Moved around several lines in the additional contributors section to make the
list in alphabetical order.
Added following persons to the additional contributor list. They have been
missing from the list but listed as maintainer of some port.
Committers should pay much more attention to maintain the contributor list!!
Andrew Stevenson <andrew@ugh.net.au>
Boyd Faulkner <faulkner@mpd.tandem.com>
Mike McGaughey <mmcg@cs.monash.edu.au>
Paul Fox <pgf@foxharp.boston.ma.us>
Rob Mallory <rmallory@csusb.edu>
Several areas of improvement:
o Incorporate some of Tatsumi's bug fixes.
o Remove the xperimnt and commerce distribution items; they haven't
been actual distributions for awhile.
o Try to sanitize the device checking code a little more.
o Cosmetic work on the network code.
Sean Eric Fagan [Sat, 7 Dec 1996 20:20:41 +0000 (20:20 +0000)]
Change how SIGINFO is handled -- set a flag (of sig_atomic_t), and check
that flag on every iteration of the loop. This avoids calling fprintf
inside a signal handler, which is always somewhat icky.
Bruce Evans [Sat, 7 Dec 1996 11:07:20 +0000 (11:07 +0000)]
Fixed `stty [-]extproc'. It set extproc to a semi-random state (always
off if there are no other options, always (?) on if another option that
changes the state is processed earlier).
John Dyson [Sat, 7 Dec 1996 05:08:37 +0000 (05:08 +0000)]
Make LKM's build again. I have added the allegedly broken
ACTUALLY_LKM_NOT_KERNEL until the real fix comes through. Whatever
flag that is used to indicate building LKMs will have be be substituted
in the future. This allows proper selection of the usage of the
efficient, single instruction in the kernel, but alas, doesn't allow
for use in userland or LKMs :-(. Them's the breaks!!!
John Dyson [Sat, 7 Dec 1996 03:46:20 +0000 (03:46 +0000)]
Change endian.h so that when a kernel is compiled without I386_CPU, and
with one of the following (I486_CPU,I586_CPU,I686_CPU) so it can take
advantage of the very quick bswap insn. This keeps LKMs from being
built to take advantage of the insn, but also makes sure that the LKMs
can be run on all CPUs. (The LKMs don't pick up the CPU options :-( ).
Add Guido von Rooij to list of security officers, remove myself from
Webmaster list since I'm not doing that anymore, remove "system administration"
category entirely since it's now a responsibility shared among multiple
individuals and, finally, add FAQ category and stick Peter da Silva's name on
it.
John Dyson [Sat, 7 Dec 1996 00:03:43 +0000 (00:03 +0000)]
Make vm_map_insert much more intelligent in the MAP_NOFAULT case so
that map entries are coalesced when appropriate. Also, conditionalize
some code that is currently not used in vm_map_insert. This mod
has been added to eliminate unnecessary map entries in buffer map.
Additionally, there were some cases where map coalescing could be done
when it shouldn't. That problem has been resolved.
List Peter da Silva explicitly as the maintainer. I feel this to be
better than using the "FAQ" alias as people have either ignored or
misused the alias in the past, and this gives the feeling that there's
an actual person behind it now (which will be a nice change in any case :-).
Bruce Evans [Fri, 6 Dec 1996 12:47:05 +0000 (12:47 +0000)]
Fixed a null pointer panic in ADDENTRY(). Unloading the lkm and joy
modules sort of works now. Their devswitch entries aren't cleaned
up, so accessing them after they have been unloaded causes a panic
in spec_open().
I asked Bill Paul why _getnetbynis() was only being called with 2 parameters,
and he said:
The 3rd agrument is new; looks like it was part of the upgrade to
a new BIND with some IPv6 support. The third argument here should be
AF_INET. In order for it to be anything else, I'd have to add new
NIS functions to support IPv6 lookups. I don't even know what those
look like yet.
So there ya go, add AF_INET as the 3rd argument to the call.
Remove more instances of passing arrays by address when they should
have simply been passed as arrays. In some cases, casts had even
been added to remove the warnings generated by such misuse! Aieee!
gethostname() returns int, not long. Answered my own question by
RTFM'ing. Either both the header files and the man pages are wrong
or this code is, and I'll take the majority decision. :-)
Bruce Evans [Thu, 5 Dec 1996 12:43:30 +0000 (12:43 +0000)]
Fixed handling of modem status changes. Only the most common case of
connect/hangup in !CLOCAL mode was handled correctly. mgetty and ppp
didn't work because they turn on CLOCAL and poll for carrier (or RI?).
Bruce Evans [Wed, 4 Dec 1996 13:42:09 +0000 (13:42 +0000)]
Fixed an off by 1 error in unix2dostime(). The first day of each month
was converted to the last day of the previous month. This bug was
introduced in the optimizations in rev.1.4.
Bump maximum username length to 16 characters. This brings us into
line with BSD/OS and Linux's username limits, making transitioning from
either operating system a lot easier than it is now. I'm currently
running with this change on my system, as are several others, and have
experienced no ill effects.
This is not for 2.2! This needs to get shaken out longer term in 3.0.
Added Pedro Giffuni to the contributor's section for his two ports:
CIM Simula Compiler using C
Isearch A text search Engine from CNIDR (to be commited later)
Nate Williams [Tue, 3 Dec 1996 21:56:15 +0000 (21:56 +0000)]
Added support for '-T' and '-B' (text and binary) file recognition by
peeking inside of Chris Torek's stdio library internals. This is
similar to the code used for other systems, but didn't work on CT's new
implementation.
Nate Williams [Tue, 3 Dec 1996 21:38:52 +0000 (21:38 +0000)]
Added my 'ddial' patches to user-PPP. The new mode tries it's darndest
to keep the link up, so it re-dials whenever it detects the link go
down. This is useful for 'dedicated' links who use PPP.
It's been used for over a year w/out problems at different sites.
Bill Paul [Tue, 3 Dec 1996 17:55:49 +0000 (17:55 +0000)]
Add support for detecting and hopefully using the passwd.adjunct.byname
NIS map which is present on SunOS NIS servers with the SunOS C2 security
hack^Woption installed. I'm convinced that the C2 security option restricts
access to the passwd.adjunct.byname map in the same way that I restrict
access to the master.passwd.{byname,buid} maps (checking for reserved ports),
which means that we should be able to handle passwd.adjunct.byname map
correctly.
If _havemaster() doesn't find a master.passwd.byname map, it will now
test for a passwd.adjunct.byname map before defaulting back to the
standard non-shadowed passwd.{byname,byuid} maps. If _pw_breakout_yp()
sees that the adjunct map was found and the password from the standard
maps starts with ##, it will try to grab the correct password field
from the adjunct map. As with the master.passwd maps, this only happens
if the caller is root, so the shadowing feature is preserved; non-root
users just get back ##username as the encrypted password.
Note that all we do is grab the second field from the passwd.adjunct.byname
entry, which is designated to be the real encrypted password. There are
other auditing fields in the entry but they aren't of much use to us.
Also switched back to using yp_order() to probe for the maps (instead
of yp_first()). The original problem with yp_order() was that it barfed
with NIS+ servers in YP compat mode since they don't support the
YPPROC_ORDER procedure. This condition is handled a bit more gracefully
in yplib now: we can detect the error and just punt on the probing.
Justin T. Gibbs [Tue, 3 Dec 1996 17:04:51 +0000 (17:04 +0000)]
Re-enable reselections as the first thing after we see a bus free. We seem
to miss reselections from some devices and since the reselection response
timeout is only 200ns, enabling reselections too late may be the cause of our
problem.
function ed_attach_NE2000_pci() in if_ed.c passes
an uninitialized block of memory (got with malloc())
to ed_attach. This prevents a proper initialization
of the device descriptor and in my case causes a panic
during the probe, while printing out device info.
Reviewed by: phk
Submitted by: Luigi Rizzo <luigi@iet.unipi.it>