Commit Graph

2568 Commits

Author SHA1 Message Date
David Greenman
3d2a8cf3d9 Added a comment. 1995-03-11 22:29:07 +00:00
David Greenman
17dda4c929 Added some additional DIAGNOSTIC code that makes sure that freed
memory addresses and types are with the valid range. Increased
MAX_COPY to 256 (used to verify no freed memory use with DIAGNOSTIC).
1995-03-11 22:28:16 +00:00
David Greenman
a14e8fd032 Fix completely bogus comment. 1995-03-11 22:25:20 +00:00
David Greenman
63635f5ae2 Clear OBJ_INTERNAL flag for device pager objects and named anonymous
objects.
1995-03-11 22:25:02 +00:00
David Greenman
9a4bf98e66 Removed gratuitous and *extremely* evil setting of OBJ_INTERNAL. This
caused a cascade of problems including kernel memory corruption, file
corruption, system hangs, and panics.
1995-03-11 22:23:14 +00:00
Poul-Henning Kamp
5511b2e4f3 A substantial change to the way this works. Not the "TEST_LABELING" is
a option you can turn on and of on the fly, as is debugging.
1995-03-11 09:32:45 +00:00
Poul-Henning Kamp
fc20488fba Moved bb stuff to support.s per Bruces suggestion. 1995-03-11 03:49:50 +00:00
David Greenman
07be9bd45f Increased default minfree to 8%. 1995-03-10 22:18:16 +00:00
David Greenman
e23c0ff5a0 The threshold for switching from time-space and space-time is too small
when minfree is 5%...so make it stay at space in this case.

Submitted by:	Kirk McKusick
1995-03-10 22:11:50 +00:00
David Greenman
57677027ac Patch to fix quota panic from Mike Karels:
allow Q_SYNC regardless of "target" uid, we allow it with -1;
fix bug that caused all ops to refer to user quotas, not group.

Submitted by:	Mike Karels
1995-03-10 22:06:39 +00:00
David Greenman
f8a0b2dd88 Reorganized an if() expression for efficiency. 1995-03-10 21:18:24 +00:00
David Greenman
914e6eb70d Whoops, back out that last change - I misread what Poul had done there. 1995-03-10 20:29:51 +00:00
David Greenman
dbd90d413f Don't thrash the name cache while trying to fill up the object cache.
(Make a new cache entry until desiredvnodes is reached).
1995-03-10 20:26:29 +00:00
David Greenman
a74a73d96c Removed some #include's of unnecessary include files. 1995-03-10 08:44:20 +00:00
Poul-Henning Kamp
9eefa8bfc4 Add a dummy ___bb_init_func for BB profiling of the kernel.
To use this: recompile src/gnu/usr.bin/cc, compile your kernel.  The
files you want to profile should be compiled with '-a -g'.  "strip -x"
the kernel and run.  You don't need to profile all files in the kernel.
My next commit is the program to extract the data from the running kernel.
1995-03-10 08:41:26 +00:00
David Greenman
b4abbbcad3 kmem_alloc() returns zero-filled memory; it isn't necessary to bzero()
it.
1995-03-10 08:13:07 +00:00
David Greenman
90c47808bb Removed unnecessary routines vm_get_pmap() and vm_put_pmap().
kmem_alloc() returns zero filled memory, so no need to explicitly
bzero() it.
1995-03-10 08:05:00 +00:00
Poul-Henning Kamp
bff1f1d0dd Clean up and improve the namecache.
1. We always keep one 16th of the vnodes on the freelist, so that the
namecache doesn't get trashed.  It used to be that it wasn't a problem, but
the only vnodes getting released these days are directories and things which
Clean up and improve the namecache.

1. We always keep one 16th of the vnodes on the freelist, so that the
namecache doesn't get trashed.  It used to be that it wasn't a problem, but
the only vnodes getting released these days are directories and things which
gets forced out of the VM/cache.  The latter is not numerous enough to keep
the pool of vnodes needed for the namecache sufficiently big.

2. Purge invalid entries in the namecache as soon as we notice them.  This
avoids a stale entry pushing out a valid entry on the LRU list.

3. Speed up the lookup in the namecache by avoid a special case branch.

4. Make the cache purge routines do the thing they're supposed to, and in
a decently efficient manner.

5. Make the size of the namecache follow the number of vnodes, so that we
can always point to all the vnodes we have in core.

6. Readability has gone way up.

7. Added a "options NCH_STATISTICS" feature that will gather more
detailed statistics on the performance of the namecache.

Reviewed by:    davidg
1995-03-09 20:27:21 +00:00
Poul-Henning Kamp
fbd6e6c9ef Clean up and improve the namecache.
1. We always keep one 16th of the vnodes on the freelist, so that the
namecache doesn't get trashed.  It used to be that it wasn't a problem, but
the only vnodes getting released these days are directories and things which
Clean up and improve the namecache.

1. We always keep one 16th of the vnodes on the freelist, so that the
namecache doesn't get trashed.  It used to be that it wasn't a problem, but
the only vnodes getting released these days are directories and things which
gets forced out of the VM/cache.  The latter is not numerous enough to keep
the pool of vnodes needed for the namecache sufficiently big.

2. Purge invalid entries in the namecache as soon as we notice them.  This
avoids a stale entry pushing out a valid entry on the LRU list.

3. Speed up the lookup in the namecache by avoid a special case branch.

4. Make the cache purge routines do the thing they're supposed to, and in
a decently efficient manner.

5. Make the size of the namecache follow the number of vnodes, so that we
can always point to all the vnodes we have in core.

6. Readability has gone way up.

7. Added a "options NCH_STATISTICS" feature that will gather more
detailed statistics on the performance of the namecache.

Reviewed by:    davidg

(cvs is dumping core on me :-(  )
1995-03-09 20:27:04 +00:00
Poul-Henning Kamp
b2e10d6d6f Clean up and improve the namecache.
1. We always keep one 16th of the vnodes on the freelist, so that the
namecache doesn't get trashed.  It used to be that it wasn't a problem, but
the only vnodes getting released these days are directories and things which
gets forced out of the VM/cache.  The latter is not numerous enough to keep
the pool of vnodes needed for the namecache sufficiently big.

2. Purge invalid entries in the namecache as soon as we notice them.  This
avoids a stale entry pushing out a valid entry on the LRU list.

3. Speed up the lookup in the namecache by avoid a special case branch.

4. Make the cache purge routines do the thing they're supposed to, and in
a decently efficient manner.

5. Make the size of the namecache follow the number of vnodes, so that we
can always point to all the vnodes we have in core.

6. Readability has gone way up.

7. Added a "options NCH_STATISTICS" feature that will gather more
detailed statistics on the performance of the namecache.

Reviewed by:	davidg
1995-03-09 20:23:45 +00:00
Ugen J.S. Antsilevich
6a8cccb088 Remove redundant IORW definition..
because of this definition never ioctl's for sound
devices worked..For me this resulted also in  loss of
snd1 device...
1995-03-08 18:43:49 +00:00
Poul-Henning Kamp
a0e8a1e29b Another little optimization to the nameicache.
If an entry is stale, ditch it.
1995-03-08 01:40:44 +00:00
Poul-Henning Kamp
2425396b27 Improve the quality of the hash used in the namei-cache. 1995-03-08 01:08:03 +00:00
Søren Schmidt
edf16d1564 Added ata.c device driver file.
Just so I don't have to repatch files.i386 again & again...
1995-03-07 21:15:30 +00:00
David Greenman
d8cb3fb153 Increased number of buffers to 1/12 of (page_count - 1024). This makes the
cache minimum closer to 10% in the usual case.
1995-03-07 19:58:02 +00:00
David Greenman
5dcf309046 Removed most of the special policy regarding the seperation of VMIO and
dir/metadata buffers as it seems to have anomolous effects.
1995-03-07 19:53:27 +00:00
David Greenman
0007244260 Set VAGE flag when pager is destroyed. This usually happens when an
object has fallen off the end of the cached list - this is likely the
last reference to the vnode and it should be reused before non file
vnodes that are already on the free list (VDIR mostly).
1995-03-07 19:04:10 +00:00
David Greenman
79f7a9e1ef Added a new flag "VAGE" to indicate that the vnode should go on the head
of the free list.
1995-03-07 19:00:47 +00:00
David Greenman
acc835fd3f Put VAGE vnodes at the head of the free list. 1995-03-07 18:59:45 +00:00
David Greenman
f2da180f8e Fixed object reference count problem that occurred in the MAP_PRIVATE
case after we rewrote vm_mmap(). Added some comments to make it easier
to follow the reference counts.
1995-03-07 17:27:49 +00:00
David Greenman
be6d5bfac7 Don't attempt to reverse collapse non OBJ_INTERNAL objects. 1995-03-07 17:25:46 +00:00
Justin T. Gibbs
06f0e1ceb8 Be careful to maintain the bits in SBLKCTL that are reserved. 1995-03-07 09:00:44 +00:00
Justin T. Gibbs
7896fbec24 Compile cleanly -Wall. Remove some code that was an attempt to make
the 294x probe more stable, but since the guess was wrong, go back to
the clearer aproach I used the first time around.
1995-03-07 08:59:28 +00:00
Justin T. Gibbs
433c166775 Compile cleanly -Wall. 1995-03-07 08:58:22 +00:00
Peter Dufault
558b7c23d1 Submitted by: bde@zeta.org.au (Bruce Evans)
Fix it so that calls to extend_get won't dereference 0 when "swap
on sd.." calls sdsize() on systems with no SCSI devices.

This probably really fixes Poul's "boot with no SCSI" problem.
1995-03-06 15:02:13 +00:00
Ugen J.S. Antsilevich
7332d95a98 Restore IPACCT out there.. 1995-03-06 10:33:34 +00:00
Poul-Henning Kamp
30f467d84a Update vfs_cache.c to use the <sys/queue.h> macros. This makes it easier
to read, but doesn't change the speed.

Reviewed by:	phk
Obtained from:	via NetBSD
1995-03-06 06:45:52 +00:00
David Greenman
7c89bce067 Removed an old (commented out) call to cldisksort(). This has been
obsolete since 4.4 clustering.
1995-03-06 05:40:44 +00:00
David Greenman
e1acba5ab1 Remove some of my old debugging junk that somehow slipped in under my nose. 1995-03-06 05:36:59 +00:00
Nate Williams
7e58dfddc1 The new ld has apparently uncovered the bug that sio.c and tty.c
both define ttydefchars[] as globals.  This was fixed in 1.1.5
by making ttydefchars[] static like it should always have been.

Submitted by:	Bruce Evans <bde@zeta.org.au>
1995-03-06 02:50:49 +00:00
Nate Williams
9617d8b1f6 Removed unnecessary define for TCPOUTFLAGS since they are not used. 1995-03-06 02:49:24 +00:00
Joerg Wunsch
2389804061 Make `p'' an explicit alias for `print'' instead of being an
implicit alias for ``panic'', since this seems to be more rational.
1995-03-05 22:56:21 +00:00
Joerg Wunsch
d83a2a270d Update to beta3.20/b22
pcvt_ioctl.h doesn't belong to here.
1995-03-05 22:26:21 +00:00
Joerg Wunsch
1076e16083 Update to beta3.20/b22 1995-03-05 22:25:44 +00:00
Joerg Wunsch
0aff94e169 Make pcvt actually work. 1995-03-05 22:24:59 +00:00
Joerg Wunsch
68cfe626f7 pcvt is still using the XSERVER option; document this. 1995-03-05 22:23:35 +00:00
Steven Wallace
66e456c58e Revert to earlier code which contains FreeBSD snd[1-7] probe information,
$Id$ information, and other code to make sound driver compile and work
correctly with FreeBSD.

Integrate changes obtained from Sujal Patel.  These changes are:
 o  local.h:  reverse option logic from EXCLUDE_* to AUDIO_*
 o  pas2_mixer.c:  small addition
 o  ad1848.c:  minor change with macro names
 o  sequencer.c:  minor change with note check
 o  many spelling corrections in comments in about every other file
1995-03-05 22:11:57 +00:00
Jordan K. Hubbard
feadd38979 Reserve cdev 63 for Brian Litzinger & his Talisman Lite MPEG decoder
Submitted by:	Brian Litzinger <brian@MediaCity.com>
1995-03-05 21:41:40 +00:00
Peter Dufault
0f48de1c30 Fix when SCSI_DEBUG turned on. Thanks to Michael Reifenberger. 1995-03-05 20:01:44 +00:00
Jordan K. Hubbard
3c931b7071 Have dmabuf.c match the roll-back in soundcard.c. We should be making
better noises again.
Submitted by:	Sujal Patel <smpatel@wam.umd.edu>
1995-03-05 08:10:28 +00:00