Michael Neumann [Sat, 10 Jan 2009 16:34:57 +0000 (17:34 +0100)]
Merge branch 'master' of ssh://crater.dragonflybsd.org/repository/git/dragonfly
Sascha Wildner [Sat, 10 Jan 2009 18:25:54 +0000 (18:25 +0000)]
Do not write two lines for swap to /etc/fstab.
Reported-by: Konstantinos Pachnis <kpachnis@freemail.gr>
Peter Avalos [Sat, 10 Jan 2009 17:13:31 +0000 (12:13 -0500)]
keyadmin hasn't been used in nearly a decade. Buh Bye.
Michael Neumann [Sat, 10 Jan 2009 16:33:15 +0000 (17:33 +0100)]
Use 5MB more for the image file to not go beyond 100%
file system usage.
Simon Schubert [Sat, 10 Jan 2009 15:39:01 +0000 (16:39 +0100)]
Allow mirror override, restructure git commands to allow other branches.
Sascha Wildner [Sat, 10 Jan 2009 16:05:03 +0000 (16:05 +0000)]
Sync iir(4) with FreeBSD.
This fixes disk recognition with the Intel RAID Controller SRCU42L, and
perhaps a few more other things.
Tested-by: Konstantinos Pachnis <kpachnis@freemail.gr>
Dragonfly-bug: <http://bugs.dragonflybsd.org/issue1128>
Michael Neumann [Sat, 10 Jan 2009 13:52:24 +0000 (14:52 +0100)]
Merge branch 'master' of ssh://crater.dragonflybsd.org/repository/git/dragonfly
Michael Neumann [Sat, 10 Jan 2009 13:51:03 +0000 (14:51 +0100)]
Do not hard-code the repository paths, instead use
variables. This makes it easier to use a mirror.
Matthew Dillon [Sat, 10 Jan 2009 01:23:13 +0000 (17:23 -0800)]
Merge branch 'master' of ssh://crater.dragonflybsd.org/repository/git/dragonfly
Sascha Wildner [Sat, 10 Jan 2009 02:01:26 +0000 (02:01 +0000)]
Fix braino on my side.
Matthew Dillon [Sat, 10 Jan 2009 01:17:03 +0000 (17:17 -0800)]
Fix a bug with undo -d. Sometimes the filesystem returns a duplicate TID
causing undo -d to choose the wrong TID. It now searches backwards looking
for the second-to-last TID to diff against, ignoring any dups.
Reported-by: "Brad Frank" <brad.frank@gmail.com>
Matthew Dillon [Fri, 9 Jan 2009 21:53:39 +0000 (13:53 -0800)]
Merge branch 'master' of ssh://crater.dragonflybsd.org/repository/git/dragonfly into devel
Matthew Dillon [Fri, 9 Jan 2009 21:51:56 +0000 (13:51 -0800)]
Fix a system livelock when ^C'ing a program sleeping in the console
driver. ERESTART must be propagated back to userland so the signal can
be processed. Trying to loop on it livelocks the machine.
Reported-by: Hasso Tepper
Sascha Wildner [Fri, 9 Jan 2009 20:08:34 +0000 (20:08 +0000)]
Clean up a bit and mention the new image stuff.
Sascha Wildner [Fri, 9 Jan 2009 20:06:58 +0000 (20:06 +0000)]
Fix typo: should be ${ISOROOT}, not ${ISODIR}.
Michael Neumann [Fri, 9 Jan 2009 16:37:56 +0000 (17:37 +0100)]
Merge branch 'master' of ssh://crater.dragonflybsd.org/repository/git/dragonfly
Sascha Wildner [Fri, 9 Jan 2009 18:29:28 +0000 (18:29 +0000)]
Revert back to installing packages from a chroot.
Also, remove the -I option which prevented the package install scripts
from being run. This is important for the LiveDVD to properly build.
Michael Neumann [Fri, 9 Jan 2009 16:34:04 +0000 (17:34 +0100)]
Cleanup USB stick image generation.
* Use $ISOROOT/mnt instead of the global /mnt.
* Automatically determine the free vn-device.
* Automatically determine the image size.
Matthias Schmidt [Fri, 9 Jan 2009 11:38:43 +0000 (12:38 +0100)]
Merge branch 'master' of git://chlamydia.fs.ei.tum.de/dragonfly
Matthias Schmidt [Fri, 9 Jan 2009 11:35:25 +0000 (12:35 +0100)]
Add IPv6 support to dntpd
This commit allows dntpd to work with either IPv4 or IPv6 NTP servers. I added
two new command line switches (-4/-6) to force dntpd to use only v4/v6 addresses.
Thanks to sephe@ for some good hints.
Reviewed-by: sephe@
Michael Neumann [Fri, 9 Jan 2009 10:36:34 +0000 (11:36 +0100)]
Fix last commit
Uncomment line that was commented in while testing.
Michael Neumann [Fri, 9 Jan 2009 09:54:10 +0000 (10:54 +0100)]
Integrate USB stick image generation into nrelease.
To build the image use:
make img release
Notice that you might have to change some of these
variables below:
IMGFILE ?= ${ISODIR}/dfly.img
IMGMNT ?= /mnt # where the image is mounted
IMGSIZE ?= 450 # the size in MB of the image
IMGVN ?= vn0 # the vn device used for the image
IMGUSBDEV ?= /dev/da0 # USB device name, required for fstab
To write the image to USB disk, use:
dd if=${ISODIR}/dfly.img of=${IMGUSBDEV} bs=1m
Sascha Wildner [Thu, 8 Jan 2009 11:29:12 +0000 (12:29 +0100)]
Add some files which were removed from the system.
Sascha Wildner [Thu, 8 Jan 2009 00:39:10 +0000 (01:39 +0100)]
Adjust for recent changes.
Sascha Wildner [Thu, 8 Jan 2009 00:31:40 +0000 (01:31 +0100)]
Remove some unneeded .Pp's.
Sascha Wildner [Thu, 8 Jan 2009 00:21:08 +0000 (01:21 +0100)]
netconfig is not a function.
Sascha Wildner [Thu, 8 Jan 2009 00:14:19 +0000 (01:14 +0100)]
Fix typo.
Sascha Wildner [Thu, 8 Jan 2009 00:03:24 +0000 (01:03 +0100)]
Use .Er
Sascha Wildner [Wed, 7 Jan 2009 22:35:37 +0000 (23:35 +0100)]
Add /usr/Makefile to 'make upgrade'.
Reported-by: David Rhodus <sdrhodus@gmail.com>
Sascha Wildner [Wed, 7 Jan 2009 22:21:01 +0000 (23:21 +0100)]
Add missing .Lb
Sascha Wildner [Wed, 7 Jan 2009 22:17:21 +0000 (23:17 +0100)]
Small mdoc cleanup.
Sascha Wildner [Wed, 7 Jan 2009 20:48:01 +0000 (21:48 +0100)]
Rename atrun.man -> atrun.8. There's nothing to process here.
Sepherosa Ziehau [Wed, 7 Jan 2009 12:08:29 +0000 (20:08 +0800)]
Add ale(4) to LINT and GENERIC.
Sepherosa Ziehau [Wed, 7 Jan 2009 11:49:56 +0000 (19:49 +0800)]
Disable TX hardware CSUM offloading.
Reported-by: mneumann@
Sepherosa Ziehau [Sat, 3 Jan 2009 12:26:36 +0000 (20:26 +0800)]
Remove sr(4) ISA part's remainder
Sepherosa Ziehau [Sat, 3 Jan 2009 12:04:05 +0000 (20:04 +0800)]
Remove ISA network device cx(4)
Sepherosa Ziehau [Sat, 3 Jan 2009 11:21:46 +0000 (19:21 +0800)]
Remove ISA part of sr(4)
Sascha Wildner [Wed, 7 Jan 2009 02:13:17 +0000 (03:13 +0100)]
Add missing type.
Simon Schubert [Tue, 6 Jan 2009 12:27:38 +0000 (13:27 +0100)]
Sort the .note.ABI-tag section always to the front of a binary.
Our kernel uses a special note section in the elf header to determine
which ABI emulation to use. However the kernel code expects this
section to reside in the first page of a binary.
Up to now, we were just lucky to get most of our binaries linked in a
way so that the note section by chance wound up in the first page.
Yet, this was never guaranteed to happen.
Recent changes to the static library build pushed the note section
after the first page (3rd order effect), so that the kernel could not
determine the correct ABI emulation anymore.
There are two possible fixes:
For one, the kernel could be changed to accept note sections located
anywhere in the binary. This however does not apply to already deployed
systems, which could face problems when upgrading.
Second, we could ensure that the note tag always winds up in the first
page of a binary. We take this approach by changing our linker scripts
to sort the .note.ABI-tag section to the front.
Simon Schubert [Tue, 6 Jan 2009 02:18:30 +0000 (03:18 +0100)]
Properly initialize the new vmspace and register its pmap.
Without this the vmspace's pmap would not be in the global
pmap_list. This later leads to a panic when sys_exec_sys_unregister
releases the vmspace.
Reported-by: Christian Sturm <athaba@inode.at>
Simon Schubert [Mon, 5 Jan 2009 23:38:23 +0000 (00:38 +0100)]
Drop old linker trickery to strip local symbols.
No clue why this ever existed. FreeBSD dropped this >5 years ago.
Simon Schubert [Tue, 6 Jan 2009 01:07:53 +0000 (02:07 +0100)]
Compile libstand for amd64 with i386 headers.
Simon Schubert [Tue, 6 Jan 2009 00:54:59 +0000 (01:54 +0100)]
Prevent build errors by reordering header files.
Simon Schubert [Tue, 6 Jan 2009 00:27:46 +0000 (01:27 +0100)]
Force compilation with -march=i386 to avoid code bloat.
Simon Schubert [Tue, 6 Jan 2009 00:26:40 +0000 (01:26 +0100)]
Properly generate fake i386 machine headers for amd64 build.
Simon Schubert [Tue, 6 Jan 2009 00:26:02 +0000 (01:26 +0100)]
Avoid spilling NXCFLAGS/NXLDFLAGS, like we do already in boot2.
Simon Schubert [Mon, 5 Jan 2009 23:39:29 +0000 (00:39 +0100)]
Adjust for amd64.
Simon 'corecode' Schubert [Sat, 3 Jan 2009 22:51:53 +0000 (23:51 +0100)]
Fold patches into contrib.
Simon 'corecode' Schubert [Sat, 3 Jan 2009 22:00:08 +0000 (23:00 +0100)]
Add i386 support for the amd64 ld.
Simon 'corecode' Schubert [Sat, 3 Jan 2009 21:59:48 +0000 (22:59 +0100)]
Pass flags as CFLAGS.
Simon 'corecode' Schubert [Sat, 3 Jan 2009 21:55:50 +0000 (22:55 +0100)]
Collapse makefile with parent to avoid duplication.
Simon 'corecode' Schubert [Sat, 3 Jan 2009 21:51:24 +0000 (22:51 +0100)]
Don't explicitly set target architecture, let as figure it out itself.
Simon 'corecode' Schubert [Mon, 29 Dec 2008 18:42:20 +0000 (19:42 +0100)]
Adjust for amd64 boot.
Simon 'corecode' Schubert [Sat, 3 Jan 2009 17:23:43 +0000 (18:23 +0100)]
Add wrapper for the cross-tools stage.
Sascha Wildner [Mon, 5 Jan 2009 22:21:06 +0000 (23:21 +0100)]
Remove useless line.
Simon Schubert [Mon, 5 Jan 2009 16:24:04 +0000 (17:24 +0100)]
Add some uhci ids.
Mostly-obtained-from: FreeBSD
Simon Schubert [Mon, 5 Jan 2009 13:32:37 +0000 (14:32 +0100)]
Even in the reinstall case make sure that /boot/modules is a directory.
Simon Schubert [Mon, 5 Jan 2009 12:37:04 +0000 (13:37 +0100)]
if_re needs miibus
Simon Schubert [Mon, 5 Jan 2009 12:20:59 +0000 (13:20 +0100)]
Add card ID for my G41 IGP.
Michael Neumann [Mon, 5 Jan 2009 14:30:49 +0000 (15:30 +0100)]
Use more specific privilege
Michael Neumann [Mon, 5 Jan 2009 14:02:50 +0000 (15:02 +0100)]
Start privilege testing framework
Hasso Tepper [Sun, 4 Jan 2009 21:14:11 +0000 (23:14 +0200)]
Fix up handling of Intel G4X chips some more.
Note that you need at least xf86-video-intel 2.4.3 for this to work.
The G4X doesn't put the GATT into the same area of stolen memory
as all the other chips and older versions of the driver didn't
handle that properly.
Obtained-from: FreeBSD
Peter Avalos [Sun, 4 Jan 2009 20:23:41 +0000 (15:23 -0500)]
Add a README.DELETED for the files removed for pam_passwdqc.
Peter Avalos [Sun, 4 Jan 2009 20:21:20 +0000 (15:21 -0500)]
Merge branch 'vendor/PAM_PASSWDQC'
Peter Avalos [Sun, 4 Jan 2009 20:17:46 +0000 (15:17 -0500)]
Remove files from pam_passwdqc that we don't need.
Peter Avalos [Sun, 4 Jan 2009 17:06:53 +0000 (12:06 -0500)]
Add pam_passwdqc to the build.
This is a pam module that does simple password strength checking. The
features are mentioned in the README and in the manual page. A sample
configuration already exists in /etc/pam.d/passwd, and is not enabled by
default.
Peter Avalos [Sun, 4 Jan 2009 17:19:57 +0000 (12:19 -0500)]
Merge commit 'origin/vendor/PAM_PASSWDQC'
Peter Avalos [Sun, 4 Jan 2009 16:25:06 +0000 (11:25 -0500)]
Import pam_passwdqc 1.0.5 from the Openwall project.
Sepherosa Ziehau [Sun, 4 Jan 2009 12:36:43 +0000 (20:36 +0800)]
White space
Sepherosa Ziehau [Sun, 4 Jan 2009 12:30:21 +0000 (20:30 +0800)]
White space and minor style changes
Sepherosa Ziehau [Sun, 4 Jan 2009 12:19:46 +0000 (20:19 +0800)]
- White space
- Put definition of BUS_DMA_MIN_ALLOC_COMP after bus_dma_tag
Sepherosa Ziehau [Sun, 4 Jan 2009 12:15:14 +0000 (20:15 +0800)]
White space
Sepherosa Ziehau [Sun, 4 Jan 2009 12:08:11 +0000 (20:08 +0800)]
kmalloc(M_INTWAIT) will not fail
Sepherosa Ziehau [Sun, 4 Jan 2009 11:50:53 +0000 (19:50 +0800)]
- Add hw.busdma sysctl tree.
- Add a node under it to show number bounce pages allocated.
Peter Avalos [Sun, 4 Jan 2009 01:30:08 +0000 (20:30 -0500)]
Sync rsh(1) with FreeBSD.
-Take out all the kerberos stuff and offload that to PAM.
-Use <paths.h> instead of a local pathnames.h.
Peter Avalos [Sun, 4 Jan 2009 01:10:04 +0000 (20:10 -0500)]
Add a PAM config for kde.
Obtained-from: FreeBSD
Peter Avalos [Sun, 4 Jan 2009 01:01:22 +0000 (20:01 -0500)]
Add a PAM config for cron(8).
Obtained-from: FreeBSD
Peter Avalos [Sat, 3 Jan 2009 23:58:03 +0000 (18:58 -0500)]
Sync atrun with FreeBSD:
-Add PAM support.
-Remove non-Dragonfly code.
-Use strlcpy().
-Use setusercontext to get login.conf settings.
Peter Avalos [Sat, 3 Jan 2009 22:30:52 +0000 (17:30 -0500)]
Sync pam configuration files with FreeBSD.
-Take advantage of pam_nologin for account management.
-Locked out and expired accounts shouldn't be accessible via remote
mailbox protocols.
-xserver doesn't seem to be used.
Peter Avalos [Sat, 3 Jan 2009 22:10:01 +0000 (17:10 -0500)]
Sync login(1) with FreeBSD. Highlights include:
-Completely PAMify login(1).
-Remove (void) casts.
-Teach login(1) to respect "hushlogin" and "nocheckmail" attributes
defined in user's $HOME/.login_conf.
-Sprinkle static keyword where needed.
Sascha Wildner [Sat, 3 Jan 2009 22:38:55 +0000 (23:38 +0100)]
Add a nice little C mode for Emacs.
Taken-from: FreeBSD (posted by DES to cvs-all some time ago)
Matthew Dillon [Sat, 3 Jan 2009 19:38:36 +0000 (11:38 -0800)]
Merge branch 'master' of ssh://crater.dragonflybsd.org/repository/git/dragonfly into devel
Matthew Dillon [Sat, 3 Jan 2009 19:37:56 +0000 (11:37 -0800)]
Add suppor for BUF_CMD_FLUSH
Suggested-by: Rumko <rumcic@gmail.com>
Matthew Dillon [Sat, 3 Jan 2009 19:37:32 +0000 (11:37 -0800)]
Remove unused field.
Peter Avalos [Sat, 3 Jan 2009 17:54:18 +0000 (12:54 -0500)]
Another temporary hack to hide reentrant netdb functions.
This should help pkgsrc before the release. After the release we should
figure out what it takes to get these thread-safe functions working with
3rd party programs.
Peter Avalos [Sun, 28 Dec 2008 17:40:34 +0000 (12:40 -0500)]
Use libutil and libypclnt in chpass for passwd manipulations.
While I'm here, WARNS and style cleanups.
Obtained-from: FreeBSD
Peter Avalos [Tue, 16 Dec 2008 16:32:02 +0000 (11:32 -0500)]
Bring in a PAM version of passwd.
Obtained-from: FreeBSD
Peter Avalos [Sun, 28 Dec 2008 06:25:56 +0000 (01:25 -0500)]
Do not dereference a null pointer for a malformed line in passwd.
Instead, just copy it silently to prevent programs from bailing. While
I'm here, remove some (void) casts.
Obtained-from: FreeBSD
Peter Avalos [Sat, 3 Jan 2009 01:07:04 +0000 (20:07 -0500)]
Remove libskey(3). libopie replaces it.
Peter Avalos [Sat, 3 Jan 2009 03:21:33 +0000 (22:21 -0500)]
Add full PAM support for account management and sessions to su(1).
Obtained-from: FreeBSD
Peter Avalos [Fri, 2 Jan 2009 20:40:49 +0000 (15:40 -0500)]
Remove S/Key key-handling apps which are replaced by OPIE's.
Peter Avalos [Fri, 2 Jan 2009 20:25:46 +0000 (15:25 -0500)]
Remove rexecd(8).
It's been obsolete since the 4.3BSD days, and it's been discouraged for
over a decade.
Peter Avalos [Fri, 2 Jan 2009 23:16:23 +0000 (18:16 -0500)]
Sync ftpd(8) with FreeBSD. Here are the highlights:
-Prevent cross-site forgery attacks on ftpd(8) due to splitting
long commands into multiple requests.
-Switch from S/Key to OPIE.
-Add PAM support for account management and sessions.
-Avoid calling uninitialized function pointers in protocol switch
code.
-Add support for RFC 2389 (FEAT) and RFC 2640 (UTF8) to ftpd(8).
-Use uniform punctuation, capitalization, and language style
in server messages wherever this doesn't contradict to a particular
message format.
-Use the standardized CHAR_BIT constant instead of NBBY.
-Let tilde expansion be done even if a file/directory doesn't exist yet.
This makes such natural commands as "MKD ~user/newdir" or "STOR
~/newfile" do what they are supposed to instead of failing miserably
with the "File not found" error.
-ANSI function declarations.
-Remove (void) casts and register keyword.
-Block SIGURG while reading from the control channel.
SIGURG is configured by ftpd to interrupt system calls, which is useful
during data transfers. However, SIGURG could interrupt I/O on the
control channel as well, which was mistaken for the end of the session.
A practical example could be aborting the download of a tiny file,
when the abort sequence reached ftpd after ftpd had passed the file
data to the system and returned to its command loop.
-Improve error handling in getline().
-Log pathname arguments to ftp commands as the user specified them;
add the working directory pathname to the log message if any of
such arguments isn't absolute. This has advantage over the old
way of logging that an admin can see what users are actually trying
to do, and where. The old code was also not too robust when it
came to a chrooted session and an absolute pathname.
-Improve handling SIGURG and OOB commands on the control channel.
The major change is to process STAT sent as an OOB command w/o
breaking the current data transfer. As a side effect, this gives
better error checking in the code performing data transfers.
-Never emit a message to stderr: use syslog instead.
When in inetd mode, this prevents bogus messages from
appearing on the control channel. When running as a
daemon, we shouldn't write to the terminal we used to
have at all.
-Don't depend on IPv4-mapped IPv6 address to bind to both IPv4
and IPv6.
-Work around a bug in some clients by never returning raw directory
contents in reply to a RETR command. Such clients consider RETR
as a way to tell a file from a directory.
-Log the actual number of bytes sent on the wire to /var/log/ftpd
instead of the disk size of the file sent. Since the log file
is intended to provide data for anonymous ftp traffic accounting,
the disk size of the file isn't really informative in this case.
Peter Avalos [Fri, 2 Jan 2009 19:36:55 +0000 (14:36 -0500)]
Use a Makefile in etc/pam.d instead of the one in etc/.
Sascha Wildner [Sat, 3 Jan 2009 15:31:16 +0000 (16:31 +0100)]
Remove unneeded variable settings (set by etc/defaults/rc.conf anyway).
Matthias Schmidt [Sat, 3 Jan 2009 13:16:58 +0000 (14:16 +0100)]
Fix a typo. Actually is 16, not 12
Matthias Schmidt [Sat, 3 Jan 2009 13:07:22 +0000 (14:07 +0100)]
Merge branch 'master' of git://chlamydia.fs.ei.tum.de/dragonfly
Matthias Schmidt [Sat, 3 Jan 2009 12:54:33 +0000 (13:54 +0100)]
Make the "Configure installed system" functionality work
This button did not work, because the installer could not determine whether
the installed fs is HAMMER or UFS. Ask the user before mounting the partition.
I chose the simple dialog rather than a automatic solution because it would
make the installer even more ugly (i.e. it would results in some grep+awk+foobar
magic).
Sascha Wildner [Sat, 3 Jan 2009 12:51:17 +0000 (13:51 +0100)]
Adjust for the portmap -> rpcbind change (fix netboot option).
While I'm here, remove some non-Dragonfly stuff and sort the lists.
Matthias Schmidt [Sat, 3 Jan 2009 12:39:38 +0000 (13:39 +0100)]
Zero out the disklabel before using it
Otherwise using a 64-bit disklabel on a disk which contains a 32-bit
disklabel is not possible and vice versa.