| Age | Commit message (Collapse) | Author |
|---|
|
is supposed to ignore the 'n' parameter and return the number of wide
characters needed to represent the given multi-byte character sequence.
However, in the special case where 'pwcs' is NULL and 'n' is zero, our
mbsrtowcs() implementation for single-byte locales mistakenly returned zero.
Before the UTF-8 locale was added, this bug was invisible to callers of
mbstowcs() because mbstowcs() handled this special case itself.
But our new mbstowcs() implementation simply forwards to the locale-specific
mbsrtowcs() implementation and expects it to do the right thing.
The "awesome" window manager's "Run:" command prompt uses mbstowcs() to
measure how many (possibly multi-byte) characters a user has typed, and
due to this bug would always be tricked into thinking the user had entered
zero characters when a single-byte locale was used.
Found after prodding by dcoppa.
ok deraadt sthen espie
|
|
Spotted by Mike Belopuhov. ok jsing@ deraadt@
|
|
plalonde (at) overnet.qc.ca
Prompted by deraadt@
|
|
repair that situation. Darn newbies...
|
|
from its vnode's buffer cache in an interrupt context. Therefore we
need interrupt protection when searching the buffer red-black tree.
ok deraadt@, thib@, art@
|
|
"total sectors" to "boundstart" and "boundend" in the list of fields
that are left alone during a RESTORE operation.
ok deraadt@
|
|
This ensures that all physical fields are filled in with current
values. Lack of values (boundstart, boundend in particular) noted
by ray@.
ok deraadt@
|
|
* correct a few obvious mistakes
* adopt some of jmc@'s recent changes to man(7)
* cut down just a little on the awful tendency
to stick a hyphen between two words.
"yes" deraadt
|
|
USB_SET_REPORT ioctls in ukbd and ums.
This allows usbhidctl to be used on these devices e.g. to dump the report
descriptor of troublesome models.
ok deraadt@
|
|
been put back in the pool. Fixes Daniel Dickman's ses device.
ok matthew@ deraadt@
|
|
in agten(4) uncovered by gcc4.
ok miod@, deraadt@
|
|
knote, remove it from the process's klist; after handling those,
remove and drop any remaining knotes from the process's klist. Ban
attaching knotes to processes that have started exiting or attaching
them via the pid of a thread other than the main thread.
ok tedu@, deraadt@
|
|
is perfectly capable to get things going again all by itself.
ok deraadt@
|
|
is perfectly capable to get things going again all by itself.
ok damien@, deraadt@
|
|
at interrupt reception; level interrupts have STS cleared when processing
is complete. Since the STS has already been reset at interrupt reception
we should not reset late..
Found by marco
|
|
ok deraadt
|
|
|
|
|
|
required; PR 6436
ok deraadt@
|
|
operations only need O_RDONLY.
ok blambert@ deraadt@
|
|
|
|
relaying the server diagnostic back to the user (eg. "500 Line too long").
|
|
during upgrades when /etc/fstab is using disklabel UID's to mount
partitions.
Feedback & trash elimination from halex@
"That seems acceptable to me" deraadt@
|
|
family checks before calling ifatoia/ifatoia6.
ok deraadt@, dlg@, mcbride@
|
|
tweaked to show both negotiated and capable as well as fixing a mask
after comments from damien@
ok deraadt@
|
|
Spotted by Mike Belopuhov, ok yasuoka@
|
|
like they need it. (Or don't, when it appears they don't need it)
|
|
|
|
|
|
|
|
EFAULT when appropriate.
ok deraadt@, miod@
|
|
by KDDI.
|
|
since that's the distinction with -U....
|
|
Seems reasonable to jsg, ok phessler, no response from reyk or pyr
|
|
- parts of the report descriptor not in a format we expect are now ignored,
instead of preventing attachment (e.g. hypothetical multi-bit modifiers).
- modifiers beyond MAXMOD are ignored.
- keycode arrays larger than MAXKEYCODE are clamped to MAXKEYCODE instead
of being rejected.
- multiple keycode arrays are ignored.
This should allow rogue keyboards to attach and be usable up to a certain
extent.
Adapted from a diff sent by Loganaden Velvindron (first name at gmail), who
has a keyboard which keycode array is larger than MAXKEYCODE (but, like most
if not all USB keyboards out there, can only report up to three simultaneous
keypresses anyway).
|
|
|
|
in the old mdoc(7) manual than in the new one, so put them back.
While here, fix the .Os synopsis and add two more .Xrs.
Do the same in man(7), as far as it is applicable.
ok jmc@
|
|
which had much less information, such that the new one gets installed.
No change to the build system required, no text change in this commit.
Doing the big move early even though a few more improvements will follow.
The duplicate information in mdoc.samples(7) will be cleaned up post-release.
ok jmc@, and kristaps@ agreed with the plan, too
|
|
wrong type), try to make it clearer what comes from where, and add a few
XXX.
Currently there is a mess of default options set in vinit(), default
options set in getremote(), stuff read from /etc/remote and options read
from the command-line, and there remains the problems with leftover
options from rotation. This all needs to be sorted out properly later.
Noticed while trying to find a possible problem reported by miod.
|
|
|
|
this is what (Open)Solaris does and it makes the i2c controller on the
blade2.5k work. From NetBSD.
Tested by sthen@, ok deraadt@
|
|
from Jakub Jelinek that never made it into upstream GCC. Fixes ld.so.
Found by drahn@ and me; ok miod@
|
|
Prevents ftpd from letting regular users logging in during anonymous-only
mode (-A).
ok millert, ray
|
|
ok jmc@
|
|
is actually zero and remove a redundant instruction.
ok jsing@
|
|
avoid a lot of extra stupid printing.
|
|
|
|
since we usually deal with directories in a global way, replacing a
directory with something else didn't work.
the fix is to record removed directories in Vstat in a "lazy" way:
we only test these objects for existence if we add something else at
the same location. Since this happens rarely, we can afford to perform
the whole SharedItems dance twice...
|
|
unbreaks opencvs build
|
|
ok nicm@
|
