Mike Smith [Fri, 3 Nov 2000 16:11:05 +0000 (16:11 +0000)]
If we can't get a command to back the bio we just took off the queue,
put the bio back, otherwise we'll drop it when we bail. This was
causing bio lossage under load, leading to eventual system lockup.
Cause prog_make targets to include a ``make obj'' rule if the -o flag
is given to crunchgen.
[ This fixes the previous commit which silently added ``make obj'' ]
A missing feature of crunchgen was the ability to supply make options on
a per program basis.
This has now been added in the following way:
* Harness the make header file that's specified with the -h argument:
- Allow the user to define $(OPTS) to specify make arguments that should
be added to every program target.
- Allow the user to define $(prog_OPTS) to specify make arguments that
should just be added to the build of 'prog'.
* Make sure that $(OPTS) and $(prog_OPTS) are defined when looking through
each program's make file to determine which object files to crunch.
* When building the crunchgen makefile add $(OPTS) and $(prog_OPTS)
to the depend and build rules for $(prog_OBJS).
Warner Losh [Fri, 3 Nov 2000 07:04:43 +0000 (07:04 +0000)]
Try to catchup with my UPDATING backlog:
o cvs 1.11
o /dev/random
o sendmail
o password default change
Please look at these entries and let me know if I've forgotten anything,
or if my understanding doesn't match reality.
This update gets it so 'lpr -r' ("remove the file after printing") will
try to move the file from the source to the destination (spool) directory.
If that succeeds, much time and disk-space will be saved by doing that
instead of copying the entire file only to remove the original. This
could be a big win on machines doing samba-service or CAP-based printing.
Note that this is about the fourth or fifth iteration of the patch, after
trying to address all possible security implications of the change.
PR: 16124
Reviewed by: freebsd-current or freebsd-hackers (some time ago)
Fold the buildopts into the existing crunch.inc file.
Deprecate the "global" crunch.inc file and the CRUNCHFLAGS global build
options. Tools not policy. Move these global settings out into each
picobsd distribution.
Bill Paul [Fri, 3 Nov 2000 00:37:45 +0000 (00:37 +0000)]
Create a pcn_setfilt() routine that twiddles the promiscuous mode
and nobroadcast bits in the mode register and call it both from
pcn_init() and pcn_ioctl(). Sometimes we need to force the state
of the nobroadcast bit after switching out of promisc mode.
Bill Paul [Fri, 3 Nov 2000 00:03:03 +0000 (00:03 +0000)]
Grrrr. Remember to bzero() the mediainfo structures after we allocate
them. If we leave garbage in them, the dc_apply_fixup() routine may
try to follow bogus pointers when applying the reset fixup.
John Baldwin [Thu, 2 Nov 2000 23:28:12 +0000 (23:28 +0000)]
The Number of Fixed Disks at memory location 0x475 is only 1 byte, not a
2 byte word. This fixes machines that probe 30-odd hard drives during boot
in the loader.
Tor Egge [Thu, 2 Nov 2000 21:42:54 +0000 (21:42 +0000)]
Clear the VFREE flag when the vnode is removed from the free list in
getnewvnode(). Otherwise routines called from VOP_INACTIVE() might
attempt to remove the vnode from a free list the vnode isn't on,
causing corruption.
PR: 18012
Implement new printcap options of sr= (aka stat.recv) and sr= (aka stat.send)
in lpd. Stat.recv is useful on a printserver, as something of a network
performance-monitoring tool. Stat.send is a minimal accounting record of
sorts for jobs going to tcp/ip based printers.
Sheldon Hearn [Thu, 2 Nov 2000 14:23:53 +0000 (14:23 +0000)]
Update the SYNOPSIS to reflect that the driver can be loaded as a module.
While here, I also updated the kernel config style, although I wouldn't
recommend doing this for the whole of section 4 yet, since our kernel
config style is still in a state of flux.
Sheldon Hearn [Thu, 2 Nov 2000 13:48:53 +0000 (13:48 +0000)]
Retire linux(8). Using shell scripts to load kernel loadable modules is
out of fashion. This particular case, unlike joy(8) and friends which
are just plain silly, did more than just load a kernel loadable module.
However, /etc/rc and the linux_base port were adjusted a while back to
cope with the absence of this script.
The only outstanding reason to hang on to it would have been for the
linux(8) manual page, which clued folks into the existence of the
Linuxulator. A new linux(4) was introduced a while back. It does
a much better job.
Post process the crunch1.mk file to allow make arguments to passed on
a per program basis allowing a greater control on what is built.
The buildopts file contains Makefile lines of form:
# Anything added to OPTS is added to every build rule.
OPTS= -DNOPAM
# These should only be added to the build of user-ppp.
ppp_OPTS= -DNOKLDLOAD -DNOINET6 -DNONAT -DNOATM -DNOSUID \
-DHAVE_DES -DNORADIUS -DNOI4B -DNONETGRAPH
Really these should be added to crunch.inc, but that file is currently
optional, and if defined masks the global one. Next step will be to move
these global settings back out into the individual builds as OPTS, and
then migrate OPTS and prog_OPTS back into the local crunch.inc file.
Sheldon Hearn [Thu, 2 Nov 2000 10:04:02 +0000 (10:04 +0000)]
Good-bye cruel world.
This software is obsolete, and its successor has interested parties
looking at it. Even if nobody was looking at kgzip, this utility
still wouldn't be useful in an environment where aout kernels
are a thing of the past.
PR: 20643
Submitted by: Peter Pentchev <roam@orbitel.bg>
Ruslan Ermilov [Thu, 2 Nov 2000 09:46:23 +0000 (09:46 +0000)]
Fixed the bug I have introduced in icmp_error() in revision 1.44.
The amount of data we copy from the original IP datagram into the
ICMP message was computed incorrectly for IP packets with payload
less than 8 bytes.
Kris Kennaway [Thu, 2 Nov 2000 06:33:57 +0000 (06:33 +0000)]
Don't use a trivially predictable temporary filename and keep recreating
it again and again, practically begging the Bad Man to insert his symlink
underneath it and send us down the path to oblivion.
Mike Smith [Thu, 2 Nov 2000 00:37:45 +0000 (00:37 +0000)]
Improve the PCI interrupt routing code. Now the process is as follows:
- Look for a hardwired interrupt in the routing table for this
bus/device/pin (we already did this).
- Look for another device with the same link byte which has a hardwired
interrupt.
- Look for a PCI device matching an entry with the same link byte
which has already been assigned an interrupt, and use that.
- Look for a routable interrupt listed in the "PCI only" interrupts
field and use that.
- Pick the first interrupt that's marked as routable and use that.
Bill Paul [Thu, 2 Nov 2000 00:00:30 +0000 (00:00 +0000)]
Fix a couple of cases where I tried to release the I/O space resource twice
(once as as an I/O space resource and once as an IRQ resource). There was
a problem with this in if_rl too, which is how I found it.
Andrew Gallatin [Wed, 1 Nov 2000 23:17:31 +0000 (23:17 +0000)]
Support for the linux ipc syscalls on the alpha, where each one has
its own syscall rather than going through a demux function like
linux_ipc() on i386
Bosko Milekic [Wed, 1 Nov 2000 23:02:19 +0000 (23:02 +0000)]
Fixup some wrong statements in the new mbuf(9) man page.
Also introduce a bunch of (missed?) macros and functions.
This man page still needs a lot of work, most likely a re-ordering
of the macros/functions, and a more complete, more accurate, listing of
available routines.
A good and worthy start nonetheless.
Andrew Gallatin [Wed, 1 Nov 2000 22:36:41 +0000 (22:36 +0000)]
fix linux_termio and linux_termios structs on alpha. alpha differences
are in the termios struct (probably because linux wants to be compatible
with the osf/1 termios struct), not the termio struct.
John Polstra [Wed, 1 Nov 2000 20:19:07 +0000 (20:19 +0000)]
At the beginning of pthread_mutex_lock(), call _thread_init() if
necessary. This works around a bug in old versions of libgcc_r.a
which are statically linked into old executables.
John Baldwin [Wed, 1 Nov 2000 18:40:02 +0000 (18:40 +0000)]
- Modify alpha_setup_intr() to take interrupt handler flags as an additional
argument. These flags include INTR_FAST, INTR_MPSAFE, etc.
- Properly handle INTR_EXCL when it is passed in to allow an interrupt
handler to claim exclusive ownership of an interrupt thread.
- Add support for psuedo-fast interrupts on the alpha. For fast interrupts,
we don't allocate an interrupt thread; instead, during dispatching of an
interrupt, we run the handler directly instead of scheduling the thread
to run. Note that the handler is currently run without Giant and must be
MP safe. The only fast handler currently is for the sio driver.
Bill Paul [Wed, 1 Nov 2000 18:26:36 +0000 (18:26 +0000)]
Close PR #21078: the aue driver was not correctly programming the
multicast filter on the Pegasus chip. Since IPv6 depends a lot
on multicasting, this caused several failures for people trying to
use IPv6 with Pegasus USB ethernet devices.
Andreas Klemm [Wed, 1 Nov 2000 13:30:24 +0000 (13:30 +0000)]
fixes:
a) the configured default printer entry might turn out to become a security hole
Although lpd isn't enabled by default in FreeBSD 4 and later versions
bad things might happen because of a simple copy and paste failure:
- fill up root-fs, if /dev/lpt0 doesn't exist
- fill up the spool dir (either root-fs or var-fs) if machine hadn't been
designed as print server
Therefore best decision: if people want printing, then configure both
1. /etc/rc.conf
2. /etc/printcap
the default entry wouldn't have served well all tastes and needs anyway...
design of most of our config files is, to have commented out suggestions
if a service is not active.
b) [Garance]
fix old and wrong documentation:
input filters are possible even if you print to a remote printer
this makes FreeBSDs implementation of lpd currently the best without
having to switch to port monster LPRng ...
c) fix pointer to wrong handbook section for a longer time .. so I doesn't fix the number
I'm only referring to the printing section, otherwise this would have to be fixed several
times, if the chapters should be reordered again ...
d) typo: chose -> choose
Submitted by: me and some suggestions by Garance
Approved by: Eivind and Garance A Drosihn <drosih@rpi.edu> in private e-mail
Doug Barton [Wed, 1 Nov 2000 07:21:56 +0000 (07:21 +0000)]
Reorder one of the new features based on experience. Since the number
of files auto-installed during an upgrade from a really old system
can get quite long, and it's piped to the PAGER already, print
that first, then print any of the 4 two-line messages that might
apply.
John Polstra [Wed, 1 Nov 2000 01:18:27 +0000 (01:18 +0000)]
Tell libgcc_r how to initialize a mutex before it uses it. This
fixes the immediate segmentation violations from some multithreaded
programs built on very up-to-date systems.