summaryrefslogtreecommitdiffstats
path: root/sys/miscfs (follow)
Commit message (Collapse)AuthorAgeFilesLines
...
* Rename vop_revoke, vn_bwrite, vop_noislocked, vop_nolock, vop_nounlockcsapuntz1998-08-0612-401/+139
| | | | | | | | | | | | to be vop_generic_revoke, vop_generic_bwrite, vop_generic_islocked, vop_generic_lock and vop_generic_unlock. Create vop_generic_abortop and propogate change to all file systems. Fix PR/371. Get rid of locking in NULLFS (should be mostly unnecessary now except for forced unmounts).
* Fix locking discipline snafu with UNIONFS. Thanks to Don Schmid (yensid@imsa.edu) for the original fix and Todd Miller.csapuntz1998-07-131-2/+4
|
* securelevels do NOT protect running binaries; only filesystem activityderaadt1998-06-271-8/+1
|
* return EOPNOTSUPP for chflags() changes; peter@freebsdderaadt1998-06-113-5/+24
|
* do not permit ptrace attach to immutable executablederaadt1998-06-091-1/+8
|
* Use __alpha__ and __mips__ instead of alpha & arcniklas1998-03-011-2/+2
|
* mips and alpha GCC does not unwind the stack beyond the current frameniklas1998-03-011-4/+7
|
* Please GCC 2.8 -Wall + some KNFniklas1998-02-231-7/+4
|
* Updates to match type changes in syscall tablestholo1998-02-087-21/+21
|
* fix a nasty code typo; rotel@indigo.iederaadt1997-12-301-2/+2
|
* I should be more careful about what I check in. Undo earlier commenting-outcsapuntz1997-12-281-5/+1
| | | | | of security check. Restricted access to /dev/memat higher secure levels now enabled.
* Fixed null_inactie so it odes the right thing.csapuntz1997-12-271-1/+5
|
* Removed i386-specific crudcsapuntz1997-11-061-93/+4
|
* Updates for VFS Lite 2 + soft update.csapuntz1997-11-0629-953/+899
|
* back out vfs lite2 till after 2.2deraadt1997-10-069-489/+602
|
* back out vfs lite2 till after 2.2deraadt1997-10-0620-410/+351
|
* VFS Lite2 Changescsapuntz1997-10-0629-953/+899
|
* Fix {KERNFS,NULLFS,UMAPFS,UNION}_DIAGNOSTIC kernel compilation errors.millert1997-09-119-55/+54
| | | | From Alan Barrett <apb@iafrica.com> with some changes.
* From NetBSD (thorpej):millert1997-08-296-21/+26
| | | | | Fix a reversed argument which caused procfs_checkioperm() to always return OK". Add a few comments to avoid further confusion.
* 1) pfs_mode should be mode_t, not u_shortmillert1997-08-167-47/+81
| | | | | | | | | | | 2) Fix procfs security hole. This is basically the NetBSD fix (which is based in part on Sean Eric Fagan's FreeBSD fix) with a few minor changes. The change creates as function, procfs_checkioperm(), that has check numbers 3 and 4 from sys_ptrace(). 3) Make procfs_control() use procfs_checkioperm() as well (it already had the same basic checks but some of them were incorrect). 4) Minor stylistic cleanups, including the use of the SET/CLR/ISSET macros to aid general grokability.
* From NetBSD (mycroft):millert1997-08-015-41/+38
| | | | | Eliminate bogus uses of V{READ,WRITE,EXEC}. Use S_I[RWX]{USR,GRP,OTH} where appropriate.
* block all write ops to init in securelevel modes; nash@mcs.comderaadt1997-06-211-1/+4
|
* struct definitions in .h files and instances in .c file please.kstailey1997-04-262-24/+27
| | | | | | | | | | | | Vic Able: > While <miscfs/kernfs/kernfs.h> has a kernfs_node, it depends > on a structure that's defined in <miscfs/kernfs/kernfs_vnops.c>. > Oops! That's not very social. :-) [I built a new kernel fine, but don't have time to run "make build" it is my intuition that this will not have any bad effects, but I didn't test. Just back it out if it hurts you and I'll try again.]
* Don't use gcc builtin __builtin_return_address(frameno) onmillert1997-04-101-2/+3
| | | | | alpha, it crashes the kernel. This is a work around until gcc is fixed. NetBSD PR#3471 fix from Ross Harvey <ross@teraflop.com>
* Fix improper cast (compiler warning).millert1997-04-061-2/+3
|
* Make kern.osrevision (sysctl) and /kern/osrev report value ofmillert1997-03-271-2/+2
| | | | OpenBSD define, not BSD which is a bit more useful.
* OpenBSD tagniklas1997-02-241-0/+1
|
* IPSEC package by John Ioannidis and Angelos D. Keromytis. Written inderaadt1997-02-201-2/+15
| | | | Greece. From ftp.funet.fi:/pub/unix/security/net/ip/BSDipsec.tar.gz
* Don't fault a page in twice (first prefault, then vm_fault); the first wastholo1997-01-221-5/+1
| | | | not neccecary (and actually caused some problems).
* fixed warning:kstailey1997-01-151-2/+2
| | | | kernfs_vnops.c:305: warning: long unsigned int format, u_int arg (arg 3)
* Fix for final ptdi panic on i386tholo1997-01-071-2/+6
|
* spec_advlock() via lf_advlock()kstailey1997-01-042-7/+8
|
* pulled out the duplicated, conditional code from both kern/vfs_syscalls.cmickey1997-01-021-1/+62
| | | | | | | | and compat/common/vfs_syscalls_43.c and placed a single copy of that code into miscfs/union/union_subr.c (seemed like a good place to put it, since it's union-fs related). as a side effect you can build unionfs in lkm. (netbsd pr#2950, Paul Goyette <paul@pgoyette.bdt.com>)
* make spec_advlock() perform indirect call to lf_advlock()kstailey1996-12-221-14/+15
|
* fixed comment in union_advlockkstailey1996-12-081-1/+2
|
* mem leak, short malloc; netbsd pr#3000, minoura@kw.netlaputa.or.jpderaadt1996-12-071-2/+6
|
* fix locking on devices to work on diskless workstationskstailey1996-12-051-3/+9
|
* remrq -> remrunqueuekstailey1996-11-231-2/+2
|
* added const to second parameter of cfprint_t routineskstailey1996-11-231-1/+5
|
* We should return 0 on non-blocking read from FIFO with no writerstholo1996-11-041-1/+4
|
* permit locking on devicesderaadt1996-11-041-2/+3
|
* and the same for the trunkmickey1996-10-271-2/+2
|
* mimic ptrace behaviour; worked out with bitbltderaadt1996-10-141-4/+25
|
* Boy do I feel like a fool. I added a bugfix to Theos' changes, and thenbitblt1996-09-271-15/+1
| | | | | | | | discovered that I had commited a bunch of nonfunctional changes Theo had made, but not debugged. This undoes both my changes, and Theo's. When this code is functional, it will be put back in. Thank you for pointing this out to me David Mazieres.
* It should be possible to trace init if securelevel is at 0, single-user mode,bitblt1996-09-261-1/+16
| | | | without having to peg the system at securelevel -1.
* still handle OLD_PIPE casederaadt1996-09-261-1/+5
|
* Bring pipe_stat prototype in scopeniklas1996-09-101-1/+6
|
* Understand DTYPE_PIPEtholo1996-09-071-1/+25
|
* -Wall & -Wstrict-prototype fixesniklas1996-07-021-3/+1
|
* use <sys/queue.h> for pfsnode list.mickey1996-06-216-22/+17
| | | | $OpenBSD$ additions.