316f45753f
Allow the DIAGNOSTIC to compile with old versions of gcc.
1996-05-17 20:53:11 +00:00
d0f2ced7d2
use %p not %x for pointers
1996-05-13 07:13:23 +00:00
6ddae4d140
Implement proper `..' locking in lookup code to avoid many potential
...
deadlocks.
copy up properly when adding a link to lower-layer only file.
(I believe there are still more deadlocks remaining.)
1996-05-10 22:57:49 +00:00
c9c8daaf7c
Add locking code to avoid deadlocks on vnode reclaim, which means the
...
addition of null_lookup, null_lock, null_unlock, null_islocked.
1996-05-10 22:50:45 +00:00
4ef330b934
remove include of <sys/cpu.h>
1996-04-22 01:38:12 +00:00
7e4d04910d
fix screw-up in the prototyping changes: print pointers with %p, NOT
...
by casting them to (unsigned int) then printing with %x.
1996-04-13 01:34:06 +00:00
2a0a8867ae
fix long-time bug in fdesc -- /dev/tty was a named pipe rather than a
...
mirror image of the real /dev/tty, a char dev. make it a char dev.
1996-04-11 11:24:29 +00:00
da749d6e09
Fix printf format follies.
1996-03-16 23:52:42 +00:00
c685105f07
Don't deref a bad ucred pointer, from Dave Carrel <carrel@cisco.com>,
...
PR #1699 .
1996-03-05 02:35:39 +00:00
183aec8fa8
GC *_nullop(). Minor nits.
1996-02-13 13:12:48 +00:00
5953a056eb
GC dead_nullop().
1996-02-13 13:02:40 +00:00
19dc07be15
close PR/2063: procfs_rw prototyped twice with different prototypes
1996-02-12 15:01:41 +00:00
631ccba638
miscfs prototype changes
1996-02-09 22:39:56 +00:00
fce3e0874f
Fix vop_link, vop_symlink, and vop_remove semantics in several ways:
...
* Change the argument names to vop_link so they actually make sense.
* Implement vop_link and vop_symlink for all file systems, so they do proper
cleanup.
* Require the file system to decide whether or not linking and unlinking of
directories is allowed, and disable it for all current file systems.
1996-02-09 18:47:48 +00:00
53fccab940
Fix vop_link, vop_symlink, and vop_remove semantics in several ways:
...
* Change the argument names to vop_link so they actually make sense.
* Implement vop_link and vop_symlink for all file systems, so they do proper
cleanup.
* Require the file system to decide whether or not linking and unlinking of
directories is allowed, and disable it for all current file systems.
1996-02-09 14:45:36 +00:00
299e1fe9df
Rename struct timespec fields to conform to POSIX.1b
1996-02-01 00:31:05 +00:00
5f8d113e8a
Don't include free blocks from the lower layer in statistics.
1996-01-30 16:45:10 +00:00
2d1739cbe1
Implement VOP_BWRITE() using vn_bwrite(), per r_friedl@informatik.uni-kl.de.
1995-10-15 05:18:12 +00:00
3dc2e10405
Fix the inode calculation in kernfs_getattr().
1995-10-09 14:25:02 +00:00
1e583dfb19
/dev/std* are of type DT_LNK.
1995-10-09 14:20:04 +00:00
d60bb4a465
Use the index number as the cookie, rather than multiplying by UIO_MX.
1995-10-09 14:03:32 +00:00
bcaf2bca34
Add support for cookies, mostly from Greg Hudson.
1995-10-09 11:18:51 +00:00
245f292fed
Prefix names of system call implementation functions with `sys_'.
1995-10-07 06:25:19 +00:00
d3f03a8c0c
Avoid reentering union_dircache() by locking the union vnode. From John Kohl.
1995-10-05 06:26:12 +00:00
36a94a031a
Lock the process in core before operating on it.
1995-08-13 09:05:51 +00:00
5482957905
splnet --> splsoftnet
1995-08-12 23:59:09 +00:00
e9d17d38b5
avoid unnecessary aging of buffers. This used to make sense, when buffer
...
caches were much smaller, but makes little sense now, and will become more
useless as RAM (and buffer cache) sizes grow. Suggested by Bob Baron.
1995-07-24 21:19:27 +00:00
3a99ca568b
this README has been unnecessary for a while.
1995-07-24 07:59:06 +00:00
c57d78c71a
Indirect off the right vnode when looking up the struct union_mount in
...
union_access(). Also, a trivial change to union_rmdir().
1995-07-13 13:19:18 +00:00
fd0f559215
add missing splx(), as suggested by enami@sys.ptg.sony.co.jp.
1995-07-08 00:42:45 +00:00
839598dfec
Make spec_read() and spec_write() vaguely consistent.
1995-07-02 07:20:50 +00:00
09e284bcca
Fix null VP references (pr 1124) - from John Kohl
1995-06-27 00:15:13 +00:00
5be81bd012
Fix null VP references (pr 1124)
1995-06-27 00:12:21 +00:00
8f62c773e8
don't assume the f_fsnamelen is nul-truncated or longer than MFSNAMELEN
1995-06-18 14:45:14 +00:00
e729deedb2
Fix a condition where union_updatevp() would reinsert the node onto the
...
wrong hash chain, by not swapping the hash values. Noted by John Kohl, but
fixed differently.
1995-06-02 02:39:20 +00:00
95ded74f58
Moved egid credential from cr_groups[0] to new field cr_gid. POSIX.1
...
requires that sgid executables and the setuid() syscall *not* change
the supplemental group list.
1995-06-01 22:43:30 +00:00
e9f4bff2cc
Add prototype.
1995-05-30 18:55:28 +00:00
2bf9d56861
Always call union_diruncache() when removing the upper object.
1995-05-30 18:52:39 +00:00
80d1ab77b5
Release the directory cache when a directory is removed or whited out.
1995-05-30 18:11:46 +00:00
0ca2e1bd7d
Fix typo.
1995-05-30 10:28:15 +00:00
7c7b203953
Make sure un_dircache is set after union_dircache(), to avoid stranding
...
references and the dircache buffer.
1995-05-30 09:37:02 +00:00
6760c93567
When a file is removed from the upper layer, make sure to release the
...
reference to it. From John Kohl.
1995-05-30 09:07:10 +00:00
eb310b67a7
Fix order of lock assertions in union_lock() and union_removed_upper().
...
Sprinkle a few vprint()s. From John Kohl.
1995-05-30 08:51:24 +00:00
df052e72a6
clean up some return-type warnings
1995-04-15 01:57:35 +00:00
13f5026553
fix timeval vs. timespec warnings
1995-04-15 01:56:43 +00:00
b88eec60bb
Allow opening a FIFO with O_RDWR.
1995-04-14 23:30:14 +00:00
99005388fd
Return EOF for old vnodes of tty devices, rather than EIO.
1995-04-10 00:48:46 +00:00
7f3a305b34
Use the new d_type field. Set VISTTY for vnodes of tty devices.
1995-04-10 00:48:10 +00:00
47ba263ebe
Emulate SCO behaviour when both FREAD and FWRITE are set, but only for SCO
...
executables.
1995-04-02 19:27:48 +00:00
6efcd1b6d4
KERNEL -> _KERNEL
1995-03-29 21:57:43 +00:00
aae5ebefc2
Add missing arguments to VOP_CLOSE() in some cases. From Frank van der
...
Linden.
1995-03-11 06:14:28 +00:00
2f805fa51b
copy*str() should use size_t.
1995-03-09 12:05:21 +00:00
aac6b08302
use u_long for copyin*
1995-03-08 01:33:20 +00:00
597bdad3df
Return EROFS rather than ENOENT in many cases. Also some cosmetic cleanup.
1995-02-03 16:18:46 +00:00
800be68af6
return EOPNOTSUPP from fhtovp and vptofh functions; doing otherwise
...
correctly is not possible.
1995-01-25 14:50:48 +00:00
e9017fd132
Clean up the code to frob mnt_stat a (tiny) bit.
1995-01-18 09:34:24 +00:00
ce22a1a443
initialize variable as pointed out by David Jones <dej@qpoint.torfree.net>
...
this should fix pr #699
1995-01-05 07:10:54 +00:00
635cd1625e
Consistency with other file system code...
1994-12-29 22:48:16 +00:00
d2e8ed4eec
Undo a few of the VCALL() changes for now.
1994-12-29 22:42:10 +00:00
609063ce2b
Format police.
1994-12-27 19:05:12 +00:00
2f0fb8ee09
Implement and use a common access checking routine
1994-12-24 16:44:12 +00:00
b52bb59b86
fix compilation error
1994-12-18 07:20:34 +00:00
dbe5168d35
Remove a_fp.
1994-12-16 03:18:03 +00:00
814ba0fc30
Call foo_statfs() from a common place when mounting.
1994-12-15 20:00:14 +00:00
9996e27813
Fix minor bug in union_setattr().
1994-12-15 19:15:06 +00:00
467e30fb72
Still more VOP_*() -> VCALL() twiddling.
1994-12-15 19:06:50 +00:00
0281dba8fa
Some more VOP_*() -> VCALL() munging.
1994-12-15 18:58:11 +00:00
7276c4ed44
Remove a_fp.
1994-12-14 18:45:21 +00:00
ef575bdbc1
Revert dup handling.
1994-12-14 18:40:27 +00:00
33c3d64e0c
Clean up deleted files.
1994-12-14 16:31:22 +00:00
82ab4c687f
Sync with CSRG.
1994-12-14 16:30:40 +00:00
a18ec3cb1b
LEASE_CHECK -> VOP_LEASE
1994-12-13 21:52:35 +00:00
bc7ed2596e
Fix editing errors.
1994-12-13 20:27:37 +00:00
527b796ff5
Turn lease_check() into a vnode op, per CSRG.
1994-12-13 20:14:30 +00:00
f821e3012a
Use VCALL() in lots of places.
1994-12-13 16:59:50 +00:00
2981ef6595
Sync with CSRG.
1994-12-13 09:58:11 +00:00
b05e959a14
Use fddupopen(), just like fdopen() does.
1994-12-04 03:13:06 +00:00
261751e81d
Reimplement union_seek().
1994-12-01 16:40:34 +00:00
1056700333
Make sure averunnable.fscale is filled before using it.
1994-12-01 10:16:53 +00:00
a1dbf4aaa5
fixed struct comment
1994-11-14 06:08:21 +00:00
923c9743a2
fixed struct comment; passed extra argument (struct file *) to open
1994-11-14 06:07:45 +00:00
8cdc99a059
fixed struct comment
1994-11-14 06:04:31 +00:00
6ac2bbfc35
be more careful with types, also pull in headers where necessary.
1994-10-30 21:43:03 +00:00
be8c7b9d80
light clean; make sure headers are properly included, types are OK, etc.
1994-10-29 07:58:24 +00:00
6b86130410
update for new syscall args description mechanism
1994-10-20 04:22:35 +00:00
a3e9e8091b
fix device aliasing and lost vnode problems.
1994-09-20 06:43:00 +00:00
76e0a7d0f4
Fix typo.
1994-09-15 03:47:19 +00:00
cab4d25c24
stat the file system at mount time, for `df -n', et al.
1994-09-15 03:42:37 +00:00
2f710a54ee
Convert process, file, and namei lists and hash tables to use queue.h.
1994-08-30 03:04:28 +00:00
f5faab89e9
Use VCALL(), not VOCALL(), and simplify union_advlock().
1994-08-30 03:01:01 +00:00
537ac5465f
Convert hash tables.
1994-08-19 11:25:29 +00:00
41761064b8
Implement /kern/msgbuf.
1994-07-21 10:22:20 +00:00
6436952bbc
Fix a null pointer dereference during rename(2).
1994-07-20 07:37:25 +00:00
988b7f5158
Support for block special files with sector sizes other than DEV_BSIZE -
...
if the device has a disklabel with a non-zero sector size value, that
value is used instead of DEV_BSIZE.
1994-07-16 11:40:56 +00:00
cb07d61edf
Fix a fencepost error.
1994-07-14 20:58:28 +00:00
cf92afd66e
New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD'
1994-06-29 06:29:24 +00:00
ab119e2183
Another update from JSP.
1994-06-17 15:21:33 +00:00
18157b489f
Remove an unneeded test.
1994-06-16 06:01:19 +00:00
82ac72f664
Merge update from JSP.
1994-06-15 23:07:54 +00:00
117527f322
Minor update from JSP after merging my changes.
1994-06-15 22:54:38 +00:00
76ee6346df
Deal with silly DIAGNOSTIC check.
1994-06-15 17:52:32 +00:00
76fe13be8e
Fix a bug pointed out by JSP.
1994-06-15 14:40:32 +00:00
7162b2c482
Use vget() for devices.
1994-06-15 03:30:44 +00:00
2e6e470fba
Fix a bug in finding the raw root device.
1994-06-15 03:05:26 +00:00
e4c7b2cebe
Clean up deleted files.
1994-06-13 17:48:29 +00:00
cde1d47595
Update to 4.4-Lite fs code, with local changes.
1994-06-08 11:33:09 +00:00
ebcaebd335
MIN -> min, MAX -> max
1994-05-24 02:35:53 +00:00
3f3d8b3889
Really fix the file size problem.
1994-05-17 06:59:17 +00:00
fa475bdca1
actually set va_size!
1994-05-17 04:10:29 +00:00
d9f8c89a76
setrun rename
1994-05-07 01:15:38 +00:00
a0a7429482
lots of changes: prototype migration, move lots of variables, definitions,
...
and structure elements around. kill some unnecessary type and macro
definitions. standardize clock handling. More changes than you'd want.
1994-05-05 05:35:42 +00:00
17f30a5d0b
kill obvious bug; glad to know this was tested!
1994-05-04 05:41:55 +00:00
4bbabd8e7a
Rename a lot of process flags.
1994-05-04 03:41:12 +00:00
d071d1cf05
some prototype cleanup, eliminate/replace bogus types (e.g. quad and
...
u_quad) -> use better types (e.g. quad_t & u_quad_t in inodes),
some cleanup.
1994-04-25 03:49:27 +00:00
4917d8beec
make fs types consistent over new kernels. also, some proto foo.
1994-04-23 07:54:38 +00:00
3dda0064a5
Convert mount, vnode, and buf structs to use <sys/queue.h>. Also,
...
some knf and structure frobbing to do along with it.
1994-04-21 07:47:31 +00:00
914b690745
forgot these...
1994-04-15 20:49:27 +00:00
93159ea799
fs types are names now.
1994-04-14 04:05:45 +00:00
656b05ccf6
be a bit smarter about determining if files shouldn't be seen by the user.
...
Also, DON'T allow a lookup to succeed on a file that's not visible!
1994-04-12 02:55:51 +00:00
bf860bb2b2
PG_COW -> PG_COPYONWRITE to match earlier changes in vm_page.h.
1994-03-17 04:10:32 +00:00
45afe06953
I now understand why name caching was redone for 4.4BSD; Net/2's
...
method of caching names has a totally bogus interface. kill *ALL*
caching of names when going through a lofs, because that's the only
way to make sure things always work right. Also, XXX the possibly-bogus
check i keep whining about. it's almost certainly wrong, but it's
too weird to modify significantly; even jsp didn't want to touch it...
1994-02-25 23:41:56 +00:00
7f77e98242
getting closer; stilly iffy though. 4.4's VFS would make this easier
1994-02-25 01:05:32 +00:00
349d681690
Undo last change; executables is `file', not `a.out'.
1994-02-15 13:32:39 +00:00
1df94d108f
Add .. entry to /kern
1994-02-14 19:46:18 +00:00
0cb2a819b2
Rename file -> a.out
1994-02-14 19:08:36 +00:00
4eab6179ba
Don't try to show a file for a process if there is none
1994-02-14 19:07:27 +00:00
bb5d305ea8
fix typo and allow creations to work right. this is in a questionably
...
correct chunk of code; author contacted.
1994-02-14 19:02:54 +00:00
b41fd90fdc
don't give bogus return code from read()
1994-02-11 01:01:51 +00:00
d43a0f37b1
don't panic if user does 'cat /kern', though result is dubious.
1994-02-09 06:32:17 +00:00
2df8c5f528
If you add files, be sure to have enough bits to encode an inode number!
1994-02-06 14:01:16 +00:00
89a86e618e
make a fpregs file.
1994-01-28 07:03:16 +00:00
7df895cf91
oops; fix that last...
1994-01-27 03:42:02 +00:00
5e12573ac2
hack from Mike Karels to deal with the last close on a controlling
...
terminal. from 4.4BSD.
1994-01-27 03:16:40 +00:00
2cee3e073f
Make procfs really work for debugging.
...
Implement not & notepg files in procfs.
1994-01-20 21:22:47 +00:00
bff20215e7
Make a printf #ifdef PORTAL_DIAGNOSTIC.
1994-01-13 18:29:03 +00:00
6570678929
fix botch in last
1994-01-12 20:01:39 +00:00
dc254a5c58
update to portal credentials handling, from jsp
1994-01-12 19:50:22 +00:00
5f7ef81c6f
Fix ugliness left over from my last mod
1994-01-11 18:02:06 +00:00
a24f445d60
Fix sign extension bug
1994-01-10 20:47:39 +00:00
c4f31d72f8
Add a missing break so my machine doesn't panic.
1994-01-10 04:58:14 +00:00
64aaba1a4e
fix some of my more recent botches, and clean up slightly.
1994-01-09 23:57:56 +00:00
5c3b96cd71
oops. fix that last
1994-01-09 20:17:06 +00:00
9f5ed8f1ef
minor cleanup; kill a few assignments
1994-01-09 20:10:51 +00:00
6202bab2db
Bug fixes and enhancements:
...
Make NFS serving work (BUT DON'T USE "attach" TO /proc/*/ctl FOR NOW!!!)
Make `curproc' a symbolic link
Add `.' and `..' entries to the directories.
Return better guesses on the size of the files.
1994-01-09 19:44:03 +00:00
903f51e908
Note that NFS mounting of fdesc doesn't make sense
1994-01-09 17:33:06 +00:00
e2efe41199
reorganization of ptrace/procfs code
1994-01-08 10:47:02 +00:00
eda5beab5a
return size of 'reg' from getattr()
1994-01-05 21:56:02 +00:00
454f368245
don't call VFS_SYNC() on underlying fs
1994-01-05 15:11:51 +00:00
058b0b8772
don't try to reclaim 'known' root vnode
1994-01-05 15:11:32 +00:00
c78cf001a2
add new portal file system from jsp@sequent.com
1994-01-05 14:23:21 +00:00
3442768ab8
add new lofs (loopback file system) from jsp@sequent.com
1994-01-05 14:15:31 +00:00
b4e04dd05d
fix return type for fifo_print
1994-01-05 11:34:44 +00:00
9d1babb688
fix UFS vs 'real' fs type mixups
1994-01-05 11:07:31 +00:00
1e5c7f966b
Clean up deleted files.
1994-01-05 11:05:17 +00:00
87693669dd
update with latest kernfs file system from jsp@sequent.com
1994-01-05 11:04:54 +00:00
4942f8a8bc
update with latest fdesc file system from jsp@sequent.com
1994-01-05 09:00:57 +00:00
3ff83b3549
make it compile (cleanly) for us
1994-01-05 08:00:05 +00:00
3bd8a96c64
Clean up deleted files.
1994-01-05 07:51:37 +00:00
699e3be9e9
add new procfs code, from Jan-Simon Pendry, jsp@sequent.com.
...
This is pretty-much "virgin", so that diffs can be done later.
1994-01-05 07:51:08 +00:00
76cce4cfdd
fix fdesc_print return type (again)
1993-12-23 17:37:33 +00:00
ae30022085
fix return type of vnode print routine
1993-12-22 13:14:10 +00:00
cc4170c743
fix return type of dead_print
1993-12-22 13:12:09 +00:00
86f960201b
pull load average and misc changes down from magnum
1993-12-20 12:39:10 +00:00
21edb9924e
Canonicalize all #includes.
1993-12-18 03:59:02 +00:00
0560892d34
Canonicalize all #includes.
1993-12-18 03:48:27 +00:00
2b3cb4657d
#ifdef out /kern/root at theo's request
1993-11-20 01:57:07 +00:00
76dbc1192b
new specfs.h and fifo.h locations
1993-11-12 05:54:12 +00:00
5e6a2e33c0
fix chris typo.
1993-10-30 16:11:07 +00:00
06c4106aec
limit block sizes requested
1993-10-29 19:25:18 +00:00
4f10c7ead8
kill volatile warning.
1993-09-16 22:08:30 +00:00
053f138dae
Changes to VFS readdir semantics
...
NFS changes for better cookie support
ISOFS changes for better Rockridge support and support for generation numbers
1993-09-07 15:40:14 +00:00
dd119ec84b
Implement setattr: mode for process entries; mode + uid/gid for the
...
PROCFS root directory.
Fixed omission in pfs_root() which came to light as a result of the above:
hold on to vnode for root dir.
1993-08-26 19:01:00 +00:00
ac815031d0
Um, last change was wrong. Instead, add 3 to the number of inodes (forget
...
about the root directory, too).
1993-08-25 14:41:19 +00:00
b3f22bcdcc
Subtract to from the free count for `.' and `..', to maintain the fiction that
...
this is a real file system.
1993-08-25 14:35:20 +00:00
1e37ac6db4
Fixed improperly initialized nfsnode in pfs_lookup()
1993-08-25 09:28:39 +00:00
8dc35fb5d6
Fill inode fields in procfs_statfs(), in stead of block fields
1993-08-24 17:49:16 +00:00
834a44f21f
copyright update.
1993-08-24 16:47:23 +00:00
0cbe77737a
Rcs Id added.
1993-08-24 16:25:07 +00:00
498189f379
Initial version of a proc filesystem.
1993-08-24 12:59:09 +00:00
742dcd5655
RLIMIT_OFILE --> RLIMIT_NOFILE
1993-08-23 16:01:56 +00:00
4543c1ebfd
changes from 0.9-ALPHA2 to 0.9-BETA
1993-08-23 04:59:38 +00:00
daa27b3c30
Make fdesc_print have a return type of void.
1993-08-02 23:01:49 +00:00
b4d9151681
Make kernfs_print have a return type of void.
1993-08-02 23:00:57 +00:00
4862b84c92
Add RCS identifiers (this time on the correct side of the branch), and
...
incorporate recent changes in netbsd-0-9 branch.
1993-08-01 19:22:24 +00:00
0e6cb953ea
ANSIfications - removed all implicit function return types and argument
...
definitions. Ensured that all files include "systm.h" to gain access to
general prototypes. Casts where necessary.
1993-06-27 06:01:27 +00:00
7b40766e79
ANSIfications - lots of function prototyping.
1993-06-27 05:58:59 +00:00
c88530d318
give various filesystems their own vnode types
1993-06-07 05:25:16 +00:00
c2bdec030c
give miscfs filesystems their own mount structure malloc type.
1993-06-07 04:56:35 +00:00
6b5412e4fa
add some more functionality: a setattr which supports chmod+chown+chgrp,
...
and the various support elsewhere to deal with it.
1993-05-28 16:47:02 +00:00
84a68c1180
add kernfs_access function, to kill kernfs security hole
1993-05-28 14:12:17 +00:00
7cef199976
make the sleeps on socket open interruptable.
1993-05-27 15:38:18 +00:00
cd1bcbc8ca
add rcs ids as necessary, and also clean up headers
1993-05-20 16:21:42 +00:00
868a2c5982
header cleanup
1993-05-20 04:01:30 +00:00
230dcf0d05
add $Id$ strings, and clean up file headers where necessary
1993-05-20 02:54:09 +00:00
6e88da586a
fix several off-by-one errors in hostname setting/reading
1993-04-27 10:06:10 +00:00
484e228599
Use EPERM when write permission is denied, not EBADF.
1993-04-27 07:21:37 +00:00
bcbf85df19
Add consistent multiple-inclusion protection.
1993-04-19 01:22:17 +00:00
b02974b037
Jay Fenlason <hack@datacube.com>:
...
1: the fi_readers and fi_writers fields of the fifoinfo structure were not
being initialized to 0. This caused the driver to not sleep the first
process to open the fifo--it thought there was already another process to
talk to (most of the time.)
2: fifo_open() was calling tsleep() without unlocking the inode of the fifo
file. This caused *any* subsequent access to the file (even an ls (!)) to
hang forever. Note that this bug was usually masked by bug #1 above.
1993-04-02 20:19:48 +00:00
d7743f6ac4
added . and ..
1993-03-30 05:16:56 +00:00
91a1e3ca32
added cdevvp (after taking it out of vfs_subr.c) because realized
...
that it was only needed here.
1993-03-27 02:00:45 +00:00
gwr