filesystem>.
Note that the <filesystem> in question must have soft updates enabled.
ps [Wed, 21 Mar 2001 08:49:02 +0000 (08:49 +0000)]
Add libssl and libcrypto compat4x libraries since the major number
changed. These were taken from the 4.2-RELEASE dist on ftp.freebsd.org.
This will be MFC'd shortly as it is required in RELENG_4 to maintain
compatability with binaries linked against these libraries.
paul [Wed, 21 Mar 2001 08:19:31 +0000 (08:19 +0000)]
Only flush rules that have a rule number above that set by a new
sysctl, net.inet.ip.fw.permanent_rules.
This allows you to install rules that are persistent across flushes,
which is very useful if you want a default set of rules that
maintains your access to remote machines while you're reconfiguring
the other rules.
mckusick [Wed, 21 Mar 2001 04:05:20 +0000 (04:05 +0000)]
Clear the fs_clean flag only when the FS_UNCLEAN flag is not set
(as is done in unmount).
Remove a snapshot inode from the superblock list when its last
name goes away rather than when its last reference goes away.
That way it will be properly reclaimed by fsck after a crash
rather than reenabled when the filesystem is mounted.
mjacob [Wed, 21 Mar 2001 00:49:37 +0000 (00:49 +0000)]
For parallel SCSI, let us now do status with the final CTIO. For the 1080,
I was hanging after sending a xfer CTIO and a status CTIO for a non-discon
INQUIRY- the xfer CTIO was returned as completed OK, but the status CTIO
was dropped on the floor. All the fields looked good. I don't know why
it got dropped. But allowing status to go back with data xfer seemed to
work. I also noticed that with a non-disconnecting command that the
firmware handle in the ATIO is zero- this leads me to believe that the
f/w really can only handle one CTIO at a time in the discon case, and
it had no idea what to do with the second (status) CTIO.
mjacob [Wed, 21 Mar 2001 00:46:44 +0000 (00:46 +0000)]
Check CT2_SENDSTATUS/CT_SENDSTATUS against cto->ct_flags, not
CAM_SEND_STATUS. Set a timeout of 2 seconds per CTIO. Make sure
that the 'real' tag value is being checked against- not the
one that also carries the firmware handle.
bmah [Tue, 20 Mar 2001 23:41:44 +0000 (23:41 +0000)]
New release notes: fxp(4) needs "device miibus", pkg_create(1)
dependency order (submitted by sobomax), "lprm -" works for remote
printers, mergemaster(8) updates, ftpd(8) updates.
Also make sure ESS Maestro-3/Allegro is mentioned in appropriate
hardware lists (submitted by scottl).
iedowse [Tue, 20 Mar 2001 18:10:19 +0000 (18:10 +0000)]
Place the call to pccard_insert_beep() in the inserted() timeout
routine instead of pccard_event(). This avoids spurious extra calls
to pccard_insert_beep() at insert or remove time which could occur
due to noise on the card-present lines.
Clean up some code in pccard_beep.c; we were depending on the order
of evaluation of function arguments, which is undefined in C. Also,
use `0' rather than `NULL' for integer values.
ru [Tue, 20 Mar 2001 18:06:09 +0000 (18:06 +0000)]
If the MANSECT variable is specified, it will override the default
manual section. If, for example, MANSECT is set to 8, the default
MAN1=${PROG}.1 feature of bsd.prog.mk becomes MAN8=${PROG}.8.
Useful for games, libexec, sbin and usr.sbin subtrees.
iedowse [Tue, 20 Mar 2001 17:22:48 +0000 (17:22 +0000)]
Reboot(8) normally waits 5 seconds after sending SIGTERMs to all
processes and then sends SIGKILLs. If a lot of processes are swapped
out, this delay may not be long enough, so processes such as an X
server may be killed before they have had time to clean up properly.
Make this delay more dynamic by waiting up to 60 seconds for swap
page-in activity to end. While I'm here, ANSIfy and remove a
`register' specifier.
alfred [Tue, 20 Mar 2001 10:27:03 +0000 (10:27 +0000)]
Unbreak LINT.
because rpcb_clnt.h is generated during buildworld and only installed into
/usr/include/rpc (and not present in src/include/rpc) we can fix it
by simply not including it when _KERNEL is defined.
this isn't the most elegant, way and might deserve some revisiting later.
bmah [Tue, 20 Mar 2001 06:08:45 +0000 (06:08 +0000)]
New release notes: Kerberos compatability with OpenSSH, dump(8)
and nodump flag, ISC DHCP client update to 2.0pl5, split(1) support
for large files, units(1) update, netstat(1) per-address packet
counting, manpage updates.
Fix up notes on Adaptec SCSI RAID and DPT SmartRAID V/VI SCSI RAID
controllers (reviewed by scottl).
gibbs [Tue, 20 Mar 2001 04:37:19 +0000 (04:37 +0000)]
This is an MFC candidate.
In our idle loop, use an or instruction to set PRELOADEN rather
than rewriting the contents of DMAPARAMS to DFCNTRL. The later
may re-enable the DMA engine if the idle loop is called to complete
the preload of at least one segment when a target disconnects on
an S/G segment boundary but before we have completed fetching the
next segment. This correts a hang, usually in message out phase,
when this situation occurs. This bug has been here for a long
time, so the situation is rare, but not impossible to reproduce.
It only affected Ultra2/U160 controllers.
Correct a few comments.
Extra Sanity. Make sure that SCSIEN is also turned off, along with
HDMAEN, at the end of the data phase.