Commit Graph

49 Commits

Author SHA1 Message Date
fvdl 1cf3a3db94 * Implement optional 32 <-> 64 bit directory cookie translation. This uses
the directory cache as translation table. See nfs_subs.c for comments.
  Makes the code a bit more complex to look at than I would have liked,
  but doesn't affect the speed of the default behavior.
* Optimize caching behavior a bit when buffers are invalidated.
* Save some RPCs in readdir operations by not bothering if there is
  a small amount left to do to fill the buffer. It'll be done in the
  next RPC with a larger chunk anyway. Wastes a bit of buffer space
  but is faster.
* Make n_vattr an allocated vattr struct. This avoids nfsnode bloat,
  and is friendlier to the malloc routines.
1997-10-19 01:46:15 +00:00
fvdl 01e7c03e2e Move cookie heuristic function inside ifdef NFS, to make a kernel with server
code but without client code link again. From Erik Bertelsen, PR 4259
1997-10-11 02:09:48 +00:00
fvdl 8aa50b9159 Fix unitialized var warning (did not appear on i386, but did on sparc). 1997-10-10 13:21:51 +00:00
fvdl 7f7d814e67 * New directory entry caching system. Provides full caching of any
directory cookie that may be thrown back at us from userspace, up
  to a size limit. Fixes double entry problem.
* Split flags for internal and external use in the NFS mount structure.
* Fix some buffer structure fields that weren're being used correctly.
* Fix missing directory cache inval call in nfs_open.
* Limit on NFS_DIRBLKSIZ no longer needed, bumped to the more reasonable
  value of 8k.
* Various other things that I forget, all related to the dir caching
  somehow, though.
1997-10-10 01:53:17 +00:00
fvdl 2c5bca83ed Don't assume that pointers into mbuf data remain valid across nfsm_dissect.
In readdirplus, don't keep such pointers but store the file attributes
in a variable instead until they are needed. Change nfsm_loadattr*
a bit so it can accept a direct pointer to an nfs_fattr structure.
1997-07-14 20:46:20 +00:00
drochner 8c0a9bab00 Don't cast 64bit (off_t) file sizes to vm_offset_t (32bit on many
architectures), truncate them intelligently instead.
The truncation is done centralized in vnode_pager.c.
This prevents from wrap-over effects when parts of large (>2^32 byte) files
are mmapped.
Don't allow to mmap above the numerical range of vm_offset_t.
This is considered a temporary solution until the vm system handles the
object sizes/offsets more cleanly.
1997-07-04 20:22:09 +00:00
fvdl 28e824c9b7 Extend lookup handling for WebNFS. This means that nfs_namei deals
with full pathname lookups if a public filehandle is used, and that
it translates the '%' escapes (URL-style) in the same case. Also,
make nfsrv_fhtovp convert the public filehandle to the vp of the
publicly exported filesystem, as stored in the nfs_pub structure.
1997-06-24 23:36:02 +00:00
mycroft e3f99a9397 Pass the vnode type to vaccess(), and use it when checking VEXEC. Make sure
that the mode bits passed to vaccess() and returned by foo_getattr() contain
only permission bits.
1997-05-08 16:19:43 +00:00
thorpej 22ea1671cf Update for new mbuf code. 1997-03-27 20:40:09 +00:00
fvdl 17d7bd9d41 Check for the use of reserved ports on a per-request basis, unless
MNT_EXNORESPORT is specified. The check is cheap and doesn't impose
any extra overhead.
1997-03-23 20:55:51 +00:00
fvdl 03fd36b882 Fixes from BSDI (thanks go to Keith Bostic). Original RCS messages:
date: 1997/02/10 18:41:14;  author: cp;  state: Exp;  lines: +110 -46
Make nfs_realign go away on sparc and add functionality to nfsm_disct.
1997-02-22 03:03:03 +00:00
thorpej dc33841893 NFSCLIENT -> NFS. 1997-01-31 02:58:48 +00:00
fvdl 2914743038 Comment change in previous made for some bad english.. 1996-12-09 11:31:11 +00:00
fvdl 46795a244d Move '#ifdef NFSSERVER' back to the right spot; NQNFS datastructures need
to be initialized on both the client and the server side. Remove misleading
comment about this being just server stuff.
1996-12-09 10:46:20 +00:00
thorpej f02e8b3cf0 Make NFSSERVER work without NFSCLIENT. This is achieved by splitting
the client and server/shared data initialization into separate functions,
and calling the server/shared initialization directly from main().
Problem noted in PR #1308 (Kenneth Stailey) and PR #1780 (Chris Demetriou).
Fix suggested in PR #1780 by Chris Demetriou, and munged a bit by me,
and OK'd by Frank van der Linden <fvdl@netbsd.org>.
1996-12-03 00:22:43 +00:00
thorpej aa25de69d1 NFS performance improvement from Doug Rabson/FreeBSD:
Improve the queuing algorithms used by NFS' asynchronous i/o.  The
existing mechanism uses a global queue for some buffers and the
vp->b_dirtyblkhd queue for others.  This turns sequential writes into
randomly ordered writes to the server, affecting both read and write
performance.  The existing mechanism also copes badly with hung
servers, tending to block accesses to other servers when all the iods
are waiting for a hung server.

The new mechanism uses a queue for each mount point.  All asynchronous
i/o goes through this queue which preserves the ordering of requests.
A simple mechanism ensures that the iods are shared out fairly between
active mount points.

Reviewed/integrated/approved by Frank van der Linden <fvdl@netbsd.org>
1996-12-02 22:55:39 +00:00
cgd ce53653783 make the namei struct members ni_dirp and ni_next, and the componentname
struct member cn_nameptr 'const', since they should never be used to
modify the path name.  (Only the pathname buffer, cn_pnbuf, should be
modified.)  Propagate the const poisoning to code that uses the namei
and componentname structs.
1996-10-25 23:13:58 +00:00
christos bcf4552fb1 revert kprintf changes 1996-10-13 01:39:03 +00:00
christos 117a51c5a5 printf -> kprintf, sprintf -> ksprintf 1996-10-10 23:31:17 +00:00
fvdl 88eff7ea92 Start XIDs at a value based on the current time, not 0. This avoids nasty
XID confusions with servers that cache them over a long period and
with clients that reboot quickly.

Problems: because of the sanity check that is done by comparing the system
time with filesystem time, XIDs will start at 0 until root is mounted,
which means it isn't completely safe for diskless setups. But it's clearly
better than it was. It would also be cleaner if all XID handling (more
generally, all RPC handling) within the kernel went through the
same functions.
1996-07-07 12:23:49 +00:00
fvdl b27bafebd5 We're only handling uio with iovcnt == 1, so don't ever attempt to increment
uio_iov, this will get us into nasty trouble. (Thanks to Matthias Drochner for
tracking this down).
1996-07-01 10:22:47 +00:00
fvdl 2847784948 * Make mounts with symlinks work (needed for direct mounts with amd). PR #1917
* Never change the NQNFS flag and/or version when just doing an update mount.
  Fixes a problem that made diskless booting impossible under some
  circumstances.
1996-05-23 22:47:27 +00:00
thorpej ea39811fd8 Make these link in the absense of "options FIFO". 1996-04-03 23:25:36 +00:00
fvdl f5e6df1ed0 Disable invalidating of directory offsets cookies. Should fix one or two
directory problems.

XXX There is no clean solution to the cookie/cookieverifier validity mess.
    Together with the disabled strict cookie check, this puts us back at
    what v2 did in this case. Slightly better solution possible by
    consequently storing 64bit cookies in other places too.
1996-03-13 00:44:17 +00:00
fvdl 5ac7df1caf Bring in a merge of Rick Macklem's NFSv3 code from Lite2 1996-02-18 11:53:36 +00:00
christos e4c93ec893 nfs prototype changes 1996-02-09 21:48:19 +00:00
jtc 299e1fe9df Rename struct timespec fields to conform to POSIX.1b 1996-02-01 00:31:05 +00:00
cgd 3e0efb4d89 changes to make this work on systems where pointers & longs are 64 bits.
This is mostly just changes to make the stuff that goes over the wire
use fixed-size types.
1995-12-19 23:07:11 +00:00
ws 4d035ffc4d No point in computing cn_hash here,
as it gets immediately recomputed in lookup
1995-09-08 13:52:23 +00:00
mycroft f5f32d8328 Exported group list now starts at offset 0, not 1. 1995-06-02 18:35:50 +00:00
jtc 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
mycroft f7c13d44bd More LIST/CIRCLEQ migration. 1994-08-18 22:47:43 +00:00
mycroft 8743009944 Convert some more lists and queues. 1994-08-17 14:43:47 +00:00
mycroft dbd174a5da Change the reply list to a TAILQ. 1994-08-17 11:41:36 +00:00
mycroft 9c55cb1c79 Set the group list length when copying credentials. 1994-07-22 23:16:36 +00:00
cgd fccfa11af5 New RCS ID's, take two. they're more aesthecially pleasant, and use 'NetBSD' 1994-06-29 06:39:25 +00:00
mycroft fee90785c2 Move a misplaced #endif. 1994-06-13 16:00:41 +00:00
mycroft cde1d47595 Update to 4.4-Lite fs code, with local changes. 1994-06-08 11:33:09 +00:00
ws ce516ff13d Make FFS optional 1994-03-09 21:21:37 +00:00
mycroft 805c52eecb Eliminate some more uses of b_actl. 1994-02-06 11:28:37 +00:00
mycroft 95b048b53a Canonicalize all #includes. 1993-12-18 00:40:47 +00:00
ws 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
mycroft 6518b0ac4a Make bpos arg to nfsm_reqh a caddr_t*, not a caddr_t**, as that's what it
is actually passed.
1993-08-02 23:11:14 +00:00
cgd a70246c1fa bpos is really a caddr_t **. doesn't really make a diff to the code
generated...
1993-07-13 10:40:07 +00:00
cgd c58c7e0a97 diskless changes made last time were hosed; were using NULL for
"no credentials" rather than NOCRED.
1993-07-13 10:04:26 +00:00
cgd 0b1df311fc changes from ws to support diskless booting... these are "OK" on inspection
and after testing...  (actually, currently, none of the changed
code is even used...)
1993-07-07 12:06:32 +00:00
cgd 769818382a add rcsid again; fix RCS+crash fuckup 1993-05-21 07:31:58 +00:00
glass 24e91b0789 migrated code to make split possible 1993-04-10 20:19:19 +00:00
cgd 61f282557f initial import of 386bsd-0.1 sources 1993-03-21 09:45:37 +00:00