nicm [Fri, 24 Apr 2015 20:58:44 +0000 (20:58 +0000)]
Set up signal handler earlier so that we don't get zombies, reported by
sobrado@.
nicm [Fri, 24 Apr 2015 20:57:51 +0000 (20:57 +0000)]
Document -h, and bring some consistency to the option descriptions. ok jmc
jmc [Fri, 24 Apr 2015 20:32:34 +0000 (20:32 +0000)]
tweaks; ok nicm
mlarkin [Fri, 24 Apr 2015 19:53:43 +0000 (19:53 +0000)]
Enable PAE mode for those CPUs that support it. This allows us to use the
NX bit for userland and kernel W^X. Unlike the previous c.2008 PAE
experiment, this does not provide > 4GB phys ram on i386 - PAE is solely
being used for NX capability this time. If you need > 4GB phys, use amd64.
Userland W^X was committed yesterday by kettenis@, and we will shortly
start reworking the kernel like we did for amd64 a few months back to get
kernel W^X.
This has been in snaps for a few days and tested by kettenis and myself
as well.
ok deraadt@, kettenis@
kettenis [Fri, 24 Apr 2015 19:41:58 +0000 (19:41 +0000)]
Make sure we keep the whole recursive mapping of the PDP instead of just the
mapping for the first page when tearing things down. Seems to fix the last
bug mlarkin@ has been chasing for a while.
ok mlarkin@
deraadt [Fri, 24 Apr 2015 18:49:54 +0000 (18:49 +0000)]
sync
nicm [Fri, 24 Apr 2015 18:06:54 +0000 (18:06 +0000)]
Trim down remote(5) to reflect what is supported by cu(1).
ok sthen jmc
deraadt [Fri, 24 Apr 2015 18:06:31 +0000 (18:06 +0000)]
remove the old tip SCCS note
nicm [Fri, 24 Apr 2015 18:05:21 +0000 (18:05 +0000)]
Remove tip(1), superseded by cu(1).
ok deraadt sthen bmercer
nicm [Fri, 24 Apr 2015 17:34:57 +0000 (17:34 +0000)]
Do not attempt to use ~/.magic if running as root (or issetugid()).
deraadt [Fri, 24 Apr 2015 17:19:06 +0000 (17:19 +0000)]
move an ugly space
nicm [Fri, 24 Apr 2015 17:10:50 +0000 (17:10 +0000)]
We need to loop over the files once to work out the width of the labels
before testing them.
nicm [Fri, 24 Apr 2015 16:47:32 +0000 (16:47 +0000)]
Trying to drop privileges means we can't handle ARG_MAX arguments, so
remove it for now.
nicm [Fri, 24 Apr 2015 16:45:32 +0000 (16:45 +0000)]
Add a couple of missing spaces (style nits).
nicm [Fri, 24 Apr 2015 16:30:06 +0000 (16:30 +0000)]
Use symbolic constants for st_mode flags, no binary change.
nicm [Fri, 24 Apr 2015 16:28:00 +0000 (16:28 +0000)]
If ~/.magic exists but can't be used, fail rather than silently falling
back to /etc/magic.
nicm [Fri, 24 Apr 2015 16:24:42 +0000 (16:24 +0000)]
Fix a couple of errors in the magic source files.
nicm [Fri, 24 Apr 2015 16:24:11 +0000 (16:24 +0000)]
New implementation of the file(1) utility. This is a simplified,
modernised version with a nearly complete magic(5) parser but omits some
of the complex builtin tests (notably ELF) and has a reduced set of
options.
ok deraadt
mpi [Fri, 24 Apr 2015 15:26:22 +0000 (15:26 +0000)]
Revert back to using GCC builtins. This code triggers an off by one in
device_unref() as found by deraadt@.
deraadt [Fri, 24 Apr 2015 14:44:17 +0000 (14:44 +0000)]
another thinkpad event to ignore; from mp39590@gmail
sf [Fri, 24 Apr 2015 12:53:35 +0000 (12:53 +0000)]
vio: If enqueue fails, drop packet.
From brad@
kettenis [Fri, 24 Apr 2015 12:52:38 +0000 (12:52 +0000)]
Enable the NX bit and use it in the PAE pmap code. PAE is still disabled
while we're chasing at least one remaining bug.
ok mlarkin@, deraadt@
sobrado [Fri, 24 Apr 2015 10:57:36 +0000 (10:57 +0000)]
improve wording.
ok jmc@
dlg [Fri, 24 Apr 2015 08:08:29 +0000 (08:08 +0000)]
read_defaults has the thing for strtonum to chew on in cp, not
optarg.
fixes a segfault introduced by the atoi to strtonum change and
reported by ajacatout@
jmc [Fri, 24 Apr 2015 06:26:49 +0000 (06:26 +0000)]
combine -Dd onto one line and update usage();
djm [Fri, 24 Apr 2015 05:26:44 +0000 (05:26 +0000)]
add ssh-agent -D to leave ssh-agent in foreground without enabling
debug mode; bz#2381 ok dtucker@
jsg [Fri, 24 Apr 2015 02:22:08 +0000 (02:22 +0000)]
Use "softintr_pic0" instead of "softintr_fakepic" when faking a
struct device so there is enough space in the buffer for a NUL and
the unit is included in the string.
discussed with deraadt and millert
deraadt [Fri, 24 Apr 2015 01:36:24 +0000 (01:36 +0000)]
2*len -> use xreallocarray()
ok djm
deraadt [Fri, 24 Apr 2015 01:36:00 +0000 (01:36 +0000)]
rename xrealloc() to xreallocarray() since it follows that form.
ok djm
millert [Thu, 23 Apr 2015 17:11:42 +0000 (17:11 +0000)]
Replace atoi() usage with strtonum(). OK deraadt@
florian [Thu, 23 Apr 2015 17:03:01 +0000 (17:03 +0000)]
We cannot log errors with relay_close() before allocating
se_log evbuffer.
(Same problem as the one just fixed in httpd(8))
OK benno
florian [Thu, 23 Apr 2015 16:59:28 +0000 (16:59 +0000)]
We cannot log errors with server_close() before allocating
clt_log evbuffer.
server_close() calls server_log() which uses ctl_log.
Crash reported by Daniel Jakots <vigdis AT chown DOT me>, thanks!
OK benno
schwarze [Thu, 23 Apr 2015 16:17:04 +0000 (16:17 +0000)]
Unify mdoc_deroff() and man_deroff() into a common function deroff().
No functional change except that for mdoc(7), it now skips leading
escape sequences just like it already did for man(7).
Escape sequences rarely occur in mdoc(7) code and if they do,
skipping them is an improvement in this context.
Minus 30 lines of code.
schwarze [Thu, 23 Apr 2015 15:35:39 +0000 (15:35 +0000)]
Get rid of two empty wrapper functions. No functional change.
mpi [Thu, 23 Apr 2015 14:42:02 +0000 (14:42 +0000)]
Fix 13 years old typo that should be responsible for the unhappiness
of UVM on PowerPC architectures by breaking pmap_is_referenced() and
friends.
ok kettenis@
dlg [Thu, 23 Apr 2015 09:56:23 +0000 (09:56 +0000)]
tedu remnants of the previous attempt to implement page zeroing in
the idle thread.
ok deraadt@
dlg [Thu, 23 Apr 2015 09:45:24 +0000 (09:45 +0000)]
replace the use of struct ifqueue in pipex with mbuf_queues.
this has a slight semantic change. previously pipex would only
process up to 128 packets on the input and output queues at a time
and would reschedule the softint if there were any left. now it
mq_delists the current set of pending packets and only processes
them. if anything is added to the queues later they'll cause the
softint to run again.
this in turn lets us deprecate sysctl_ifq since nothing uses it
anymore. because niqueues are mostly wrappers around mbuf_queues,
we can provide sysctl_mq and just #define sysctl_niq to it.
pipex bits are ok yasuoka@
deraadt [Thu, 23 Apr 2015 07:45:50 +0000 (07:45 +0000)]
use reallocarray instead of calloc; avoid the zero before infill
ok nicm
deraadt [Thu, 23 Apr 2015 06:11:19 +0000 (06:11 +0000)]
Do not need to buf[0] = 0 before strlcpy(buf, ...
deraadt [Thu, 23 Apr 2015 05:58:14 +0000 (05:58 +0000)]
delete a NOTDEF of unused code
deraadt [Thu, 23 Apr 2015 05:26:33 +0000 (05:26 +0000)]
KNF and other style adjustments to make this more scrutible.
reviewed by millert
dtucker [Thu, 23 Apr 2015 05:01:19 +0000 (05:01 +0000)]
Add a simple regression test for sshd's configuration parser. Right now,
all it does is run the output of sshd -T back through itself and ensure
the output is valid and invariant.
dtucker [Thu, 23 Apr 2015 04:59:10 +0000 (04:59 +0000)]
Two small fixes for sshd -T: ListenAddress'es are added to a list head so
reverse the order when printing them to ensure the behaviour remains the same,
and print StreamLocalBindMask as octal with leading zero. ok deraadt@
dtucker [Thu, 23 Apr 2015 04:53:53 +0000 (04:53 +0000)]
Check for and reject missing arguments for VersionAddendum and ForceCommand.
bz#2281, patch from plautrba at redhat com, ok djm@
jsg [Thu, 23 Apr 2015 02:55:15 +0000 (02:55 +0000)]
avoid an uneeded deref
ok blambert@ guenther@
deraadt [Thu, 23 Apr 2015 02:13:18 +0000 (02:13 +0000)]
error in strtonum() conversion; found by Nathanael Rensen
dlg [Thu, 23 Apr 2015 00:49:37 +0000 (00:49 +0000)]
dont need \n on panic strings
ratchov [Wed, 22 Apr 2015 20:29:56 +0000 (20:29 +0000)]
Remove forgotten empty files, from miod, thanks
bluhm [Wed, 22 Apr 2015 18:07:32 +0000 (18:07 +0000)]
Expand SLIST macro of unix pcb members before printing the pointer.
Fixes a gcc warning. No binary change.
Found by and OK deraadt@
schwarze [Wed, 22 Apr 2015 16:25:31 +0000 (16:25 +0000)]
Use .Ar for placeholders and .Cm for arguments to be provided verbatim.
No text change.
Inconsistent markup noticed by Alex dot Greif at greifdesign dot net.
OK jmc@ ajacoutot@.
nicm [Wed, 22 Apr 2015 15:32:33 +0000 (15:32 +0000)]
Make session_has return a flag, returning the first winlink found is a
recipe for errors.
nicm [Wed, 22 Apr 2015 15:30:11 +0000 (15:30 +0000)]
Change the windows array into an RB tree and fix some places where we
were only looking at the first winlink for a window in a session.
nicm [Wed, 22 Apr 2015 15:05:03 +0000 (15:05 +0000)]
window_index is only used in one place (window_destroy) so inline it there.
nicm [Wed, 22 Apr 2015 12:50:12 +0000 (12:50 +0000)]
Update termtypes.master to terminfo.src from ncurses-5.9-
20150418.
jsg [Wed, 22 Apr 2015 11:39:04 +0000 (11:39 +0000)]
Init a variable to zero so values aren't or'd into uninitialised memory.
From jeundery at gmail dot com
mpi [Wed, 22 Apr 2015 06:44:17 +0000 (06:44 +0000)]
Add the necessary glue to keep carp(4) working while other pseudo-drivers
are converted to if_input(), this time without breaking parent-less setup.
ok sthen@, dlg@
mpi [Wed, 22 Apr 2015 06:42:11 +0000 (06:42 +0000)]
Add missing header for NBRIDGE chunk, found the hardway and diff from
Rafael Zalamena.
mpi [Wed, 22 Apr 2015 06:39:03 +0000 (06:39 +0000)]
Implement the MI atomic API for PowerPC to avoid using gcc builtins that
include extra sync operations.
ok kettenis@
mlarkin [Wed, 22 Apr 2015 06:26:23 +0000 (06:26 +0000)]
Reduce differences between pae and no-pae modes.
discussed with deraadt
deraadt [Wed, 22 Apr 2015 05:54:46 +0000 (05:54 +0000)]
sync
jsg [Wed, 22 Apr 2015 04:12:22 +0000 (04:12 +0000)]
Init rt to NULL in arpresolve(). Not a problem in practice
as la will be NULL when rt is uninitialised which should cause
the function to return, but makes the code easier to follow.
ok mpi@ miod@
jsg [Wed, 22 Apr 2015 04:02:06 +0000 (04:02 +0000)]
Init error to 0 for sdmmc_mem_single_{read,write}_block
otherwise the value would be uninitialised in the unlikely
case of being called with length 0.
dlg [Wed, 22 Apr 2015 03:48:52 +0000 (03:48 +0000)]
having macros provide semicolons is dangerous.
djm [Wed, 22 Apr 2015 01:38:36 +0000 (01:38 +0000)]
use correct key for nested certificate test
djm [Wed, 22 Apr 2015 01:24:01 +0000 (01:24 +0000)]
unknown certificate extensions are non-fatal, so don't fatal
when they are encountered; bz#2387 reported by Bob Van Zant;
ok dtucker@
nicm [Tue, 21 Apr 2015 22:42:27 +0000 (22:42 +0000)]
Look up indexes as number before name, makes more sense if windows are
named starting with numbers. From Thomas Adam.
nicm [Tue, 21 Apr 2015 22:38:49 +0000 (22:38 +0000)]
Always format real layout even when zoomed.
nicm [Tue, 21 Apr 2015 22:32:40 +0000 (22:32 +0000)]
Do not die on USR1 if any of the socket parent directories are
missing. Reported by Robin Powell.
nicm [Tue, 21 Apr 2015 22:21:41 +0000 (22:21 +0000)]
Simplify error messages when socket connect fails, suggested by "Karthik K".
nicm [Tue, 21 Apr 2015 21:31:02 +0000 (21:31 +0000)]
The free callback could end up being fired before the done callback
(happens on Cygwin), so use a reference count instead of a single
flag. SF bug 188 reported by "iceboy".
nicm [Tue, 21 Apr 2015 21:24:49 +0000 (21:24 +0000)]
Fix moving windows to nonexistent indexes when renumber-windows is
off. From Thomas Adam, reported by Daniel Levai and Theo Buehler.
mlarkin [Tue, 21 Apr 2015 18:47:57 +0000 (18:47 +0000)]
Remove an extra lcr3 that snuck into pmap_switch during yesterday's
cleanup, responsible for various reaper panics pointed out on bugs@ this
morning.
ok deraadt@
mikeb [Tue, 21 Apr 2015 16:34:59 +0000 (16:34 +0000)]
Improve divert-to specification parsing w.r.t. rule address family.
ok henning
florian [Tue, 21 Apr 2015 16:32:24 +0000 (16:32 +0000)]
According to RFC 4861 and the actuall implementation rltime may not
exceed 9000 seconds.
Reported by Alexander Lobodzinski on bugs@, thanks!
OK benno@
schwarze [Tue, 21 Apr 2015 16:13:54 +0000 (16:13 +0000)]
Avoid a use after free when the target node is deleted during validation.
Bug reported by jsg@.
nicm [Tue, 21 Apr 2015 15:34:32 +0000 (15:34 +0000)]
Bind mouse dragging so that it is passed through to applications if they
want it rather than entering copy mode.
nicm [Tue, 21 Apr 2015 15:21:41 +0000 (15:21 +0000)]
Don't eat the mouse event that triggers a drag end because we may want
to pass it on to application inside the pane.
nicm [Tue, 21 Apr 2015 15:18:38 +0000 (15:18 +0000)]
Put mouse_any_flag back, don't know where it went to (still in man page).
nicm [Tue, 21 Apr 2015 15:18:06 +0000 (15:18 +0000)]
Pass mouse events through to commands for if-shell.
nicm [Tue, 21 Apr 2015 15:16:06 +0000 (15:16 +0000)]
cmd_mouse_pane can return NULL, check for that.
schwarze [Tue, 21 Apr 2015 14:27:07 +0000 (14:27 +0000)]
Section 1, not section one: section numbers are always numerical.
For example, "man -s one" wouldn't work either.
Patch from Alex Greif <alex dot greif at greifdesign dot net>.
dlg [Tue, 21 Apr 2015 13:15:54 +0000 (13:15 +0000)]
disable *8 again for now. incoherent archs arent having much fun with it.
henning [Tue, 21 Apr 2015 10:58:28 +0000 (10:58 +0000)]
append the setdir ("5.7/amd64", "snapshots/vax") as get-parameter "path"
when fetching the autoinstall response file.
a webserver that serves static files doesn't give a damn. if I map that
to something that dynamically creates the response file, i can use that to
construct the sets path, or play other arch and/or version dependent
games.
ok krw uwe
schwarze [Tue, 21 Apr 2015 10:46:48 +0000 (10:46 +0000)]
Back in January, Kaspars at Bankovskis dot net reported that the wc(1)
manual claims that -m would count characters, while our version of the
command actually doesn't, lacking POSIX multibyte support. Since
nobody stood up to fix the code, sync the docs with reality for now.
OK jmc@
schwarze [Tue, 21 Apr 2015 10:24:22 +0000 (10:24 +0000)]
When diff(1) finds differences, it returns an exit status of 1.
In that particular case, refrain from printing "diff: exit code 1"
because that exit status doesn't indicate an error condition.
Issue noticed by and patch OK'ed by ajacoutot@.
"I agree with the goal, and I suspect the diff actually achieves it" guenther@.
mpi [Tue, 21 Apr 2015 09:35:32 +0000 (09:35 +0000)]
Revert previous. Packets going through carp_input() but not destinated
to the carp interface would end up enqueued without being decapsulated.
Found the hard way by shten@
jsg [Tue, 21 Apr 2015 07:01:00 +0000 (07:01 +0000)]
Add back a backslash removed in rev 1.42 so KEX_SERVER_ENCRYPT will
include aes again.
ok deraadt@
mlarkin [Tue, 21 Apr 2015 04:40:40 +0000 (04:40 +0000)]
Remove a duplicate variable and #define that snuck in, in a previous
commit.
ok deraadt@
jsg [Tue, 21 Apr 2015 01:49:19 +0000 (01:49 +0000)]
fix a memory leak if tls_read() fails. ok henning@
jsg [Tue, 21 Apr 2015 01:46:57 +0000 (01:46 +0000)]
fix a memory leak. ok blambert@
jsg [Tue, 21 Apr 2015 01:44:47 +0000 (01:44 +0000)]
No need to do "size = (size_t)sb.st_size" both before and after
a fstat() call.
ok mikeb@ markus@
jsg [Tue, 21 Apr 2015 01:41:42 +0000 (01:41 +0000)]
init both fds passed to pipe as -1 instead of initing one twice
ok deraadt@ guenther@ miod@ millert@
guenther [Tue, 21 Apr 2015 01:38:41 +0000 (01:38 +0000)]
The ELF psABI for PPC specifies that the stack shall always be 16-byte aligned.
ok mpi@ deraadt@
dlg [Tue, 21 Apr 2015 01:31:51 +0000 (01:31 +0000)]
rework sgi mutexes to use the owner pointer as the lock.
this is like src/sys/arch/alpha/alpha/mutex.c r1.14.
this changes sgi mutexes so they record which cpu owns the lock
rather than just if the lock is held or not. the diagnostics compare
the owner to the current cpus curcpu() address so they can actually
tell if the current cpu holds the lock instead of whether any cpu
holds the lock.
instead of using custom asm to implement a cas this uses atomic_cas_ptr.
while im here i also shuffled the code. on MULTIPROCESSOR systems
instead of duplicating code between mtx_enter and mtx_enter_try,
mtx_enter simply loops on mtx_enter_try until it succeeds.
this also provides an alternative implementation of mutexes on
!MULTIPROCESSOR systems that avoids interlocking opcodes. mutexes
wont contend on UP boxes, theyre basically wrappers around spls.
we can just do the splraise, stash the owner as a guard value for
DIAGNOSTIC and return. similarly, mtx_enter_try on UP will never
fail, so we can just call mtx_enter and return 1.
tested by and ok miod@
guenther [Tue, 21 Apr 2015 01:24:28 +0000 (01:24 +0000)]
Eliminate the last uses of *fork's second syscall return register; the pid
is zero in the child
ok deraadt@ miod@
mlarkin [Tue, 21 Apr 2015 00:07:51 +0000 (00:07 +0000)]
Reduce differences between i386 pmap modes.
ok kettenis@
miod [Mon, 20 Apr 2015 19:08:52 +0000 (19:08 +0000)]
Remove most static, gives us much better ddb tracebacks from secondary
processors.
mikeb [Mon, 20 Apr 2015 17:22:18 +0000 (17:22 +0000)]
Log if we refuse to continue the exchange when another one that corresponds
to the same policy is already active. OK markus, hshoexer
nicm [Mon, 20 Apr 2015 15:41:32 +0000 (15:41 +0000)]
Remove unused-but-set variables, from Thomas Adam.
nicm [Mon, 20 Apr 2015 15:34:56 +0000 (15:34 +0000)]
Support for multiple key tables to commands to be bound to sequences of
keys. The default key bindings become the "prefix" table and -n the
"root" table. Keys may be bound in new tables with bind -T and
switch-client -T used to specify the table in which the next key should
be looked up. Based on a diff from Keith Amling.