wpaul [Tue, 26 Jan 1999 15:18:26 +0000 (15:18 +0000)]
When selecting the correct EEPROM offset to use for probing the station
address, account for cards which report the Texas Instruments PCI vendor ID
in addition to Compaq and Olicom. (I don't actually have a card that
reports the TI vendor/device ID, but it appears that some Racore adapters
work this way, and failing to account for it when we have the ID listed
in the supported devices list is a bug.)
jkh [Tue, 26 Jan 1999 09:12:21 +0000 (09:12 +0000)]
Frob the upgrade target to be a bit more inclusive. This appears to
be a good bit more successful at doing automated upgrades than the
previous aout-to-elf target, anyway. I'll MF3 in a couple of days
depending on how this does.
dillon [Tue, 26 Jan 1999 07:36:20 +0000 (07:36 +0000)]
Remove use of kern.conf_dir sysctl. conf_dir is left as a local
variable and propogates back to /etc/rc where it will be used to
locate the rc.local file. The local variable will also be used by
/etc/rc.conf. Note that /etc/rc.conf reverts to its prior operation
of accessing /etc/rc.conf.local if run standalone.
dillon [Tue, 26 Jan 1999 07:32:59 +0000 (07:32 +0000)]
Remove sysctl's from rc.conf, there seems to be a concensus that
rc.conf should only contain simple ops. We still keep the conf_dir
override, however, and this will be used when rc.conf is run from
/etc/rc in a diskless configuration.
peter [Tue, 26 Jan 1999 05:14:36 +0000 (05:14 +0000)]
Move reading of rc.conf sooner as requested by Greg. I'm a tad nervous
about this becase that makes it get run *before* the filesystems are
mounted. If people have added stuff to their rc.conf or rc.conf.local
that uses stuff outside of /bin and /sbin, this will break.
asami [Tue, 26 Jan 1999 03:40:12 +0000 (03:40 +0000)]
Move bsd.port.*mk to under ${PORTSDIR}/Mk (already repository copied).
There are only skeletons left here; they merely serve as a backup to
include the real versions under ${PORTSDIR}/Mk while we update the ports
tree to include them directly.
imp [Tue, 26 Jan 1999 03:13:22 +0000 (03:13 +0000)]
o Add info about Julian's Linux Threads checkin (one of these
days I'm going to write a make kvm top level target).
o Add warning about the
short lived boot block breakage and briefly what to do to
fix it.
julian [Tue, 26 Jan 1999 02:49:52 +0000 (02:49 +0000)]
Mostly remove the VM_STACK OPTION.
This changes the definitions of a few items so that structures are the
same whether or not the option itself is enabled. This allows
people to enable and disable the option without recompilng the world.
As the author says:
|I ran into a problem pulling out the VM_STACK option. I was aware of this
|when I first did the work, but then forgot about it. The VM_STACK stuff
|has some code changes in the i386 branch. There need to be corresponding
|changes in the alpha branch before it can come out completely.
what is done:
|
|1) Pull the VM_STACK option out of the header files it appears in. This
|really shouldn't affect anything that executes with or without the rest
|of the VM_STACK patches. The vm_map_entry will then always have one
|extra element (avail_ssize). It just won't be used if the VM_STACK
|option is not turned on.
|
|I've also pulled the option out of vm_map.c. This shouldn't harm anything,
|since the routines that are enabled as a result are not called unless
|the VM_STACK option is enabled elsewhere.
|
|2) Add what appears to be appropriate code the the alpha branch, still
|protected behind the VM_STACK switch. I don't have an alpha machine,
|so we would need to get some testers with alpha machines to try it out.
|
|Once there is some testing, we can consider making the change permanent
|for both i386 and alpha.
|
[..]
|
|Once the alpha code is adequately tested, we can pull VM_STACK out
|everywhere.
|
julian [Tue, 26 Jan 1999 02:38:12 +0000 (02:38 +0000)]
Enable Linux threads support by default.
This takes the conditionals out of the code that has been tested by
various people for a while.
ps and friends (libkvm) will need a recompile as some proc structure
changes are made.
billf [Mon, 25 Jan 1999 21:08:13 +0000 (21:08 +0000)]
Update pkg_add's remote package feature to reflect the new structure of
wcarchive. This needs to be updated when:
(a) The directory structure changes on wcarchive
(b) getosreldate() has a new significance
(ie. get ELF packages/get a.out packages)
(c) Branches are changed around
As we stand right now, 3.0-postELFday and 4.0-x all point to the same
directory, but when 4.0 starts to have its own packages, this file will
need to be changed.
dillon [Mon, 25 Jan 1999 18:26:09 +0000 (18:26 +0000)]
Add kern.conf_dir sysctl. This is a R+W string used to specify the
directory containing rc.conf.local and rc.local, and possibly other
things in the future.
This sysctl is used by the diskless startup code and new rc.conf. If
it cannot be found or is empty, the system should revert to using /etc.
fenner [Mon, 25 Jan 1999 16:58:56 +0000 (16:58 +0000)]
Port NetBSD's 19990120-accept bug fix. This works around the race condition
where select(2) can return that a listening socket has a connected socket
queued, the connection is broken, and the user calls accept(2), which then
blocks because there are no connections queued.
rnordier [Mon, 25 Jan 1999 10:34:35 +0000 (10:34 +0000)]
Correctly record the end of the a.out symbol table. In practice, a
bootinfo structure where bi_esymtab < bi_symtab was being passed
to the kernel. In the case of older 2.x kernels, this was causing
garbage to be printed to the video console, followed by an exception.
This should resolve a problem reported on -current by Peter Jeremy
<peter.jeremy@auss2.alcatel.com.au>.
markm [Mon, 25 Jan 1999 06:49:43 +0000 (06:49 +0000)]
Link everything against libcrypt. ELF builds complain without it.
Clean up the master makefile a bit and add a "dekerberise" target
for those who have shot themselves in the foot.
dillon [Mon, 25 Jan 1999 04:40:53 +0000 (04:40 +0000)]
Introduce rc script for BOOTP 'diskless' boot. Well, not quite diskless
since the kernel must be booted from something ( like a floppy ). This
script must occur near the beginning of the rc file in order to support
read-only NFS mounts, which in turn allows all the BOOTP machines to use
the same / and /usr.
peter [Mon, 25 Jan 1999 04:08:28 +0000 (04:08 +0000)]
Force the order of the setdefs* so that make -jN doesn't build the
setdefs* multiple times at the same time that it's compiling setdefs[0].c.
I think this was leading to unterminated linker sets etc.
gallatin [Mon, 25 Jan 1999 03:26:45 +0000 (03:26 +0000)]
gethostbyname2() was broken for lookups via NIS on FreeBSD/alpha
due to _gethostbynis() setting h.h_length to sizeof(u_long), which
works out to 8 on alphas. And 8!= NS_INADDRSZ.
markm [Sun, 24 Jan 1999 13:51:26 +0000 (13:51 +0000)]
Replace a bunch of "ln foo bar"'s with "ln -f foo bar".
This means that if(when) you go "sh MAKEDEV all" in /dev
the devices get remade; you don't get errors.
A lot of the changes are for info only; they are commented out.
peter [Sun, 24 Jan 1999 07:40:56 +0000 (07:40 +0000)]
Check the patch obtained from sendmail.org for the header denial-of-service
attack into the vendor branch. It is a little unusual doing it this way
but it will eliminate (or minimize anyway) conflicts when 8.9.3 comes out.
Obtained from: sendmail.org (as posted on bugtraq, but without broken tabs)
bde [Sun, 24 Jan 1999 06:28:37 +0000 (06:28 +0000)]
Go back to only supporting revoke() for bdevs and cdevs. It is very
buggy for fifos, and no one seems to have investigated its behaviour
on other types of files. It has been broken since the Lite2 merge
in rev.1.54.
Nagged about by: Brian Feldman (green@unixhelp.org)
msmith [Sun, 24 Jan 1999 06:06:21 +0000 (06:06 +0000)]
From the PR:
I added a FICL_TRACE-conditioned trace facility based on "see".
It is ugly because words' functions are almost all static, and ficlExec,
where the trace has to be located, can't get their pointers. So, #ifdef
this staticization, and add most of see's body into ficlExec. Duplication
of code, uglyness, etc. But it is cleanly #ifdef'ed, and works like a
charm.
It does not provide "step" facility, though, just trace. It is
tunable at run-time through "trace!". If anyone (most likely me :) ever
wants a step facility, I'll add it. Should be easy.
PR: bin/9652
Submitted by: "Daniel C. Sobral" <dcs@newsguy.com>
dillon [Sun, 24 Jan 1999 06:00:31 +0000 (06:00 +0000)]
vm_map_split() used to dirty the page manually after calling
vm_page_rename(), but never pulled the page off PQ_CACHE if it was on
PQ_CACHE. Dirty pages in PQ_CACHE are not allowed and a KASSERT was
added in -4.x to test for this... and got hit.
In -4.x, vm_page_rename() automatically dirties the page. This commit
also has it deal with the PQ_CACHE case, deactivating the page in that
case.
msmith [Sun, 24 Jan 1999 05:58:18 +0000 (05:58 +0000)]
From the PR:
FICL's TYPE copies the counted string to HERE, as abial has
remarked. Answering to abial's question, this is NOT garanteed to have
enough space.
...
We have dynamic memory. Even before memory-alloc got in, we
already had dynamic memory. Use it, then! (ficlMalloc is sysdep, so I
suppose that's why it was not used for TYPE; ficl is probably designed
to work without a working ficlFree).
Submitted by: "Daniel C. Sobral" <dcs@newsguy.com>
dillon [Sun, 24 Jan 1999 03:49:58 +0000 (03:49 +0000)]
Addendum: The original code that the last commit 'fixed' actually did
not have a bug in it, but the last commit did make it more readable so
we are keeping it.
dillon [Sun, 24 Jan 1999 02:32:15 +0000 (02:32 +0000)]
vm_pager_put_pages() is passed an rcval array to hold per-page return
values. The 'int' return value for the procedure was never used and
not well defined in any case when there are mixed errors on pages, so
it has been removed. vm_pager_put_pages() and associated vm_pager
functions now return void.
dillon [Sun, 24 Jan 1999 00:55:04 +0000 (00:55 +0000)]
Get rid of unused old_m in vm_fault. Add INVARIANTS to test whether
page is still busy after all the hell vm_fault goes through.. it is
supposed to be, and printf() if it isn't. don't panic, though.
dillon [Sun, 24 Jan 1999 00:51:11 +0000 (00:51 +0000)]
Don't try to calculate B_CACHE for an NFS related bp that has a
> 0 b_validend. This will screw up small-writes, causing lots of
little writes out the network.
msmith [Sun, 24 Jan 1999 00:12:04 +0000 (00:12 +0000)]
biosdisk.c
Implement a new variable 'root_disk_unit' which supersedes
'num_ide_disks' and makes it possible to explicitly set the
root device unit number regardless of type considerations.
bootinfo.c
If we can't calculate a dev_t for the root disk, complain and
don't proceed to boot with an invalid boot device.
roger [Sat, 23 Jan 1999 11:32:06 +0000 (11:32 +0000)]
Added detection of Hauppauge IR remote control.
and MSP34xx Audio chip. Fixed i2c read error.
Hauppauge supplied details of new Tuner Types.
Added tuner type 0x1a.
Danny Braniss <danny@cs.huji.ac.il> submitted Bt878
AverMedia detection with PCI subsystem vendor id.