* Add $FreeBSD$ to limerick and zippy
* Delete trailing white space in all its forms. In addition to being
bad style in general, it also causes formatting, and other problems
for various third party items (like xscreensaver) which use fortunes
for their own purposes.
Ian Dowse [Sun, 28 Apr 2002 11:59:30 +0000 (11:59 +0000)]
Fix a number of warnings. A gcc preprocessor bug relating to macros
with variable numbers of arguments made this slightly harder than
it should be. Avoid the bug by not doing string concatenation within
the macros, and instead add a new function to syslog or print the
error messages.
Alan Cox [Sun, 28 Apr 2002 06:07:54 +0000 (06:07 +0000)]
o Introduce and use vm_map_trylock() to replace several direct uses
of lockmgr().
o Add missing synchronization to vmspace_swap_count(): Obtain a read lock
on the vm_map before traversing it.
Add code to emulate arithmetic, comparison and conversion operations
on long double, which are not implemented in hardware on any UltraSPARC
chip that I know of. This just calls into the existing floating point
emulator, which is still needed to emulate other floating point operations
in certain conditions. Without this gcc has to generate the quad floating
point instructions directly, which sometimes causes internal compiler
errors.
Add many prototypes and ANSI-ize 'patch. Not a complete job of it, but it gets
our patch to look a little more like NetBSD's, and has the nice characteristic
that the object code is exactly the same after the change as before it (even in
patch.c and pch.c, which have pesky 'assert' statements in them).
Reviewed by: /sbin/md5 on i386, alpha, sparc64
MFC after: 3 days
Peter Wemm [Sun, 28 Apr 2002 00:15:48 +0000 (00:15 +0000)]
We do not necessarily need to map/unmap pages to zero parts of them.
On systems where physical memory is also direct mapped (alpha, sparc,
ia64 etc) this is slightly harmful.
Ian Dowse [Sat, 27 Apr 2002 22:41:41 +0000 (22:41 +0000)]
Avoid the user-visible effect of setting SA_NOCLDWAIT when the
SIGCHLD handler is SIG_IGN. This is a reimplementation of the
problematic revision 1.131 of kern_exit.c. To avoid accessing process
UPAGES, we set a new procsig flag when the SIGCHLD handler is SIG_IGN
and use that instead.
Peter Wemm [Sat, 27 Apr 2002 22:32:57 +0000 (22:32 +0000)]
Finish fixing hints. Remember the use_kenv state for the next run.
Otherwise we fall back to using the static hints the next time around.
We still have the leftover fallback code there which meant that we skipped
the use_hints checking on the second and subsequent calls. Also, be a bit
more careful about walking off the end of the envp array.
I've extracted this from a larger diff. I hope I didn't miss anything...
Ian Dowse [Sat, 27 Apr 2002 22:20:33 +0000 (22:20 +0000)]
Remove a stale comment saying that the vnode lock must be the first
element in the structure pointed to by vp->v_data; the vnode lock
is now within the vnode structure itself.
Ian Dowse [Sat, 27 Apr 2002 22:10:16 +0000 (22:10 +0000)]
Remove the nfs_{lock,unlock,islocked} functions and the associated
definitions; they have been unused and #if 0'd out since the Lite/2
merge and we are unlikely to want them in the future.
Alan Cox [Sat, 27 Apr 2002 22:01:37 +0000 (22:01 +0000)]
o Begin documenting the (existing) locking protocol on the vm_map
in the same style as sys/proc.h.
o Undo the de-inlining of several trivial, MPSAFE methods on the vm_map.
(Contrary to the commit message for vm_map.h revision 1.66 and vm_map.c
revision 1.206, de-inlining these methods increased the kernel's size.)
Emulate ldq and stq (load/store long double) instructions. GCC has started
using these to load long doubles, but they aren't implemented in hardware
on (at least) UltraSPARC I and II machines.
Emulate popc in the user trap handler as well.
Re-arrange slightly to make support functions more accessible.
Warner Losh [Sat, 27 Apr 2002 16:03:25 +0000 (16:03 +0000)]
This patch fixes my breakage of ssid matching. I introduced the
breakage when I tried to merge OpenBSD wi_hostap changes into the
tree. Skibo found the problem and submitted these patches. Thanks!
Warner Losh [Sat, 27 Apr 2002 16:01:26 +0000 (16:01 +0000)]
Better names for the PCI cards. The biggest change is that we now
identify the Intersil Prism 2.5 PCI native card as that, rather than
Linksys, the first folks to get it to market.
Add a global sx sigio_lock to protect the pointer to the sigio object
of a socket. This avoids lock order reversal caused by locking a
process in pgsigio().
sowakeup() and the callers of it (sowwakeup, soisconnected, etc.) now
require sigio_lock to be locked. Provide sowwakeup_locked(),
soisconnected_locked(), and so on in case where we have to modify a
socket and wake up a process atomically.