Jeffrey Altman [Mon, 10 Dec 2007 20:33:34 +0000]
windows-buffers-
20071210
LICENSE MIT
after removing cm_buf_t objects from the hash tables, clear the HT list
pointers to make debugging easier.
when modifying data version values in buf_ForceDataVersion() grab a
read lock on buf_globalLock instead of a write lock. A write lock is
only required if the hash table linked lists are being modified.
The dataVersion itself is protected by the cm_scache_t mx mutex being
held.
Carsten Jacobi [Mon, 10 Dec 2007 18:26:30 +0000]
pam-expires-setting-should-be-32bit-
20071210
LICENSE IPL10
store as a 32 bit value explicitly so 32 bit routines don't leave garbage around here.
Mike Meffie [Mon, 10 Dec 2007 18:20:50 +0000]
uss-lex-more-forward-decls-
20071210
LICENSE IPL10
forward declare this so picky compilers will do their jobs.
Derrick Brashear [Sun, 9 Dec 2007 23:33:42 +0000]
cnvldb-prototypes-
20071209
LICENSE IPL10
forward declare these
Derrick Brashear [Sun, 9 Dec 2007 06:05:51 +0000]
aklog-no-dummy-write-on-aix5-
20071209
LICENSE IPL10
don't do the dummy write on aix5. we are better off actually working
Derrick Brashear [Sat, 8 Dec 2007 17:57:33 +0000]
writedcache-enforce-xdcache-writelock-
20071208
LICENSE IPL10
FIXES 59136
in order that the writes to CacheItems not be addled, we need to *actually*
hold a *write* lock when we write. not even just read as the comment says.
and certainly not none as was happening in the dir package.
Derrick Brashear [Sat, 8 Dec 2007 17:44:00 +0000]
linux24-kernel-configure-test-
20071208
LICENSE IPL10
on linux 2.4 the try kbuild macro needs to try harder to notice what
symbols we don't have
Jeffrey Altman [Fri, 7 Dec 2007 21:32:53 +0000]
windows-wix-afslogon-
20071207
LICENSE MIT
Move afslogon.dll and afscpcc.exe to the OpenAFS\Client\Program directory
Jeffrey Altman [Fri, 7 Dec 2007 16:03:56 +0000]
windows-follow-backup-path-
20071207
LICENSE MIT
Add a registry value, FollowBackupPath, that provides the Windows cache
manager with functionality equivalent to the UNIX afsd -backuptree option.
Jeffrey Altman [Wed, 5 Dec 2007 20:38:38 +0000]
viced-cbstuff-
20071205
LICENSE IPL10
Keep track of AddCallBack calls in cbstuff.
Increment cbstuff.DeleteCallBacks after holding H_LOCK
Jeffrey Altman [Wed, 5 Dec 2007 20:31:56 +0000]
cbd-
20071205
LICENSE IPL10
Windows requires open() to be called with O_BINARY otherwise Ctrl-Z
means EOF which breaks dumpfile processing.
Be consistent about writing/reading time as afs_uint32 instead of
writing as afs_uint32 and reading as time_t since time_t can be 32-bit
or 64-bit depending on the platform.
Derrick Brashear [Tue, 4 Dec 2007 21:08:10 +0000]
largefile-avoid-wraparound-write-to-old-server-
20071204
LICENSE IPL10
FIXES 73720
apply analogous check that's in fileserver to clients writing to old servers
Derrick Brashear [Tue, 4 Dec 2007 21:01:12 +0000]
inode-convertro-dont-include-volser-code-in-viced-
20071204
LICENSE IPL10
FIXES 77574
viced isn't an fssync client; avoid the convert volser code there
Derrick Brashear [Tue, 4 Dec 2007 20:39:18 +0000]
linux-nfs-translator-dont-lose-nfs-pag-on-write-
20071204
LICENSE IPL10
avoid issue where nfs writes through the translator could become unpag'd because a background daemon came into play
Derrick Brashear [Tue, 4 Dec 2007 20:18:50 +0000]
afsdb-avoid-cross-task-stack-use-
20071204
LICENSE IPL10
really for aix, but simplified code universally.
Jeffrey Altman [Tue, 4 Dec 2007 15:44:56 +0000]
windows-giveupallcallbacks-
20071204
LICENSE IPL10
Add a registry value, GiveUpAllCallBacks, that can be used to enable/disable
the use of the GiveUpAllCallBacks RPC. Due to the fact that calling this RPC
can destablize pre-1.4.6 file servers since they do not include DELTA
giveupallcallbacks-locking-
20071121, the default is being set to disabled.
We can re-enable the functionality by default only after sufficient time
has been given for orgs to update their file servers.
Tracy Di Marco White [Sat, 1 Dec 2007 05:40:51 +0000]
nbsd-new-version-
20071201
LICENSE IPL10
FIXES 75602
add support for new version(s) of netbsd
Derrick Brashear [Fri, 30 Nov 2007 19:13:38 +0000]
aix-pin-small-allocs-
20071130
LICENSE IPL10
we assume the small alloc pool is pinned. maybe we should pin it.
Hartmut Reuter [Thu, 29 Nov 2007 21:37:00 +0000]
namei-allow-ufs-
20071129
LICENSE IPL10
allow ufs again for namei. oops.
Jeffrey Altman [Thu, 29 Nov 2007 20:53:53 +0000]
windows-aklog-no-panic-if-no-kfw-
20071129
LICENSE MIT
Jeffrey Altman [Thu, 29 Nov 2007 19:25:04 +0000]
viced-drop-core-on-shutdownandcore-
20071129
LICENSE IPL10
FIXES 74708
when we call ShutDownAndCore, actually drop a core.
Jeffrey Altman [Thu, 29 Nov 2007 18:40:12 +0000]
clean-felater-when-rechaining-
20071129
LICENSE IPL10
FIXES 74708
in order that this not be continually rechained, clear this here; we
are handling it now.
Derrick Brashear [Thu, 29 Nov 2007 18:32:29 +0000]
zero-uio-on-zero-read-
20071127
LICENSE IPL10
avoid leaking random data in a read of a 0 length object.
Jeffrey Altman [Thu, 29 Nov 2007 16:10:24 +0000]
windows-volstat-headers-
20071129
move the enum volstat definition to cm_volstat.h from cm_volume.h.
fix the copyright.
Jeffrey Altman [Thu, 29 Nov 2007 16:06:48 +0000]
windows-smb-out-of-memory-
20071129
It turns out that LsaCallAuthenticationPackage can succeed but still
fail with an extended error. The one case that we have seen reported
is STATUS_NO_MEMORY which would be produced when the paging file is
full. This could be a transient error that will be resolved after
Windows increases the paging file size.
Instead of crashing under this condition we should force the authentication
to fail and permit the client to retry.
WER
567093715-1-
0405174244
Jeffrey Altman [Thu, 29 Nov 2007 16:00:19 +0000]
windows-daemon-shutdown-
20071129
The cm_Daemon() thread did not check after each operation whether or
not the service is in the process of shutting down. As a result the
thread could access data, locks, etc. that had been freed by other
threads.
WER
572098933-1-
0407879071
Jason Edgecombe [Thu, 29 Nov 2007 05:05:37 +0000]
man-page-symlink-
20071128
Add man pages for all the symlink commands (Windows-only).
Russ Allbery [Thu, 29 Nov 2007 04:57:05 +0000]
man-page-vos-convertrotorw-fix-
20071128
Don't say that inode file servers are only available on Solaris and AIX.
Jason Edgecombe [Thu, 29 Nov 2007 04:17:10 +0000]
man-page-fs-sysname-improvements-
20071128
Update the fs sysname man page to document the behavior of multiple @sys
values.
Jason Edgecombe [Thu, 29 Nov 2007 03:54:44 +0000]
man-page-fs-rxstatpeer-
20071128
New man page for fs rxstatpeer.
Steven Jenkins [Thu, 29 Nov 2007 03:32:52 +0000]
man-page-fileserver-auditlog-
20071128
Remove duplicate -auditlog in the fileserver synopsis.
Jason Edgecombe [Thu, 29 Nov 2007 03:30:41 +0000]
man-page-additions-
20071128
Add new man pages for vos convertROtoRW, vos copy, and read_tape and add
a troubleshooting section to the fileserver man page.
Russ Allbery [Thu, 29 Nov 2007 02:56:13 +0000]
man-fs-uuid-tweaks-
20071128
Some minor capitalization and formatting tweaks to the fs uuid man page.
Derrick Brashear [Wed, 28 Nov 2007 05:03:03 +0000]
getpag-prototype-
20071127
forward declare this function
====================
This delta was composed from multiple commits as part of the CVS->Git migration.
The checkin message with each commit was inconsistent.
The following are the additional commit messages.
====================
returns void, but, whatever
Mike Meffie [Tue, 27 Nov 2007 20:44:47 +0000]
linux-try-kbuild-deal-with-old-modpost-
20071127
the pre 2.6.20.1 modpost issued warnings and not errors for unresolvable symbols.
not overly useful.
Derrick Brashear [Tue, 27 Nov 2007 20:01:00 +0000]
macos-no-read-on-directories-
20071127
thanks to quentin@mit.edu for pointing this out.
no read() on directories
Marc Dionne [Tue, 27 Nov 2007 19:26:42 +0000]
linux-hash-inodes-for-mmap-
20071127
FIXES 78544
inodes are only added to the superblock dirty list if hashed.
do so.
Asanka Herath [Tue, 27 Nov 2007 17:28:44 +0000]
windows-nim-afscred-double-free-
20071127
fix a double free on error condition
Derrick Brashear [Mon, 26 Nov 2007 22:19:02 +0000]
redhat-cleanup-authlibs-correctly-
20071126
when not packaging authlibs, correctly discard them
Derrick Brashear [Mon, 26 Nov 2007 22:10:34 +0000]
volprocs-alloc-vnode-instead-of-using-stack-
20071126
don't use stack here; alloc memory for short term use
Rainer Toebbicke [Mon, 26 Nov 2007 21:47:00 +0000]
namei-avoid-spurious-emfile-
20071126
FIXES 77531
before we did not handle the fd cache runnign dry. "oops"
Christof Hanke [Mon, 26 Nov 2007 21:03:31 +0000]
assorted-warning-cleanup-
20071126
add missignheader files, missing return values, correct a few calls to printf
Christof Hanke [Mon, 26 Nov 2007 20:38:14 +0000]
kauth-ubik-wrap-pad-requisite-args-out-
20071126
of course, if Date is larger than 32 bits, the wrapper is broken anyway. we should really get rid of the wrapper
Derrick Brashear [Fri, 23 Nov 2007 14:29:51 +0000]
enhance-afsdb-debugging-
20071123
allow cmdebug to show locks on cells. add fstrace event for afsdb
Sean O'Malley [Fri, 23 Nov 2007 13:55:49 +0000]
krb5-klog-solaris-
20071123
FIXES 77575
make this work on solaris
Marc Dionne [Fri, 23 Nov 2007 13:37:24 +0000]
linux-2624-
20071123
FIXES 77370
support linux 2.6.24
Derrick Brashear [Fri, 23 Nov 2007 13:00:42 +0000]
macos-buildpkg-cleanup-cruft-
20071123
don't leave the old dmg around since it makes a new one bomb
Jeffrey Altman [Wed, 21 Nov 2007 19:30:51 +0000]
giveupallcallbacks-locking-
20071121
FIXES 74708
Hold H_LOCK across calls into GiveUpAllCallbacks_r
Jeffrey Altman [Wed, 14 Nov 2007 17:18:25 +0000]
windows-dv-logging-
20071114
64-bit data version logging updates
Jeffrey Altman [Wed, 14 Nov 2007 06:22:04 +0000]
windows-64bit-data-version-
20071114
Add support for 64-bit AFSFetchStatus data version values
Derrick Brashear [Wed, 14 Nov 2007 04:34:53 +0000]
dafs-inode-convert-
20071113
the dafs version of this
Derrick Brashear [Tue, 13 Nov 2007 22:15:03 +0000]
inode-convert-ro-to-rw-
20071113
so this functionality works with inode fileserver, also. corrected abstraction from previous versions so vol and volser have correct bits.
Derrick Brashear [Tue, 13 Nov 2007 00:46:09 +0000]
redhat-spec-no-strip-
20071112
suggested by simon. debuginfo support in rpmbuild strips for us. deal here.
Simon Wilkinson [Tue, 13 Nov 2007 00:29:31 +0000]
redhat-reunrevert-fedora-build-change-
20071112
"oops"
Derrick Brashear [Tue, 13 Nov 2007 00:00:57 +0000]
redhat-packaging-updates-
20071112
FIXES 76890
various updates to deal with fedora 8 and remove obsolete stuff
Jeffrey Altman [Mon, 12 Nov 2007 19:03:15 +0000]
viced-accurately-track-file-callbacks-
20071112
FIXES 74708
this should fix the issue with cbfree being corrupted. a more exhaustive pointer-blunting patch should probably be pulled in, but, this is simple and obvious
Kevin McBride [Mon, 12 Nov 2007 18:25:54 +0000]
autoconf-strip-debug-options-sanity-
20071112
To prevent stripping, specify the '--disable-strip-binaries' option on
the ./configure command line. fileserver and volserver are never stripped.
When --enable-debug is specified, binaries will not be stripped by default.
Tom Keiser [Mon, 12 Nov 2007 18:17:43 +0000]
dafs-optimize-shutdown-
20071112
optimize shutdown case as is done elsewhere, for dafs.
Jeffrey Altman [Mon, 12 Nov 2007 03:24:59 +0000]
windows-readonly-volume-callbacks-
20071111
disable the option by default. require the daemonCheckVolCBInterval
registry value be set in order to activate it.
do not renew .readonly volumes whose callbacks are less than 30 minutes
old.
Russ Allbery [Mon, 12 Nov 2007 00:12:28 +0000]
man-page-name-underscore-more-
20071111
Missed fs apropos when fixing NAME entries to eliminate spaces.
Russ Allbery [Sun, 11 Nov 2007 22:54:56 +0000]
man-page-name-underscore-
20071111
The NAME heading for man pages can't contain a space in the program side
or the man pages won't index with some man implementations.
Jeffrey Altman [Sat, 10 Nov 2007 22:50:44 +0000]
windows-readonly-volume-callbacks-
20071110
avoid a deadlock
Simon Wilkinson [Sat, 10 Nov 2007 20:13:27 +0000]
libafsrpc-makefile-fix-
20071110
FIXES 76836
fix double / in makefile
Dan Hyde [Sat, 10 Nov 2007 00:33:41 +0000]
viced-missing-lock-
20071109
FIXES 76713
Add missing lock that needs to protect an in use 'host' from having
HOSTDELETED added to the hostFlags and HWNO_INPROGRESS removed.
Jeffrey Altman [Sat, 10 Nov 2007 00:18:11 +0000]
cmdebug-human-readable-expiration-time-
20071109
add -ctime parameter which produces human readable expiration times.
Doesn't work with OpenAFS for Windows prior to 1.5.28 because time()
was not used in those versions for tracking expiration times.
Jeffrey Altman [Sat, 10 Nov 2007 00:14:26 +0000]
windows-readonly-volume-callbacks-
20071109
A .readonly volume callback applies to the entire volume. Track it
in the cm_volume_t cbExpiresRO field and apply it to all cm_scache_t
objects with valid callbacks that are associated with that volume upon
each daemon callback check. This will prevent premature callback
expiration.
Also, attempt to automatically refresh the callbacks every 30 minutes
by obtaining a callback on the volume root. This value is configurable
with the "daemonCheckVolCBInterval" registry value.
Change from osi_Time() to time() for expiration values in order to
permit conversion to human readable values in cmdebug.
Derrick Brashear [Fri, 9 Nov 2007 20:15:46 +0000]
clean-platform-
20071109
clean platform dir
Derrick Brashear [Fri, 9 Nov 2007 19:47:18 +0000]
macos-packaging-updates-
20071109
make the readme correct. make dmgs which have a useful name
Derrick Brashear [Fri, 9 Nov 2007 19:18:49 +0000]
hpux11-thread-lock-tracking-
20071109
do this for hpux also
Derrick Brashear [Thu, 8 Nov 2007 14:38:48 +0000]
darwin-vfs-context-pullback-
20071108
again, commit fix we have in 1.4
Derrick Brashear [Thu, 8 Nov 2007 14:26:47 +0000]
reconvert-aklog-to-afs-com-err-
20071108
make this match what we're shipping in 1.4
Derrick Brashear [Thu, 8 Nov 2007 14:21:06 +0000]
aix5-revert-kcred-
20071108
pull this back to 1.5 and head
Derrick Brashear [Wed, 7 Nov 2007 04:23:47 +0000]
shutdown-cleanup-not-yet-
20071106
Jeffrey Altman [Wed, 7 Nov 2007 04:01:36 +0000]
linux-nsec-timestamp-zero-
20071106
don't leave garbage in nsec
Anders Kaseorg [Tue, 6 Nov 2007 18:30:19 +0000]
linux-splice-support-
20071106
support splice in kernels new enough to support it (sendfile is deprecated)
Derrick Brashear [Mon, 5 Nov 2007 22:09:12 +0000]
aklog-no-kimpersonate-in-more-cases-
20071105
aix's NAS kerberos can't support kimpersonate. deal
Derrick Brashear [Mon, 5 Nov 2007 21:43:00 +0000]
krb5-klog-no-err-table-yet-
20071105
right now we don't have this everywhere, until we reorder libraries, and
test usefully with configure, forget it.
Thomas Mueller [Mon, 5 Nov 2007 20:27:24 +0000]
cbd-use-callback-size-for-callbacks-
20071105
FIXES 76408
cast correctly
Jeffrey Altman [Mon, 5 Nov 2007 18:25:33 +0000]
windows-bplus-tree-
20071105
Migrate search key into thread local storage
Jeffrey Altman [Mon, 5 Nov 2007 18:23:30 +0000]
windows-afsd-notification-
20071105
fidp->cell is not a pointer
Jeffrey Altman [Mon, 5 Nov 2007 00:23:54 +0000]
windows-afsd-notification-
20071104
the difference between osi_assert() and osi_assertx() is whether or not
a description is specified as part of the call. When no message is
specified afsd_notifier() is called with a NULL msgp parameter. This
results in a null pointer dereference during the EventLog() call.
As a result, none of the information describing the assertion is written
to the log file.
This commit sets a default message for use when no other message is
provided and it replaces all calls to osi_assert with osi_assertx and
adds descriptions.
Jeffrey Altman [Mon, 5 Nov 2007 00:16:58 +0000]
windows-osi-lock-assert-msg-
20071104
modify lock_AssertXXX macros to call osi_assertx() and provide a
descriptive message.
Asanka Herath [Sat, 3 Nov 2007 16:31:50 +0000]
windows-scache-syncop-waiters-
20071103
One of the issues that has become a serious problem since the addition
of the local directory updates is that although cm_SyncOp synchronizes
operations, it does not preserve the order of requests. This has always
been a problem in that it has been possible for a request to fail to
complete due to its worker thread's bad luck. When a request takes
longer than the Windows SMB Redirector's timeout, the SMB Redirector
tears down the SMB virtual circuit.
When using the local directory updates it is really important that
the directory update operations complete in the order that they were
sent to the file server. If they don't, then the local directory
state and the file server state will not match and the local directory
state must be discarded which in turn forces a new read of the entire
directory contents over the network.
This patch adds a new cm_scache_waiter_t object that is used to store
the current thread, buffer, and syncop flags within a waiters queue
on each cm_scache_t object. If a thread is forced to sleep in cm_SyncOp,
upon waking it will check to see if there are any other threads waiting
that are attempting to perform a similar task ahead of it in the queue.
If yes, the thread goes back to sleep. If not, it goes ahead and
enters the cm_SyncOp conflict resolution block.
This patch has the additional side effect of reducing the number of
competing threads that must obtain the cm_scache_t mutex and process
the cm_SyncOp conflict resolution block. As a result, the overall
CPU utilization of the service and the clock time associated with
processing requests will be reduced.
Jeffrey Altman [Sat, 3 Nov 2007 16:18:14 +0000]
windows-dir-
20071103
Reorganize the locking for cm_BeginDirOp and cm_EndDirOp.
There are a number of locations where locks are obtained, dropped, and
reobtained. This reorganization attempts to accomplish several things:
(1) be optimistic for the most common case so it will be fast
(2) add consistency checks after each location where locks are dropped
and re-obtained. If we lose a race in cm_BeginDirOp and the bplus
tree is out of date, retry until we get to a consistent state
that we can use.
(3) Ensure that all operations take place with the correct locks.
Jeffrey Altman [Sat, 3 Nov 2007 16:08:03 +0000]
windows-buf-
20071103
assert that the cm_scache_t mutex is held when calling
buf_ForceDataVersion().
Jeffrey Altman [Sat, 3 Nov 2007 16:05:43 +0000]
windows-bplus-tree-
20071103
rename findNode to leafNode in bplus_Lookup
replace all OutputDebugString calls with osi_LogX calls
modify bestMatch to special case the return values for leaf nodes.
If an entry is above or below the values available in the leaf node
return BTLOWER or BTUPPER instead of BTERROR.
In insert and delete operations check for BTLOWER/BTUPPER and isleaf,
if true convert to either slot 0 or Max and perform the insertion.
This produces easier to read code when performing lookups.
Jeffrey Hutzelman [Sat, 3 Nov 2007 15:59:12 +0000]
windows-cellservdb-
20071103
GCO Public CellServDB 25 Oct 2007
Jeffrey Altman [Sat, 3 Nov 2007 01:05:53 +0000]
windows-install-des-prototypes-
20071102
required by auth-getkey-prototyping-
20071102
Derrick Brashear [Fri, 2 Nov 2007 18:01:32 +0000]
auth-getkey-prototyping-
20071102
make this sane
Ken Raeburn [Thu, 1 Nov 2007 19:12:38 +0000]
viced-new-host-reup-h-lock-
20071101
FIXES 69384
don't leave lock dropped
Dale Ghent [Thu, 1 Nov 2007 17:05:33 +0000]
solaris-cc-deal-with-new-64-bit-arch-specifier-
20071101
FIXES 75908
switch from -xmarch=amd64 to -m64
Dale Ghent [Thu, 1 Nov 2007 16:57:26 +0000]
solaris-nevada-update-
20071101
FIXES 75896
update for current solaris nevada
Simon Wilkinson [Thu, 1 Nov 2007 16:09:22 +0000]
auth-cleanup-
20071101
kill warnings in auth directory
Simon Wilkinson [Thu, 1 Nov 2007 15:58:59 +0000]
rxgen-ansi-again-
20071101
another look at lettign rxgen do ansification
Tom Keiser [Thu, 1 Nov 2007 14:59:38 +0000]
dafs-salvage-deal-with-clones-
20071101
FIXES 75591
cope with trying to salvage a clone correctly
Chaskiel M Grundman [Thu, 1 Nov 2007 03:23:09 +0000]
linux-afs-unlinked-file-dentry-flagging-
20071031
fix case:
perl -e 'open FOO, ">foo"; unlink(foo); @stat = stat(FOO); print "inode for foo(unlinked) is $stat[1]\n"; system("/bin/ls -ali | grep $stat[1]");'
Simon Wilkinson [Wed, 31 Oct 2007 22:31:38 +0000]
more-anonymous-pointer-cleanup-
20071031
more char*->void* conversion
Jeffrey Altman [Wed, 31 Oct 2007 15:33:27 +0000]
windows-btree-
20071031
Add additional validation and error handling code after each call to
getSlot(). If an invalid slot is returned, return NONODE. If the
invalid slot is returned when extracting a data node, invalidate the
tree.
Modify compareKeys() to always perform a case-insensitive comparison
and only perform a case sensistive comparison if the case-insensitive
one matches. This ensures the ordering is consistently reported.
Add lock assertions to ensure that all calls are being performed with
the correct locks being held. There have been some crash reports that
provide stack data that does not appear to be possible unless there is
a race. However, there are no obvious locations where the race is
taking place and the test suite indicates that all of the correct locks
are being held. We shall see what happens in the field.
For consistency replace all calls to findKey in which the range is
(1,numentries) with calls to getSlot().
Optimize the depth search loop by testing the slot value in the for
statement instead of forcing the loop to be broken later.
Jeffrey Altman [Wed, 31 Oct 2007 15:23:42 +0000]
windows-begindirop-
20071031
Avoid a race condition in cm_BeginDirOp() caused by the failure to
hold the cm_scache_t mutex while copying status data from the scp to
the dirop
Jeffrey Altman [Wed, 31 Oct 2007 15:20:14 +0000]
windows-cm_analyze-ewouldblock-
20071031
Add support for EWOULDBLOCK to cm_Analyze. If the file server returns
EWOULDBLOCK, retry the request every two seconds for up to the RDRtimeout.