1 Mon Oct 18 17:43:29 1999 H.J. Lu <hjl@lucon.org>
3 * nfs-utils.spec (Obsoletes): Fixed a typo, change knfsd-client
7 Mon Oct 18 14:56:22 1999 H.J. Lu <hjl@lucon.org>
9 * Initial version 0.1 released.
11 * configure.in (VERSION): Set to "nfs-utils 0.1".
12 * configure: Regenerated.
14 Mon Oct 18 14:54:57 1999 H.J. Lu <hjl@lucon.org>
16 * utils/mountd/mountd.c (get_exportlist): Cleanup.
18 * utils/exportfs/exportfs.c (unexport_all): Unexport from
19 kernel only if the entry is exported to kernel.
20 (unexportfs): Likewise.
22 Wed Sep 08 16:49:32 1999 Neil Brown <neilb@cse.unsw.edu.au>
24 1/ utils/mountd/rmtab.c::mountlist_list
26 This routine stats the rmtab file to see if it has changed. It
27 if has, it cleans up it's old copy of the data. But it still
28 always re-read the file, thus returning multiple copies of the
29 data on consecutive calls without intervening changes.
30 "Showmount -a" didn't show this as it appears to sort/unique the
31 data, but 'strace showmount -a' showed that the size of the
32 datagram that it received grew.
34 I moved the getrmtabent loop inside the mtime test.
36 2/ utils/exportfs/exportfs.c
38 Many routines used the m_path field of m_export instead of
40 According to the comment in nfslib.h, m_path should only
41 be used when processing a mount request (i.e. in mountd)
42 where the mountpoint may be a subdirectory of the export point.
44 I changed all occurances of m_path to e_path
47 3/ utils/exportfs/exportfs.c:main
49 extra arguments are not meaningful with -a or -r, but
50 exportfs accepted them and then ignored the -a/-r, expect that
51 -r would still unexport everything first.
53 I generate an error if there are extra args and f_all
55 4/ utils/exportfs/exportfs.c:main
56 extract dump out as a special case.
58 5/ utils/exportfs/exportfs.c
59 made f_reexport a local variable.
62 6/ utils/exportfs/exportfs.c:main,exportall
64 support/export/rmtab.c
65 only mayexport on newly created entries, don't set xtabent at all
67 7/ support/include/nfslib.h
69 add #define _PATH_PROC_EXPORTS to be /proc/fs/nds/exports
71 8/ support/export/xtab.c
73 xtab_mount_read loads data from _PATH_PROC_EXPORTS if it exists,
77 9/ support/export/xtab.c
79 xtab_mount_read now sets m_exported, and NOT
82 removed the append arguement from xtab_write as it was
85 added is_export flag to xtab_write similar to xtab_read
86 if is_export, only write entries with m_xtabent or m_addxtab
87 if !is_export, only write entries with m_exported
89 10/ support/export/export.c::export_allowed_internal
91 added test for exp->m_mayexport, as the export tree
92 may have entries that are no longer allowed to be exported,
93 and so shouldn't caused deduced exported by rmtab_read
95 11/ utils/exportfs/exportfs.c::main
96 error checking of flags.
98 12/ utils/exportfs/exportfs.c
100 total rewrite of export and unexport logic.
102 - build an exportslist of valid exports, based on
103 current etab file and arguments,
104 - read rmtab to instantiate relevant wild card entries
105 - read etab to find out what is currently exported
106 - synchronise intention with reality
107 - write out etab and xtab
110 discard the m_addxtab flag
111 add m_changed flag so we know what to report in exportfs
113 14/ utils/mountd/auth.c:auth_authenticate
115 the value returned by gethostbyaddr was trusted.
117 It now follows this with a call to gethostbyname
118 and checks that the address is in the list.
120 15/ support/export/nfsctl.c::cltsetup,expsetup
122 force client names to lowercase as kernel is
125 16/ quietened a few compiler warnings
127 17/ support/export/client:client_lookup
129 look for pre-existing client with same name before creating
132 18/ support/include/exportfs.h
134 The ordering of the MCL_* enum was:
135 ANONYMOUS, FQDN, SUBNETWORK, WILDCARD, NETGROUP
137 I moved ANONYMOUS to the end.
139 The ordering is significant when an export entry is being searched for to
140 match a given address. There are two problems with ANONYMOUS being first.
142 1/ if a directory is exported rw to a couple of hosts and ro to everyone else,
143 then the ro case will always be found first and the privileged hosts won't get
145 2/ When mountd gets a request to mount an ANONYMOUSly exported tree, it creates a FQDN
146 export entry for the specific host, and writes it to xtab.
147 When another request comes from the same host, the ANONYMOUS entry is found again,
148 before the new FQDN entry, so it creates another FQDN entry and writes it to xtab
149 again. If causes bloat in xtab.
151 Putting ANONYMOUS at the end reflects it's nature as a catch-all
153 19/ utils/exportfs/exportfs.man
154 many updates to the man page to reflect changes to the code
156 -----------------------
162 - allow exportfs to modify rmtab file
163 - make sure kernel never gets two clients with same IP address
164 - possible kernel should reject
165 - needs to be some way to lookup client in kernel by IP address
166 - maybe get kernel to do case-insensitive comparisons on client names
167 - remove unused clients from kernel
169 - change etab to xtab and xtab to xtab.active
171 - timestamp and/or statd-stamp in rmtab for removing old entries.
173 Mon Oct 18 11:48:07 1999 H.J. Lu <hjl@lucon.org>
175 * linux-nfs: New directory.
176 * linux-nfs/ChangeLog: Moved from ..
177 * linux-nfs/INSTALL: Likewise.
178 * linux-nfs/KNOWNBUGS: Likewise.
179 * linux-nfs/NEW: Likewise.
180 * linux-nfs/README: Likewise.
181 * linux-nfs/THANKS: Likewise.
182 * linux-nfs/TODO: Likewise.
184 * Starting from knfsd 1.4.7.