The cursor is only inside the scrolling region when we are in origin
mode. In that case, it should use originreg instead of scrollreg. It is
completely valid to place the cursor outside the scrolling region.
rwatson [Thu, 1 Jan 2009 20:03:01 +0000 (20:03 +0000)]
Temporary workaround for the limitations of the mbuf flowid field: zero
the field in the mbuf constructors, since otherwise we have no way to
tell if they are valid. In the future, Kip has plans to add a flag
specifically to indicate validity, which is the preferred model.
marius [Thu, 1 Jan 2009 14:01:21 +0000 (14:01 +0000)]
- Currently the PMAP code is laid out to let the kernel TSB cover the
whole KVA space using one locked 4MB dTLB entry per GB of physical
memory. On Cheetah-class machines only the dt16 can hold locked
entries though, which would be completely consumed for the kernel
TSB on machines with >= 16GB. Therefore limit the KVA space to use
no more than half of the lockable dTLB slots, given that we need
them also for other things.
- Add sanity checks which ensure that we don't exhaust the (lockable)
TLB slots.
Replace syscons terminal renderer by a new renderer that uses libteken.
Some time ago I started working on a library called libteken, which is
terminal emulator. It does not buffer any screen contents, but only
keeps terminal state, such as cursor position, attributes, etc. It
should implement all escape sequences that are implemented by the
cons25 terminal emulator, but also a fair amount of sequences that are
present in VT100 and xterm.
A lot of random notes, which could be of interest to users/developers:
- Even though I'm leaving the terminal type set to `cons25', users can
do experiments with placing `xterm-color' in /etc/ttys. Because we
only implement a subset of features of xterm, this may cause
artifacts. We should consider extending libteken, because in my
opinion xterm is the way to go. Some missing features:
- Keypad application mode (DECKPAM)
- Character sets (SCS)
- libteken is filled with a fair amount of assertions, but unfortunately
we cannot go into the debugger anymore if we fail them. I've done
development of this library almost entirely in userspace. In
sys/dev/syscons/teken there are two applications that can be helpful
when debugging the code:
- teken_demo: a terminal emulator that can be started from a regular
xterm that emulates a terminal using libteken. This application can
be very useful to debug any rendering issues.
- teken_stress: a stress testing application that emulates random
terminal output. libteken has literally survived multiple terabytes
of random input.
- libteken also includes support for UTF-8, but unfortunately our input
layer and font renderer don't support this. If users want to
experiment with UTF-8 support, they can enable `TEKEN_UTF8' in
teken.h. If you recompile your kernel or the teken_demo application,
you can hold some nice experiments.
- I've left PC98 the way it is right now. The PC98 platform has a custom
syscons renderer, which supports some form of localised input. Maybe
we should port PC98 to libteken by the time syscons supports UTF-8?
- I've removed the `dumb' terminal emulator. It has been broken for
years. It hasn't survived the `struct proc' -> `struct thread'
conversion.
- To prevent confusion among people that want to hack on libteken:
unlike syscons, the state machines that parse the escape sequences are
machine generated. This means that if you want to add new escape
sequences, you have to add an entry to the `sequences' file. This will
cause new entries to be added to `teken_state.h'.
- Any rendering artifacts that didn't occur prior to this commit are by
accident. They should be reported to me, so I can fix them.
Discussed on: current@, hackers@
Discussed with: philip (at 25C3)
dougb [Thu, 1 Jan 2009 12:09:57 +0000 (12:09 +0000)]
General Improvements
====================
1. List the command line options in a more standard way
2. Improve the explanations of some of the arguments (-A, -D)
3. Add ARCHSTRING and MTREEDB to the example rc file
4. Re-sort some of the examples according to the existing distinction
of "has a command line version" vs. "does not have a command line version"
Document changes for r186678
===========================
1. /usr/src/etc -> /usr/src where needed [1]
2. Add IGNORE_FILES to the example rc [2] (and remove IGNORE_MOTD)
3. Update the EXIT STATUS section for [3] and [4]
dougb [Thu, 1 Jan 2009 11:41:13 +0000 (11:41 +0000)]
Maintenance and updates
=======================
1. Various improvements to the mtree (-U) feature:
a. Seperate the notion of directory and file (user can override db path)
b. Only check for the existence of the mtree file if -U is set
c. Use mktemp to create the new version of the file
d. More safely install the new file
e. Standardize error messages a bit
2. Remove the last of the MAKEDEV stuff (RIP)
New Features
============
1. Switch to using the top level (e.g., /usr/src) Makefile, and specify
that we should use the *.mk files from the source directory instead of
the installed versions. [1][2] This allows easier cross builds and
simplifies (or in some cases permits) upgrading.
2. Check for the deprecated 'nodev' option in /etc/fstab [3]
3. Add support for the IGNORE_FILES variable [4] and deprecate IGNORE_MOTD
accordingly.
4. Before installing a file check to make sure that the target does not
already exist as a directory [5]
5. Check to be sure that the file installed and error out if not
dougb [Thu, 1 Jan 2009 10:55:26 +0000 (10:55 +0000)]
Revert 184781, 184804, and 184832 (automatic installation of files
that differ only by VCS Id) for the following reasons:
1. It was added without my consent, review, or even a heads up
2. It is something that I've repeatedly said I do not want, and certainly
do not want as the default
3. It is poorly implemented (much too complex, produces false positives
e.g., /etc/mail/helpfile)
Given that this is a situation that comes up very infrequently (usually
only for a major version upgrade) and can usually be handled simply
enough on a one-off basis, I will once again point out that I think
this is a Bad Idea. I would be willing to consider a better implementation
as an option that is off by default.
kientzle [Thu, 1 Jan 2009 02:29:57 +0000 (02:29 +0000)]
Don't try to read the next Gzip header after we reach the
end of the compressed stream. This is desirable behavior,
but the implementation here is very broken and causes strange
problems, so disable it for now.
Thanks to Simon L. Nielsen for reporting this problem.
When sysctl() is being called with a buffer that is too small, it will
return ENOMEM. Unfortunately the changes I made the other day sets the
error number to 0, because it just returns the error number of the
copyout(). Revert this part of the change.
kensmith [Wed, 31 Dec 2008 20:47:48 +0000 (20:47 +0000)]
Explicitly tell the users what a "down side" to using the boot manager
is, consistently call it the boot manager, and switch the order the
options are presented so no boot manager is first in the list (and
hence more or less the default).
This area will probably be rototilled more before 8.0 comes out.
luigi [Wed, 31 Dec 2008 14:25:14 +0000 (14:25 +0000)]
add some USB scanner IDs for recent Epson multifunction devices.
Add support to uscanner.c for known-working devices
(the same should be done for uscanner2.c).
rwatson [Wed, 31 Dec 2008 11:12:24 +0000 (11:12 +0000)]
Merge OpenBSM alpha 4 from OpenBSM vendor branch to head, both
contrib/openbsm (svn merge) and src/sys/{bsm,security/audit} (manual
merge). Add libauditd build parts and add to auditd's linkage;
force libbsm to build before libauditd.
OpenBSM history for imported revisions below for reference.
MFC after: 1 month
Sponsored by: Apple Inc.
Obtained from: TrustedBSD Project
OpenBSM 1.1 alpha 4
- With the addition of BSM error number mapping, we also need to map the
local error number passed to audit_submit(3) to a BSM error number,
rather than have the caller perform that conversion.
- Reallocate user audit events to avoid collisions with Solaris; adopt a
more formal allocation scheme, and add some events allocated in Solaris
that will be of immediate use on other platforms.
- Add an event for Calife.
- Add au_strerror(3), which allows generating strings for BSM errors
directly, rather than requiring applications to map to the local error
space, which might not be able to entirely represent the BSM error
number space.
- Major auditd rewrite for launchd(8) support. Add libauditd library
that is shared between launchd and auditd.
- Add AUDIT_TRIGGER_INITIALIZE trigger (sent via 'audit -i') for
(re)starting auditing under launchd(8) on Mac OS X.
- Add 'current' symlink to active audit trail.
- Add crash recovery of previous audit trail file when detected on audit
startup that it has not been properly terminated.
- Add the event AUE_audit_recovery to indicated when an audit trail file
has been recovered from not being properly terminated. This event is
stored in the new audit trail file and includes the path of recovered
audit trail file.
- Mac OS X and FreeBSD dependent code in auditd.c is separated into
auditd_darwin.c and auditd_fbsd.c files.
- Add an event for the posix_spawn(2) and fsgetpath(2) Mac OS X system
calls.
- For Mac OS X, we use ASL(3) instead of syslog(3) for logging.
- Add support for NOTICE level logging.
OpenBSM 1.1 alpha 3
- Add two new functions, au_bsm_to_errno() and au_errno_to_bsm(), to map
between BSM error numbers (largely the Solaris definitions) and local
errno(2) values for 32-bit and 64-bit return tokens. This is required
as operating systems don't agree on some of the values of more recent
error numbers.
- Fix a bug how au_to_exec_args(3) and au_to_exec_env(3) calculates the
total size for the token. This buge.
- Deprecated Darwin constants, such as TRAILER_PAD_MAGIC, removed.
imp [Wed, 31 Dec 2008 07:41:42 +0000 (07:41 +0000)]
When no driver attaches to a card, don't power down the card. We can
now read config registers of cardbus cards that are inserted, but
aren't attached to a driver.
Also, add a power related comment...
alc [Tue, 30 Dec 2008 19:48:03 +0000 (19:48 +0000)]
Move the implementation of the vm map's fast path on address lookup from
vm_map_lookup{,_locked}() to vm_map_lookup_entry(). Having the fast path
in vm_map_lookup{,_locked}() limits its benefits to page faults. Moving
it to vm_map_lookup_entry() extends its benefits to other operations on
the vm map.
rwatson [Tue, 30 Dec 2008 16:09:57 +0000 (16:09 +0000)]
Rename mbcnt to mbcnt_delta in uipc_send() -- unlike other local
variables named mbcnt in uipc_usrreq.c, this instance is a delta
rather than a cache of sb_mbcnt.
kib [Tue, 30 Dec 2008 12:51:56 +0000 (12:51 +0000)]
Clear the pointers to the file in the struct filedesc before file is closed
in fdfree. Otherwise, sysctl_kern_proc_filedesc may dereference stale
struct file * values.
kib [Tue, 30 Dec 2008 12:51:14 +0000 (12:51 +0000)]
In r185557, the check for existing negative entry for the given name
did not compared nc_dvp with supplied parent directory vnode pointer.
Add the check and note that now branches for vp != NULL and vp == NULL
are the same, thus can be merged.
Reported and reviewed by: kan
Tested by: pho
MFC after: 2 weeks
luigi [Tue, 30 Dec 2008 09:40:50 +0000 (09:40 +0000)]
A number of small changes to make the 'save choice to disk' safer,
and re-enable it as default.
In particular:
+ re-enable the 'update' flag in the Makefile (of course!);
+ commit Warner's patch "orb $NOUPDATE,_FLAGS(%bp)"
to avoid writing to disk in case of a timeout/default choice;
+ fix an off-by-one count in the partition scan code that would
print the wrong name for unknown partitions;
+ unconditionally change the boot prompt to 'Boot:' instead of 'Default:'
to make room for the extra code/checks/messages. Some of the changes
listed below are also made to save space;
+ rearrange and fix comments for known partition types. Right now we
explicitly recognise *BSD, Linux, FAT16 (type 6, used on many USB keys),
NTFS (type 7), FAT32 (type 11).
Depending on other options we also recognise Extended (type 5),
FAT12 (type 1) and FAT16 < 32MB (type 4).
+ Add an entry "F6 PXE" when the code is built with -DPXE (which is
a default now). Technically, F6 boots through INT18, so the prompt 'PXE'
is a bit misleading. Unfortunately the name INT18
is too long and does not fit in - we could use ROM perhaps.
The reason I picked 'PXE' is that on many (I believe) new systems
INT18 calls PXE.
Apart from the choice of the name for PXE/ROM/INT18, this should close
pending issues on the 1-sector boot0 code and we should be able to
move the code to RELENG_7 when it reopens.
ed [Mon, 29 Dec 2008 19:24:00 +0000 (19:24 +0000)]
Fix compilation. Also move ogetkerninfo() to kern_xxx.c.
It seems I forgot to remove `int error' from a single piece of code. I'm
also moving ogetkerninfo() to kern_xxx.c, because it belongs to the
class of compat system information system calls, not the generic sysctl
code.
kib [Mon, 29 Dec 2008 13:25:58 +0000 (13:25 +0000)]
When the insmntque() in the pfs_vncache_alloc() fails, vop_reclaim calls
pfs_vncache_free() that removes pvd from the list, while it is not yet
put on the list.
Prevent the invalid removal from the list by clearing pvd_next and
pvd_prev for the newly allocated pvd, and only move pfs_vncache list
head when the pvd was at the head.
ed [Mon, 29 Dec 2008 12:58:45 +0000 (12:58 +0000)]
Push down Giant inside sysctl. Also add some more assertions to the code.
In the existing code we didn't really enforce that callers hold Giant
before calling userland_sysctl(), even though there is no guarantee it
is safe. Fix this by just placing Giant locks around the call to the oid
handler. This also means we only pick up Giant for a very short period
of time. Maybe we should add MPSAFE flags to sysctl or phase it out all
together.
I've also added SYSCTL_LOCK_ASSERT(). We have to make sure sysctl_root()
and name2oid() are called with the sysctl lock held.
kib [Mon, 29 Dec 2008 12:45:11 +0000 (12:45 +0000)]
vm_map_lock_read() does not increment map->timestamp, so we should
compare map->timestamp with saved timestamp after map read lock is
reacquired, not with saved timestamp + 1. The only consequence of the +1
was unconditional lookup of the next map entry, though.
Tested by: pho
Approved by: des
MFC after: 2 weeks
kib [Mon, 29 Dec 2008 12:12:23 +0000 (12:12 +0000)]
Drop the pseudofs vnode lock around call to pfs_read handler. The handler
may need to lock arbitrary vnodes, causing either lock order reversal or
recursive vnode lock acquisition.
Tested by: pho
Approved by: des
MFC after: 2 weeks
kib [Mon, 29 Dec 2008 12:07:18 +0000 (12:07 +0000)]
After the pfs_vncache_mutex is dropped, another thread may attempt to
do pfs_vncache_alloc() for the same pfs_node and pid. In this case, we
could end up with two vnodes for the pair. Recheck the cache under the
locked pfs_vncache_mutex after all sleeping operations are done [1].
This case mostly cannot happen now because pseudofs uses exclusive vnode
locking for lookup. But it does drop the vnode lock for dotdot lookups,
and Marcus' pseudofs_vptocnp implementation is vulnerable too.
Do not call free() on the struct pfs_vdata after insmntque() failure,
because vp->v_data points to the structure, and pseudofs_reclaim()
frees it by the call to pfs_vncache_free().
rwatson [Sun, 28 Dec 2008 22:40:42 +0000 (22:40 +0000)]
Vendor import of OpenBSM 1.1 alpha4, which incorporates the following
changes since the last imported OpenBSM release:
OpenBSM 1.1 alpha 4
- With the addition of BSM error number mapping, we also need to map the
local error number passed to audit_submit(3) to a BSM error number,
rather than have the caller perform that conversion.
- Reallocate user audit events to avoid collisions with Solaris; adopt a
more formal allocation scheme, and add some events allocated in Solaris
that will be of immediate use on other platforms.
- Add an event for Calife.
- Add au_strerror(3), which allows generating strings for BSM errors
directly, rather than requiring applications to map to the local error
space, which might not be able to entirely represent the BSM error
number space.
- Major auditd rewrite for launchd(8) support. Add libauditd library
that is shared between launchd and auditd.
- Add AUDIT_TRIGGER_INITIALIZE trigger (sent via 'audit -i') for
(re)starting auditing under launchd(8) on Mac OS X.
- Add 'current' symlink to active audit trail.
- Add crash recovery of previous audit trail file when detected on audit
startup that it has not been properly terminated.
- Add the event AUE_audit_recovery to indicated when an audit trail file
has been recovered from not being properly terminated. This event is
stored in the new audit trail file and includes the path of recovered
audit trail file.
- Mac OS X and FreeBSD dependent code in auditd.c is separated into
auditd_darwin.c and auditd_fbsd.c files.
- Add an event for the posix_spawn(2) and fsgetpath(2) Mac OS X system
calls.
- For Mac OS X, we use ASL(3) instead of syslog(3) for logging.
- Add support for NOTICE level logging.
OpenBSM 1.1 alpha 3
- Add two new functions, au_bsm_to_errno() and au_errno_to_bsm(), to map
between BSM error numbers (largely the Solaris definitions) and local
errno(2) values for 32-bit and 64-bit return tokens. This is required
as operating systems don't agree on some of the values of more recent
error numbers.
- Fix a bug how au_to_exec_args(3) and au_to_exec_env(3) calculates the
total size for the token. This bug resulted in "unknown" tokens being
printed after the exec args/env tokens.
- Support for AUT_SOCKET_EX extended socket tokens, which describe a
socket using a pair of IPv4/IPv6 and port tuples.
- OpenBSM BSM file header version bumped for 1.1 release.
- Deprecated Darwin constants, such as TRAILER_PAD_MAGIC, removed.
Obtained from: TrustedBSD Project
Sponsored by: Apple Inc.
bz [Sun, 28 Dec 2008 21:18:01 +0000 (21:18 +0000)]
For consistency use LLE_IS_VALID() in this 4th place that is actually
interested in the (void *)-1 return value hack.
This way we can easily identify those special parts of the code.
bz [Sun, 28 Dec 2008 11:04:24 +0000 (11:04 +0000)]
Include std.ixp425 for "cpu" and comment out duplicate memory map
options. Using the already included std.avila is not considered
to be entirely right (and the options slightly differ) but the best
match we currently have. Upcoming work should fit better.
Reorder another variable to match the layout of other configs.
Reviewed by: sam, warner (earlier version with options removed)
bz [Sat, 27 Dec 2008 23:24:59 +0000 (23:24 +0000)]
Like in the rest of the file and the network stack use inp as
variable name for the inpcb.
For consistency with the other *_hdrsiz functions use 'size'
instead of 'siz' as variable name.
bz [Sat, 27 Dec 2008 22:58:16 +0000 (22:58 +0000)]
Non-functional (style) changes:
- Always use round brackets with return ().
- Add empty line to beginning of functions without local variables.
- Comments start with a capital letter and end in a '.'.
While there adapt a few comments.
bz [Sat, 27 Dec 2008 17:19:16 +0000 (17:19 +0000)]
Hide detect_virtual() along with the accompanying string
arrays under #ifndef XEN to make XEN config compile again.
In case of Xen vm_guest is hard coded.
Move the list for the vm_guest sysctl out of the restictive
bounds as the sysctl is there in either case.
rik [Sat, 27 Dec 2008 15:22:22 +0000 (15:22 +0000)]
Add support for the Oxford OX16PCI958-based card.
Note, that the patch provided with this card for the Linux states that
the card uses DEFAULT_RCLK * 2, while in fact it is '* 10'. So probably
we should also use the subdevice/subvendord here. For now just ignore
that fact.
PR: kern/129665
Submitted by: bsam
Obtained from: united efforst with bsam
bz [Sat, 27 Dec 2008 15:07:51 +0000 (15:07 +0000)]
Permit digits at the beginning and end of kernel config file names for
`make universe'. This catches a few more arm and, once enabled, mips
configs and permits having local configs named like NOINET6.
lulf [Sat, 27 Dec 2008 14:32:39 +0000 (14:32 +0000)]
- Fix an issue with access permissions to underlying disks used by a gvinum
plex. If the plex is a raid5 plex, and is being written to, parity data might
have to be read from the underlying disks, requiring them to be opened for
reading as well as writing.