David Malone [Sun, 17 Mar 2002 14:03:17 +0000 (14:03 +0000)]
Describe the format of the allow and deny files. This is more-or-less
the patch Matthew submitted, but I broke the lines in a more FreeBSD
way and made one small wording change.
PR: 31265
Submitted by: Matthew D. Fuller <fullermd@over-yonder.net>
MFC after: 3 weeks
- try to open the file
- prompt "Insert the disk labelled <whatever> and press any key..."
- try to open the file
- return error if file could not be located
RE team is free to use this feature in the upcoming 5.0-DP1.
Mark Murray [Sun, 17 Mar 2002 10:05:57 +0000 (10:05 +0000)]
Allow "make lint" to mostly work. Our sources are very unclean WRT
lint, so this is turned off by default. Setting WANT_LINT will turn
on generation of lint libraries for /usr/libdata/lint/*.ln.
Clean up the i4b kernel part: remove unmaintained #if(def)s for NetBSD,
OpenBSD and BSD/OS and respective code, remove pre $FreeBSD CVS id's,
remove #if(def)s and respective code for FreeBSD versions < 5 .
Murray Stokely [Sun, 17 Mar 2002 08:48:08 +0000 (08:48 +0000)]
Add a new entity for the previous stable release. For the most part,
the release notes compare 5.0-CURRENT to 4.0-RELEASE, but in at least
one place it is beneficial to mention the last stable release of
FreeBSD (4.5-RELEASE).
Murray Stokely [Sun, 17 Mar 2002 08:04:02 +0000 (08:04 +0000)]
Add wireless devices to device_names array. Now that ifconfig can set
session IDs, and sysinstall can load modules from the MFSROOT, it
should be possible to install FreeBSD over a wireless link.
Doug Barton [Sun, 17 Mar 2002 07:35:51 +0000 (07:35 +0000)]
Answer the question posed in 1.126. amd won't start without either a
conf file, or command line options. I brought this up in PR 12432,
which (ironically) obrien assigned to me after I became a committer. :)
Murray Stokely [Sun, 17 Mar 2002 07:22:51 +0000 (07:22 +0000)]
* Document the 'floppies' target.
* Document the LOCAL_SCRIPT option.
* Document the NOPORTREADMES option.
* Be more specific in a comment.
* Be more specific about the ftp.1 and cdrom.1 targets.
* Clarify the usage of the CVSROOT variable.
* Clarify the usage of the NODOC variable.
Alan Cox [Sun, 17 Mar 2002 07:01:42 +0000 (07:01 +0000)]
Undo part of revision 1.57: Now that (o)sendsig() doesn't call useracc(),
the motivation for saving and restoring the map->hint in useracc() is gone.
(The same tests that motivated this change in revision 1.57 now show that
there is no performance loss from removing it.) This was really a hack and
some day we would have had to add new synchronization here on map->hint
to maintain it.
Warner Losh [Sun, 17 Mar 2002 05:05:42 +0000 (05:05 +0000)]
Add support for the home Wireless Network airway wireless pcmcia card.
I cannot find where I got these patches, but I think it was from the
bsd-nomads mailing list. Please contact me if you posted them there.
Alan Cox [Sun, 17 Mar 2002 04:21:19 +0000 (04:21 +0000)]
o Stop calling useracc() in (o)sendsig() now that we use copyout()
to copy the sigframe to the user's stack. Useracc() takes a non-trivial
amount of time. Eliminating it speeds up signal delivery by 15% or more.
o Update some comments.
Alan Cox [Sun, 17 Mar 2002 03:19:31 +0000 (03:19 +0000)]
Acquire a read lock on the map inside of vm_map_check_protection() rather
than expecting the caller to do so. This (1) eliminates duplicated code in
kernacc() and useracc() and (2) fixes missing synchronization in munmap().
Jake Burkholder [Sun, 17 Mar 2002 01:53:51 +0000 (01:53 +0000)]
Don't demap the requested page from the tlb in pmap_kenter or pmap_kremove,
even on the local cpu. These are no longer used unsafely in MI code, and
the MD code has been adjusted to compensate.
Jake Burkholder [Sun, 17 Mar 2002 01:51:32 +0000 (01:51 +0000)]
Fix a problem where kernel text could become unmapped when clearing out all
the user mappings from the tlb due to the context numbers rolling over. The
store to the internal mmu register must be followed by a membar #Sync before
much else happens to "avoid data corruption", so we use special inlines which
both disable interrupts and ensure that the compiler will not insert extra
instructions between the two. Also, load the tte tag and check if the context
is nucleus context, rather than relying on the priviledged bit which doesn't
actually serve any purpose in our design, and check the lock bit too for
sanity.
Jake Burkholder [Sun, 17 Mar 2002 01:45:29 +0000 (01:45 +0000)]
Use the tlb data access register to map the kernel tsb, rather than the data
in register. The latter uses the random replacment algorithm to pick the
slot, we want a specific slot.
Kirk McKusick [Sun, 17 Mar 2002 01:25:47 +0000 (01:25 +0000)]
Add a flags parameter to VFS_VGET to pass through the desired
locking flags when acquiring a vnode. The immediate purpose is
to allow polling lock requests (LK_NOWAIT) needed by soft updates
to avoid deadlock when enlisting other processes to help with
the background cleanup. For the future it will allow the use of
shared locks for read access to vnodes. This change touches a
lot of files as it affects most filesystems within the system.
It has been well tested on FFS, loopback, and CD-ROM filesystems.
only lightly on the others, so if you find a problem there, please
let me (mckusick@mckusick.com) know.
Jake Burkholder [Sun, 17 Mar 2002 00:56:41 +0000 (00:56 +0000)]
Convert all pmap_kenter/pmap_kremove pairs in MI code to use pmap_qenter/
pmap_qremove. pmap_kenter is not safe to use in MI code because it is not
guaranteed to flush the mapping from the tlb on all cpus. If the process
in question is preempted and migrates cpus between the call to pmap_kenter
and pmap_kremove, the original cpu will be left with stale mappings in its
tlb. This is currently not a problem for i386 because we do not use PG_G on
SMP, and thus all mappings are flushed from the tlb on context switches, not
just user mappings. This is not the case on all architectures, and if PG_G
is to be used with SMP on i386 it will be a problem. This was committed by
peter earlier as part of his fine grained tlb shootdown work for i386, which
was backed out for other reasons.
Maxim Sobolev [Sat, 16 Mar 2002 23:35:51 +0000 (23:35 +0000)]
When loading a font allow suffix specifying its size be omited, in which
case use size of the currently displaying font as a suffix. For example,
when the when the size of the currently displayed font is 8x8 the
following command will load koi8-r-8x8.fnt.
Warner Losh [Sat, 16 Mar 2002 23:02:41 +0000 (23:02 +0000)]
Don't call the bios if the interrupt appaers to be already routed. Some
older PCI BIOSes hate this and this leads to panics when it is done. Also,
assume that a uniquely routed interrupt is already routed. This also
seems to help some older laptops with feable BIOSes cope.
Warner Losh [Sat, 16 Mar 2002 22:59:15 +0000 (22:59 +0000)]
o Rework the identify routine a little, merging it with NetBSD's wi.
o Add exerpimental support for identifying lucent cards. All of mine
come back with ID of 1, but NetBSD committed code for 5. So accept
both.
o rename wi_prism2_ver to wi_firmware_ver so that we could, if necessary,
do special things for lucent cards too.
o Bring in a small part of the changes from airtools: The wi_cmd function
now takes two additional arguments. I didn't bring in their ioctls yet.
o eliminate the use of LE16TOH, and remove its define.
o Print the firmware as if there were 100 versions instead of 10. This means
that 6.1 and 6.10 aren't confusing to people. We now print 6.01 in the
former case.
# A good junior hacker project would be to merge the NetBSD, FreeBSD, and
# OpenBSD drivers into one source base.
Crist J. Clark [Sat, 16 Mar 2002 20:01:25 +0000 (20:01 +0000)]
Only put standard FreeBSD directories in the PATH. If the
administrator wishes to run commands outside of the PATH, he should
use a full pathname for the executable or set the PATH as appropriate
in any local startup scripts.
Josef Karthauser [Sat, 16 Mar 2002 12:44:21 +0000 (12:44 +0000)]
Merge from NetBSD:
ohcivar.h (1.22), uhcivar.h (1.29):
============================================================
date: 2000/04/25 09:20:55; author: augustss;
Move the size of the mapped bus_space region into the bus
independent softc.
============================================================
ohci.c (1.88), uhci.c (1.112):
============================================================
date: 2000/04/25 14:28:13; author: augustss;
Insert (very conservative!) bus_space_barrier() calls at
all register accesses.
The bus_space(9) man page says you've gotta have them...
============================================================
Josef Karthauser [Sat, 16 Mar 2002 12:06:01 +0000 (12:06 +0000)]
Huge merge from NetBSD:
usbdi.c (1.61):
===================================================================
revision 1.61
date: 2000/01/31 20:13:07; author: augustss; lines: +20 -4
Change the way the HC done method is invoked a little.
===================================================================
usbdi.c (1.65):
===================================================================
revision 1.65
date: 2000/03/08 15:34:10; author: augustss; lines: +4 -2
Get the status right when a polled transfer times out.
===================================================================
ohci.c (1.79), uhci.c (1.89), uhcivar.h (1.24), usb_port.h (1.22),
usbdivar.h (1.48):
===================================================================
date: 2000/03/23 07:01:46; author: thorpej;
New callout mechanism with two major improvements over the old
timeout()/untimeout() API:
- Clients supply callout handle storage, thus eliminating problems of
resource allocation.
- Insertion and removal of callouts is constant time, important as
this facility is used quite a lot in the kernel.
The old timeout()/untimeout() API has been removed from the kernel.
===================================================================
uhci.c (1.80), usbdi.c (1.66):
===================================================================
date: 2000/03/23 18:59:10; author: thorpej;
Shake out some bugs from the callout changes.
===================================================================
ohci.c (1.80), uhci.c (1.91), uhcivar.h (1.25), usb_port.h (1.23),
usbdi.c (1.67), usbdivar.h (1.49):
===================================================================
date: 2000/03/24 22:03:30; author: augustss;
Some cleanup and renaming of the callouts used in USB drivers.
===================================================================
uhci.c (1.92), uhcivar.h (1.26):
===================================================================
date: 2000/03/24 22:57:58; author: augustss;
Two major changes:
Make each xfer have its own intr_info. This is necessary if we want
to queue multiple xfers on an endpoint. This should get rid of the
(mostly harmless) DIAGNOSTICs about intr_infos (not) being done.
Change (again!) how xfers are aborted. Aborting a TD is a nightmare
on the braindead UHCI controller. (Unless you stop the HC, thereby
losing isoc traffic.) Hopefully I got it right this time.
===================================================================
usbdivar.h (1.50):
===================================================================
revision 1.50
date: 2000/03/25 00:10:19; author: augustss; lines: +4 -2
GC an unsued field and add some DIAGNOSTIC in xfer.
===================================================================
ums.c: Use the callout functions instead of the timeout ones.
uhci.c (1.93):
===================================================================
revision 1.93
date: 2000/03/25 00:11:21; author: augustss;
lines: +26 -1
Add more DIAGNOSTIC when aborting isoc.
===================================================================
uhci.c (1.94), usbdivar.h (1.51):
===================================================================
date: 2000/03/25 07:13:05; author: augustss;
More DIAGNOSTIC.
Initialize a callout handle I forgot.
===================================================================
ohci.c (1.81), uhci.c (1.96), uhcivar.h (1.27), usb_subr.c (1.68),
usbdi.c (1.68), usbdivar.h (1.52):
===================================================================
date: 2000/03/25 18:02:33; author: augustss;
Rename and move around callout handles to make it more sane.
Add some DIAGNOSTIC.
Fix buglet in isoc abort on UHCI.
===================================================================
uhci.c (1.98):
===================================================================
revision 1.98
date: 2000/03/27 07:39:48; author: augustss; lines: +12 -4
Make it compile without DIAGNOSTIC.
===================================================================
uhci.c (1.100):
===================================================================
revision 1.100
date: 2000/03/27 09:41:36; author: augustss; lines: +13 -3
Don't mess with QH in bulk abort for the moment.
===================================================================
uhci.c (1.102):
===================================================================
revision 1.102
date: 2000/03/27 22:42:57; author: augustss; lines: +66 -26
Be a little more careful when aborting.
Preallocate some TDs for large buffers.
===================================================================
uhci.c (1.103):
===================================================================
date: 2000/03/28 09:47:10; author: augustss; lines: +11 -1
Another patch for xfer abort...
XXX The current xfer queueing and aborting semantics should really
XXX be changed. It cannot be implemented in a sane way on UHCI.
XXX One day when I have lots of time I'll redesign it...
===================================================================
uhci.c (1.104): Correct a debug message.
uhci.c (1.105): Be more defensive in a DIAGNOSTIC test.
uhci.c (1.106):
===================================================================
revision 1.106
date: 2000/03/29 01:49:13; author: augustss; lines: +14 -309
*SIGH* Revert back to the old method of aborting xfers.
I had tested the new stuff for two months now, but as soon as I commited
it the problems started to appear. Murphy, no doubt...
===================================================================
usb_subr.c (1.70), usbdi.c (1.71), usbdivar.h (1.53):
===================================================================
revision 1.70
date: 2000/03/29 01:45:20; author: augustss; lines: +2 -1
Do not accept new xfers for queuing while a pipe is aborting.
===================================================================
Warner Losh [Sat, 16 Mar 2002 04:26:46 +0000 (04:26 +0000)]
Add device ID for Xircom modem. Also add work around from gwk@sgi.com
to put the device into 8 bit mode a second time. This appears to have
no ill effects on other devices, and appears to be necessary for the
xircom modem.
Submitted by: gwk@sgi.com, many others that found his patch in the archives.
Move the definition of PT_[GS]ET{,DB,FP}REGS from the MD ptrace.h to the
MI ptrace.h, since all platforms define them. Keep the MD ptrace.h around
for FIX_SSTEP (which is currently only needed on Alpha).