Bruce Evans [Sat, 30 Jan 1999 14:28:10 +0000 (14:28 +0000)]
Rewrote the changes in rev.1.53. Don't ignore errors from makeinfo.
Use fancy substitutions instead of a huge if statement to handle the
rather delicate quoting.
Put a comment here about using suser() to determine super-user-ness.
Using suser() means that the p_acflag will (can) be set, to show that
a process used super-powers during execution. It also makes it easier
to restrict roots we don't trust later on.
Matt Jacob [Sat, 30 Jan 1999 07:29:00 +0000 (07:29 +0000)]
Implement and use Fast Posting for both parallel && fibre. Redo a bit of
the startup code. Implement a call to outer framework function so that
asynchronous events can be handled (e.g., speed negotiation, target mode).
Mark Newton [Sat, 30 Jan 1999 06:29:48 +0000 (06:29 +0000)]
Emulator KLD for SysVR4 executables grabbed from NetBSD.
See http://www.freebsd.org/~newton/freebsd-svr4 for limitations,
capabilities, history and TO-DO list.
Mark Newton [Sat, 30 Jan 1999 06:27:59 +0000 (06:27 +0000)]
New pseudo-device to provide for a way of obtaining a socket by open()ing
something in /dev.
Obtained from: Christos Zoulas, with FreeBSD modifications
Mark Newton [Sat, 30 Jan 1999 06:26:35 +0000 (06:26 +0000)]
Changed struct socket to include a new field (at the end, so as not
to break existing software) acting as a pointer to emulator-specific
state data that some emulators may (or may not) need to maintain about
a socket.
Used by the svr4 module as a place for maintaining STREAMS emulation state.
Discussed with: Mike Smith, Garrett Wollman back in Sept 98
Matthew Dillon [Fri, 29 Jan 1999 22:59:43 +0000 (22:59 +0000)]
*_execsw static structures cannot be const due to the way they interact
with EXEC_SET, DECLARE_MODULE, and module_register. Specifically,
module_register. We may eventually be able to make these const, but
not now.
Brian Somers [Fri, 29 Jan 1999 22:46:31 +0000 (22:46 +0000)]
o Send a CHAP challenge of 16 random digits when RADIUS is
configured. This isn't strictly necessary according to the
rfc, but it's suggested there....
o Don't forget to include our authname when sending a
CHAP challenge when RADIUS is configured.
o Don't supply the ``16'' representing the chap answer
length to radius_Authenticate() - libradius does this
for us.
o When we successfully authenticate via radius_Authenticate(),
continue with datalink_AuthOk() as expected.
Sponsored by: Internet Business Solutions Ltd., Switzerland
Brian Somers [Fri, 29 Jan 1999 22:44:47 +0000 (22:44 +0000)]
Install -C radlib.h
Don't insist that RAD_USER_PASSWORD is supplied before
calling rad_send_request(). Instead, insist on only one
of RAD_USER_PASSWORD and RAD_CHAP_PASSWORD.
Sponsored by: Internet Business Solutions Ltd., Switzerland
Kazutaka YOKOTA [Fri, 29 Jan 1999 18:09:38 +0000 (18:09 +0000)]
Unload the splash decoder module when a screen saver module is loaded.
Strictly speaking, it is not necessary; the screen saver will load
even if the splash module is still in memory. But still, it is the right
thing to do, otherwise the splash decoder module just wasts the kernel space.
Bill Paul [Fri, 29 Jan 1999 16:25:20 +0000 (16:25 +0000)]
Add two additional cards to the supported adapters list: the Racore 8165
and Racore 8148 adapters are now supported by the ThunderLAN driver.
The 8165 is just a plain vanilla 10/100 card; the 8148 is a 'multi-
personality' adapter which can support 10baseT, 100baseTX and 100baseFX
if you include the proper modules.
Also update the tl man page to mention the Racore cards.
Bruce Evans [Fri, 29 Jan 1999 11:31:45 +0000 (11:31 +0000)]
Removed a bogus cast to v_caddr_t. This is part of terminating
v_caddr_t with extreme prejudice. Here the bogons were originally
the same as for c_caddr_t (half-baked K&R support), but rev.1.95
changed one wrong cast and one harmless cast to 2 wrong casts,
and rev.1.96 only fixed the originally wrong cast.
Bruce Evans [Fri, 29 Jan 1999 10:57:18 +0000 (10:57 +0000)]
Fixed comment and whitespace lossage in recent commits. The -W flag
is not implied by -Wall as claimed by gcc.1. Adding it causes a
measly 7193 new warnings for LINT, mostly for "unused parameter" and
"comparison between signed and unsigned".
Bruce Evans [Fri, 29 Jan 1999 09:04:27 +0000 (09:04 +0000)]
Cast to `const char *' instead of to c_caddr_t. This is part of
terminating c_caddr_t with extreme prejudice. Here we depended
on the "opaque" type c_caddr_t being precisely `const char *'
to do unportable pointer arithmetic.
Matthew Dillon [Fri, 29 Jan 1999 08:36:45 +0000 (08:36 +0000)]
More -Wall / -Wcast-qual cleanup. Also, EXEC_SET can't use
C_DECLARE_MODULE due to the linker_file_sysinit() function
making modifications to the data.
Bruce Evans [Fri, 29 Jan 1999 08:29:05 +0000 (08:29 +0000)]
Removed bogus casts to c_caddr_t. This is part of terminating
c_caddr_t with extreme prejudice. Here the original casts to
caddr_t were to support K&R compilers (or missing prototypes),
but the relevant source files require an ANSI compiler.
Matthew Dillon [Fri, 29 Jan 1999 08:12:49 +0000 (08:12 +0000)]
Commit a solution for the SYSINIT vs C_SYSINIT conundrum. The
problem and solution is outlined in the comments, but basically
we needed a way to allow the SYSINIT mechanism to handle const void *
arguments and function pointers as well as non-const arguments and
function pointers while still maintaining the compiler's ability to
issue warnings if you try to use a bad combination.
Bruce Evans [Fri, 29 Jan 1999 08:10:35 +0000 (08:10 +0000)]
Removed a bogus cast to c_caddr_t. This is part of terminating
c_caddr_t with extreme prejudice. Here the point of the original
cast to caddr_t was to break the warning about the const mismatch
between write(2)'s `const void *buf' and `struct uio's `char
*iov_base' (previous bitrot gave a gratuitous dependency on caddr_t
being char *). Compiling with -Wcast-qual made the cast a full
no-op.
This change has no effect on the warning for discarding `const'
on assignment to iov_base. The warning should not be fixed by
splitting `struct iovec' into a non-const version for read()
and a const version for write(), since correct const poisoning
would affect all pointers to i/o addresses. Const'ness should
probably be forgotten by not declaring it in syscalls.master.
Bruce Evans [Fri, 29 Jan 1999 07:23:53 +0000 (07:23 +0000)]
Removed a bogus cast to c_caddr_t. This is part of terminating
c_caddr_t with extreme prejudice. Here we want to convert from
`const char *' to `const char *'. Casting through c_caddr_t is
not the way to do this. The original cast to caddr_t was apparently
to break warnings about const mismatches in other versions of BSD
(in 4.4BSDLite2, the conversion is from `const char *path' to
plain caddr_t).
Mike Smith [Fri, 29 Jan 1999 01:18:15 +0000 (01:18 +0000)]
Allow the interrupt allocated to the PCIC to be tuned from the kernel
environment by setting the variable machdep.pccard.pcic_irq. If
the IRQ is invalid, the PCIC will run in polled mode.
Mike Smith [Fri, 29 Jan 1999 01:16:50 +0000 (01:16 +0000)]
Document the machdep.pccard.pcic_irq tunable, which allows the user to tune
the interrupt which will be given to the PCIC. If the value supplied is
illegal or not available, interrupts will be turned off and polled mode
used instead.
Bill Fumerola [Thu, 28 Jan 1999 20:17:32 +0000 (20:17 +0000)]
Fix nasty bug where getpackagesite() will return an integer if it doesn't
know what revision of FreeBSD is being run. This case should never happen,
but just in case.
Garrett Wollman [Thu, 28 Jan 1999 20:03:31 +0000 (20:03 +0000)]
Now that newsyslog is capable of doing this at a specific time,
let it rotate /var/log/wtmp again, and update monthly/200.accounting to
take this into account. (Some sites might want to change the parameters
of the rotation; it's easier to do this when it's all centralized in
newsyslog.conf.)
Roger Hardiman [Thu, 28 Jan 1999 17:47:47 +0000 (17:47 +0000)]
Support for MSP3410D / MSP3415D Stereo/Mono audio using the audio format
Auto Detection Mode. This leaves MSP3400C owners still unsupported.
Thanks to Gerd Knorr <kraxel@cs.tu-berlin.de> for providing some early
assistance and sample code in the linux bttv driver.
Nicolas Souchu <nsouch@freebsd.org> ported the msp_read/write/reset
functions to smbus/iicbus.
METEOR_INPUT_DEV2 now selects a composite camera on the SVIDEO port.
For true SVIDEO, use METEOR_INPUT_DEV_SVIDEO.
If you get a monochrome image from the SVIDEO port, you have
seleted the wrong input type.
Garrett Wollman [Thu, 28 Jan 1999 17:41:02 +0000 (17:41 +0000)]
Fix synopsis to match 1003.2. Add text describing the way in which
our implementation does not meet 1003.2 (rather than the now outdated
``is expected to comply' language).
Matthew Dillon [Thu, 28 Jan 1999 17:30:51 +0000 (17:30 +0000)]
Clarify the SYSINIT problem by breaking SYSINIT's up into a void *
version and a const void * version. Currently the const void * version
simply calls the void * version ( i.e. no 'fix' is in place ).
A solution needs to be found for the C_SYSINIT ( etc...) family of
macros that allows const void * without generating a warning, but
does not allow non-const void *.
Mark Newton [Thu, 28 Jan 1999 11:45:49 +0000 (11:45 +0000)]
Sun Bug ID 1251858 (on http://sunsolve1.sun.com) discusses the way that
Sun implemented iBCS2 compatibility on Solaris >= 2.6: The emulator
runs in user-mode, patching the LDT so that client programs making
syscalls through the old iBCS2 call gate get handled by the emulator
process. Unemulated syscalls therefore need their own call-gate that
bypasses the emulator. Sun chose LDT entry 4 to implement this, which
is what we've been using as LUDATA_SEL, so we need to change LUDATA_SEL
if we want to run Solaris executables.
KATO Takenori [Thu, 28 Jan 1999 11:24:36 +0000 (11:24 +0000)]
The "easy" fixes for compiling the kernel -Wunused: remove unreferenced static
and local variables, goto labels, and functions declared but not defined.
Kazutaka YOKOTA [Thu, 28 Jan 1999 10:55:55 +0000 (10:55 +0000)]
- Fixed the bug which always ignored Ctrl-Pause/Break on the AT 101
keyboard.
- Translate some keycode for the 84 keyboard so that the 84 keyboard
and the 101 keyboard become more compatible in terms of keycodes.
- Updated the built-in keymaps so that it is in line with the recent
changes in share/syscons/keymaps.
- Added some comment on the Pause/Break key on the 101 keyboard.
Kazutaka YOKOTA [Thu, 28 Jan 1999 10:48:29 +0000 (10:48 +0000)]
- Fix errors in some keymap files.
- The numpad * key should always generate *.
- The numpad - is fkey52 and should not generate a control code (0x1f).
- The numpad 5 is fkey54, not fkey61.
- The numpad 6 is fkey55 and should not generate a control code (0x1e).
- Fix Spanish keymap.
PR: i386/9532
Submitted by jose@we.lc.ehu.es.
- Added Croatian keymap. It is the same as the Slovenian keymap.
PR: misc/9706
Pointed out by: Damjan Marion
- Addef Finnish keymap. It is the same as the Swedish keymap.
PR: bin/9632
Submitted by: Martti Kuparinen
- Assign special functions consistently in all keymap files.
101 keyboard 84 keyboard function
------------------------------------------------------------------------
Ctrl-Alt-Delete Ctrl-Alt-Delete reboot
Ctrl-Alt-Esc Ctrl-Alt-Esc debug
Ctrl-Alt-Space Ctrl-Alt-Space susp
ScrollLock ScrollLock slock
PrintScreen Shift-(Numpad *)/PrintScreen nscr
Ctrl-PrintScreen Shift-Ctrl-(Numpad *)/PrintScreen debug
Alt-PrintScreen/SysRq SysRq nop
Pause Ctrl-NumLock slock
Shift-Pause Shift-Ctrl-NumLock saver
Alt-Pause Alt-Ctrl-NumLock susp
Ctrl-Pause/Break Ctrl-ScrollLock/Break nop
Left W*ndow key NA fkey62
Right W*ndow key NA fkey63
Menu NA fkey64
NOTE: us.unix.kbd and us.emacs.kbd are very much different from the
other keymaps, thus, I didn't touch them.
Brian Somers [Thu, 28 Jan 1999 01:56:34 +0000 (01:56 +0000)]
Initial RADIUS support (using libradius). See the man page for
details. Compiling with -DNORADIUS (the default for `release')
removes support.
TODO: The functionality in libradius::rad_send_request() needs
to be supplied as a set of routines so that ppp doesn't
have to wait indefinitely for the radius server(s). Instead,
we need to get a descriptor back, select() on the descriptor,
and ask libradius to service it when necessary.
For now, ppp blocks SIGALRM while in rad_send_request(), so
it misses PAP/CHAP retries & timeouts if they occur.
Only PAP is functional. When CHAP is attempted, libradius
complains that no User-Password has been specified... rfc2138
says that it *mustn't* be used for CHAP :-(
Sponsored by: Internet Business Solutions Ltd., Switzerland
Matthew Dillon [Wed, 27 Jan 1999 23:49:45 +0000 (23:49 +0000)]
Fix but in devfs_strategy(). Switch cases were falling through
instead of breaking out, so a VCHR devices would run the VCHR
routine and then fall through and run the VBLK routine. Fixed.