Commit Graph

42848 Commits

Author SHA1 Message Date
Matthew Dillon
87c8f0c0f2 Add in required instruction serialization prior to releasing the
MP lock for the last time.  The use of a locked instruction to
    cpu-private memory is 3x faster then CPUID and 3x faster then the
    use of a locked instruction to shared memory (the lock itself).

    Instruction serialization is required to ensure that any pending
    memory ops are properly flushed prior to the release of the lock,
    due to out-of-order instruction execution by the cpu.
1999-11-23 18:46:09 +00:00
Marcel Moolenaar
84eadd6305 Add .ORDER which is needed for the previous commit to work with -jN.
Pointed out by: bde
1999-11-23 17:18:52 +00:00
Daniel C. Sobral
8a3f7b6ba9 Add machine-specific include path to ficl's sysdep.h. Wishes I had
gone to bed earlier.

Pointed by: peter
1999-11-23 16:30:48 +00:00
Daniel C. Sobral
0ae220095f Revert ill-considered simplification in 1.13.
Pointed by: peter
1999-11-23 16:28:39 +00:00
Daniel C. Sobral
83b912c6e6 Improve Makefile by using .PATH.
Taught by: peter
1999-11-23 15:55:28 +00:00
Daniel C. Sobral
8967d5f42e Improve on ficl port to alpha.
Files sysdep.[ch] are now in ${MACHINE_ARCH} subdirectory. Internal
#if's used to identify the platform where removed.

Make rule for target testmain was greatly simplified, because it was
easier simplifying it than changing it to support the new location of
sysdep.[ch].

(a repo-copy was done on sysdep.[ch], of course)
1999-11-23 15:24:30 +00:00
Daniel C. Sobral
99915e5b91 (Hopefully) make all necessary changes for ficl to support alpha. 1999-11-23 11:17:37 +00:00
Marcel Moolenaar
494b6f4cba Improve backward compatibility.
make(1) uses sysctlbyname() to find out if it is running on a PC98
machine. This check has been added on 1998/9/9. The MIB variable was
added on 1998/8/31. At that time __FreeBSD_version was 300003. So,
only perform the check *if* __FreeBSD_version is defined and if it's
greater than 300003.

For the record: sysctlbyname was added on 1997/5/30 so this change
automaticly handles that...
1999-11-23 10:35:24 +00:00
Marcel Moolenaar
5f644e4c72 vnode_if.sh makes both vnode_if.c and vnode_if.h. Add vnode_if.c to
the LHS of the rule for vnode_if.h. This solves a "dunno how to make"
error.
1999-11-23 10:17:36 +00:00
Alan Cox
2ed14a92db Correct the following error: vm_map_pageable() on a COW'ed (post-fork)
vm_map always failed because vm_map_lookup() looked at
 "vm_map_entry->wired_count" instead of "(vm_map_entry->eflags &
 MAP_ENTRY_USER_WIRED)".  The effect was that many page
 wiring operations by sysctl were (silently) failing.
1999-11-23 06:51:28 +00:00
Yoshinobu Inoue
e1da8747e7 Removed IPSEC and IPV6FIREWALL because they are not ready yet. 1999-11-23 05:42:36 +00:00
Andrey A. Chernov
d65bc34042 Document single_mountd_enable 1999-11-23 04:14:35 +00:00
Brian Feldman
9a46ff715e Fix a confusion between osigcontext and ucontext_t in the previous commit.
Since an osigcontext is smaller, if you check for a valid (much larger sized)
ucontext_t and it fails, we bogusly would reject the osigcontext as per
rev 1.378.  Instead, check for osigcontext range validity first, and
ucontext_t later.  This unbreaks Netscape.

Pointed to the right commit by:	peter
1999-11-23 04:09:13 +00:00
Andrey A. Chernov
41a1d0d2b5 Allow man to parse MANLOCALES but not to store it 1999-11-23 03:38:41 +00:00
Andrey A. Chernov
1ed4b3e952 #ifdef MAIN locale additions to build man 1999-11-23 03:34:17 +00:00
Andrey A. Chernov
08abdae453 Remove man_locles - goes to manpath.config 1999-11-23 03:33:25 +00:00
Andrey A. Chernov
0a0be98ce3 Remove man_locales - goes to manpath.config 1999-11-23 03:24:20 +00:00
Andrey A. Chernov
6e31a7f7fc Use manpath -L for man locales 1999-11-23 03:21:19 +00:00
Andrey A. Chernov
597ccab43e Use manpath -L for locales 1999-11-23 03:11:54 +00:00
Andrey A. Chernov
f052fd3563 Add -L: print MANLOCALES variable for man locales list 1999-11-23 03:06:18 +00:00
Yoshinobu Inoue
af4f75caf0 Temporaly re-enable IPSEC and IPV6FIREWALL to make the kernel buildable
with INET6
(I'll fix those dependency later, Sorry)

PR:		kern/15053
Submitted by:	ssar@stacken.kth.se
1999-11-23 02:01:49 +00:00
Nick Hibma
1f1669ea57 Use releaseNo, not release. release is evil and should go away. 1999-11-23 01:16:10 +00:00
Brian Somers
2d904a8ba9 Oops, typo 1999-11-23 00:26:03 +00:00
Brian Somers
e6c5a96cee Document pppoed options 1999-11-23 00:22:58 +00:00
Brian Somers
6c62fd3e55 Add pppoed startup options 1999-11-23 00:22:25 +00:00
Brian Somers
97cba131dc Add a ``-P pidfile'' option 1999-11-23 00:21:20 +00:00
Brian Somers
be8bbe8809 Don't setuid() 'till we've called ID0setproctitle() 1999-11-23 00:14:50 +00:00
Cameron Grant
54671c035c fix panic for large writes in non-blocking mode 1999-11-22 21:16:01 +00:00
David E. O'Brien
ab473c8a50 Missed two.
Submitted by:	Philippe Charnier <charnier@xp11.frmug.org>
1999-11-22 20:03:01 +00:00
Nick Hibma
0ee137258d Flip the sequence of vendor, product around, to make it consistent
with the rest of the world. This avoids confusion when talking about
VID.PID.RID vs. PID.VID.RID. The former is the sequence normally used.
1999-11-22 19:08:36 +00:00
Nick Hibma
6739e1ad25 Clarify the functionality of usbd a bit more. 1999-11-22 19:04:58 +00:00
Peter Wemm
6e24b2a8e2 Fix (I believe) Bill Paul's core dump when using gdb on shared executables.
This is an addition to rev 1.5.  If lm == NULL, new->lm is uninitialized.
1999-11-22 18:04:53 +00:00
Cameron Grant
29618c3ae2 fix mixer to return properly scaled values 1999-11-22 16:20:48 +00:00
Peter Wemm
3ac8285030 Add pcm entries for the alpha too, since they were taken out of
conf/files and moved to i386/conf/files.i386

Forgotten by: tanimura
1999-11-22 16:10:26 +00:00
Peter Wemm
9565e721cb Update the Alpha makefile to be more in line with the x86 version. It
was getting quite dated.  Attempt to minimize diffs between the two so
it's not as painful to do this next time.
1999-11-22 15:39:13 +00:00
Poul-Henning Kamp
24e7ab7c0b Isolate the swapdev_vp "not quite" vnode in the only source file which
needs it now that /dev/drum is gone.

Reviewed by: eivind, peter
1999-11-22 15:27:09 +00:00
Peter Wemm
0171aaede2 Use %ll instead of %q as gcc moans bitterly about it. 1999-11-22 15:23:58 +00:00
Poul-Henning Kamp
8f52148e89 Isolate the swapdev_vp "not quite" vnode in the only source file which
needs it now that /dev/drum is gone.

Reviewed by: eivind, peter
1999-11-22 15:17:59 +00:00
Peter Wemm
33f86cee91 Zap some latent problems hidden by differences between KERNEL and _KERNEL.
The KAME code intruduces _KERNEL, which exposes some of them.
1999-11-22 15:14:56 +00:00
Nick Hibma
4239682e7f Update the manpage with the changed usb_devinfo structure. 1999-11-22 14:41:02 +00:00
Nick Hibma
28d38bdf15 Fix some bugs in user-end output and add a reference to the original
copyright in the resulting file.
1999-11-22 14:40:04 +00:00
Nick Hibma
c577d56d33 Feh, kind of went wrong the previous commit. dev should child (in some
cases) plus a typo.
1999-11-22 14:39:21 +00:00
Nick Hibma
4755ea009e Add MOUSE_MSC_MAXBUTTON and MOUSE_IF_USB. 1999-11-22 14:38:11 +00:00
Andrew Gallatin
4c6aa17cdc Allow config to generate proper ioconf.c files when devices request DMA
channel 0.

Submitted by: dfr
1999-11-22 14:31:55 +00:00
Andrew Gallatin
0e7e521cf3 Allow a DMA channel of 0. This gets the on-board mss audio device working
on Digital AlphaStations 200 and 500 machines (and probably others as well).

Submitted by: dfr
1999-11-22 14:30:41 +00:00
Yoshinobu Inoue
255717e065 Temporaly remove IPSEC and IPV6FIREWALL related options because they are not
ready for build yet.

Commented by: des
1999-11-22 13:25:21 +00:00
Eivind Eklund
637bc59616 Allow empty UIDs if we are processing NIS records. I am not entirely
happy with how this end up and will re-visit the entire empty field
problem, but this patch solves the NIS problem for now.

Submitted by:	Dan Nelson <dan@emsphone.com>
PR:	14865,14984
1999-11-22 12:42:38 +00:00
Yoshinobu Inoue
51f4c15279 move INET6 option from GENERIC to LINT.
Thanks for Brian Fundakowski Feldman and Dag-Erling Smorgrav,
to give me the comment and the patch.

Submitted by:Dag-Erling Smorgrav
1999-11-22 11:13:00 +00:00
Poul-Henning Kamp
38224dcd59 Convert various pieces of code to use vn_isdisk() rather than checking
for vp->v_type == VBLK.

In ccd: we don't need to call VOP_GETATTR to find the type of a vnode.

Reviewed by:    sos
1999-11-22 10:33:55 +00:00
David E. O'Brien
06adfd11a1 This commit was generated by cvs2svn to compensate for changes in r53574,
which included commits to RCS files with non-trunk default branches.
1999-11-22 10:31:17 +00:00