STABLE14-aklog-stds-krb5-order-20060802

kfm needs this first so when you include com_err.h from afs as a side effect of krb5.h we don't explode

(cherry picked from commit aebadcd5478994fec011368597004bef2439ccb6)

STABLE14-endian-fixes-20060802

macos needs us to do weird stuff so -arch i386 -arch ppc works

(cherry picked from commit fd099b3e763b6b2c5bfb87db0cae8a900f031ad3)

STABLE14-macos-use-platform-copy-of-afssettings-20060802

we moved this to be made for real; use it

(cherry picked from commit 4ea8536c908871f173d776fc475064761d47d85d)

STABLE14-macos-platform-update-20060801

minor cleanup of the platform stuff

(cherry picked from commit 4d67cb2f1645239d1ef73a8847b17275e4cb3feb)

STABLE14-darwin-platform-fix-20060801

get rid of tabs

(cherry picked from commit d1da98ac49860379f772bc27bdd86ac410592ad3)

STABLE14-openafs-stable-1-4-2b3-20060801

make openafs 1.4.2 b3

STABLE14-viced-capability-writelock-20060801

add write loc king capability

(cherry picked from commit 80e88d6a80fc3fc0d15650f20ea7fc3fece1a1e4)

STABLE14-macos-installer-20060801

partial updates

STABLE14-platform-build-fixes-20060801

Add dummy makefiles for all the other supported architectures and fix a
few issues with makefile inclusion in the Makefile.in and
DARWIN/Makefile.in.

(cherry picked from commit 088e1409f8d318697023341e2050a3e58708b0fd)

STABLE14-macos-move-afssettings-20060801

build this as part of the build

(cherry picked from commit 86a29ebd32fd9cc48e367930709276a7c489e085)

STABLE14-windows-server-ping-20060216

add logging of server addresses, status and capabilities

prevent multiple threads from pinging the same server simultaneously.
instead, threads block and then use the result of the ping that was
already taking place.

(cherry picked from commit 424ae50d1f7dfe0c3557d98f8ae9bc03a8265a21)

STABLE14-windows-largefile-support-20060623

This patch adds large file support (> 2GB) to the CIFS server and
the cache manager.

It also corrects a significant oversight where the definition of
struct rx_call was inconsistent due to a failure to define AFS_PTHREAD_ENV
prior to including rx.h in all source files.  This prevented all errors
during rx_Read and rx_Write to be ignored.

(cherry picked from commit 015f9ce287c22df7a0b967ee0357d0816b4e0e54)

STABLE14-ubik-call-sucks-20060704

make ubik_Call removal work on Windows.

(cherry picked from commit 941ac04165a9239a11b4ab908350dbfae24edfb9)

STABLE14-kreltime-guard-against-null-tm-20060731

so we can memset 0 and get defaults

(cherry picked from commit 67f04fa714580016bdbb707d962ce763023e31aa)

STABLE14-fslog-dpf-20060731

turns out it's nice to log why we exit

(cherry picked from commit 75922872e6c29bd059db3b07b7b274e3c4172715)

STABLE14-viced-write-locking-fix-20060425

According to src/libacl/prs_fs.h the PRSFS_LOCK privilege is only supposed
to control read-locks.  The PRSFS_WRITE and PRSFS_INSERT privileges are
supposed to control write-locks.

(cherry picked from commit ecdc5aca34b0f0dc6bf67eb95f8e6506c556ca26)

STABLE14-va-copy-instead-of-just-pointering-20060731

FIXES 35559

va_list isn't integral on ppc

(cherry picked from commit da4775909094e8dc1e14fa9cdbf8ba3baaebf23a)

STABLE14-aklog-man-page-20060731

Rewrite the aklog man page in POD and add documentation of the new flags
and .xlog.

RCS file: /cvs/openafs/src/aklog/Attic/aklog.1,v
Working file: src/aklog/aklog.1
head: 1.2
branch:
locks: strict
access list:
keyword substitution: kv
total revisions: 3; selected revisions: 0
description:

====================
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.
====================

Fix unterminated B<>.

(cherry picked from commit 5f434b8087d3c73123860f853d15a0f51c03bf73)

STABLE14-ubik-call-sucks-20060703

ubik_Call's pseudo-varargs construction is hideous

put it out of its misery where possible

STABLE14-macos-packaging-update-20060405

FIXES 29226

copy cellservdb if there is none

(cherry picked from commit b3120246765e9f5d144ac470b66269ccbbce48a9)

STABLE14-fileserver-hates-pruclient-20060626

right now there is one struct ubik_client shared by all threads in the fileserve
r, which sucks when one has it locked

let's, uh, fix that.

(cherry picked from commit 791b2141809fc0ad2da78ab3cd91303ac0ec482d)

STABLE14-ubik-call-sucks-20060703

ubik_Call's pseudo-varargs construction is hideous

put it out of its misery where possible

STABLE14-namei_mapper-20060731

add chaskiel's namei mapping tool to cvs and make a rule to build it

(cherry picked from commit 8c4496602badc1b0aa2642e061a720d4d1f108b8)

STABLE14-bucoord-64-bit-cleanup-20060701

However, in bucoord/commands.c, bc_VolRestoreCmd(), bc_DiskRestoreCmd(),
bc_VolsetRestoreCmd(), and bc_DumpCmd() were passing '0' as arguments
instead of NULL.  Because there is no prototype for bc_StartDmpRst() this
results in a crash on 64-bit Linux.  (when running 'volrestore' from an
interactive backup session)

STABLE14-fbsd-ifnet-flags-20060731

FIXES 35757

i should really fix this right

(cherry picked from commit c04c2a0b1aa266c67d2e7e056e6c3f9518a39a2a)

STABLE14-fbsd61-20060221

Param file for FreeBSD 6.1. Builds, untested.

(cherry picked from commit d7a60f50a40da4acc28fe7ebbd145a25a9ba8b6d)

STABLE14-libafsauthent-avoid-global-host-20060731

FIXES 36411

defining host globally in the rmtsysc interface: dumb

(cherry picked from commit e6f6adc1bc310482baa868046e68c386ab1c822b)

STABLE14-vos-avoid-freeing-null-pointers-20060731

FIXES 36441

pretty obvious. don't.

(cherry picked from commit 77c502047025d8073e4cd35c5e4851fd1d2eb369)

STABLE14-call-bzero-in-convertvolumeinfo-with-correct-size-for-dotreadonly-20060731

FIXES 36442

and slipstream in bzero() killing

(cherry picked from commit 8d2dbc51f40e61cbe6157c910dff18606cf84282)

STABLE14-rx-readpacket-leak-20060728

avoid just allocating packets forever

(cherry picked from commit e199597d529a035f50e57e7dd653a8095313d246)

STABLE14-windows-version-20060728

1.4.2-beta3 == 1.4.0153.0

STABLE14-windows-notes-20060728

updates for 1.4.2-beta2

STABLE14-windows-remove-bad-asserts-20060728

remove assert()s that should simply be if()

(cherry picked from commit a4211966d7f9d16dd2f93c9122ba09b4920dfd8e)

STABLE14-windows-integrated-logon-20060726

 delta windows-integrated-logon-20060724 would cause Windows 2000
 systems to blue screen with an exception in winlogon.exe.

(cherry picked from commit 3167f707fdf582c8ca851a353ca0c82b92ccc947)

STABLE14-des-no-obsolete-encrypt-20060725

this version of encrypt() conflicts with current linux (which returns void)

we don't use it, so, buh-bye

(cherry picked from commit 68e8e996b6cc82d5097e82a4c9c8af9eb8c3c093)

STABLE14-cellservdb-dl-sanity-check-20060725

protect me from myself

(cherry picked from commit 238ac51e80551bf57334867521817103fddd2cbf)

STABLE14-windows-integrated-logon-20060724

 * apply ACL restrictions to credential cache immediately after
   initialization

 * move file from system temp directory to user temp directory
   prior to starting executable as user to perform credential
   import

(cherry picked from commit 1a0329fef030fedce3fb12d9c641825b0d49f053)

STABLE14-aix5-curpag-20060719

curpag() exists in too many places
anyway, fix em

(cherry picked from commit 72abe41e1df0e530ae73e50b0b952b26d4e26683)

STABLE14-viced-checkrights-loglevel-20060719

increase the log level on the "CheckRights: len=%u, for host=%s:%d\n"
message to 5.

(cherry picked from commit 14ed562c4b576bcfe6b0147b86fa5f7899d22c16)

STABLE14-rxkad-typeindex-20060718

clean up an off by one error for type index

(cherry picked from commit 51d0c3c8d1daf63e2a9d07ff5be0e649eb821b0c)

STABLE14-windows-wix-2_0_4221-20060718

 * Updates to WIX installer to support the 2.0.4221 build

 * Removes unnecessary files from Release Notes

(cherry picked from commit 436e8fe17b2051db806e386db86b3de789afb2f0)

STABLE14-aix-largefile-20060714

FDH_WRITE can return something great \7fer than an int32 size for largefile

(cherry picked from commit 7cb94029ce2092f2117742047cffd3c6b58292cd)

STABLE14-linux-posix-lock-file-api-fix-20060714

fix lock enrollment for older kernels

(cherry picked from commit 3a8a83c061965195c4842d89b888e37cd4b33ee3)

STABLE14-vpurge-needs-no-lock-20060713

don't hold the VOL_LOCK during vpurgevolume
volserver's transaction stuff (NewTrans, actually) precludes any problem

(cherry picked from commit 6878e0b6348a649768274adbce77500fbd45caa4)

STABLE14-rxkad-private-reorder-structures-20060713

based on fix from tkeiser@gmail.com for aix segfaulting issue; keep offsets of
cprivate and sprivate's copies of type and level the same

(cherry picked from commit 845b4ac1cf8f97c5c4d84e36109ef477edef4a31)

STABLE14-tsm-interface-simple-uid-20060713

for now we use afs_uint32 as a uid in the exported interface

(cherry picked from commit 8b898261c1c682f819856966f8158aba36d9abdb)

STABLE14-aix-lam-20060712

tsm for aix5

STABLE14-fix-includes-merge-20060712

Now that we include arpa/inet.h, also include its prerequisites,
and fix the ordering.

STABLE14-windows-wait-rx-hard-dead-time-20060710

Don't timeout based upon the LAN Workstation Session Timeout value.
Instead wait for the rx hard dead timeout.

(cherry picked from commit 8f2aba65819a668fc4a78c0ab929932c23cbc381)

STABLE14-windows-release-notes-20060706

Update the release notes for the 1.4.2 build.  Generate from WinWord
as "filtered html".  This should make the files smaller.

STABLE14-openafs-1-4-2b2-20060705

make openafs 1.4.2-b2

STABLE14-windows-vlserver-makefile-correction-20060705

so make clean works right (remove vldbint.h)

(cherry picked from commit 3aa506fe502f010135d25e0d6996f5cdc38a13bd)

STABLE14-libafsrpc-export-xdr-20060705

"       ld: warning: relocation warning: R_SPARC_COPY: file
       /usr/local/lib/libafsrpc.so: symbol
        rx_enable_stats: relocation bound to a symbol with STV_PROTECTED
       visibility
       Undefined                       first referenced
        symbol                             in file
       xdr_afs_uint32
       /usr/local/lib/libafsauthent.a(ptint.xdr.o)
       ld: fatal: Symbol referencing errors. No output written to slapd
       collect2: ld returned 1 exit status
"

one can guess how that might be less than useful

(cherry picked from commit 3d4b12626136a59bd876648e0d1360f629998df8)

STABLE14-windows-version-20060705

1.4.2-beta2 == 1.4.1052

STABLE14-viced-uninitialized-20060704

tcon in XStatsVersion.

(cherry picked from commit 15963ecb3459f34025c0e8fdbf8f1055e1e4a294)

STABLE14-sol10u2-rename-20060703

FIXES 34774

In s10u2 the vnode path cache code was modified so that we're no longer
guaranteed of having a valid path cache pointer for the lofs mount case.
Attached patch changes gafs_rename to deal with this. Patch also removes
unnecessary MODLOAD32 build rule from sun4x_510 case.

(cherry picked from commit a2d5bb078f3df6dc6fe992fdcaaf8b5a5607c0d6)

STABLE14-windows-notes-20060703

notes

STABLE14-windows-afslogon-pr_end-20060702

* do not attempt to copy kfw krb5 creds if integrated logon disabled

* disable use of automatic registration from afslogon.dll.  calling
  pr_Initialize() results in rx_Init() being called which creates
  threads that prevent safe unloading of the dll.  This is causing
  crashes in third party applications that call CreateProcessAsUser()
  which in turn temporarily loads afslogon.dll.

* add life_to_time to afskfw.c in order remove the link to the rxkad
  and des libraries

(cherry picked from commit 61df8d9fc834b87d758d8581e73fa7c05198ecf4)

STABLE14-windows-nsis-sdk-20060702

Move SDK to Program Files\OpenAFS\SDK

(cherry picked from commit 5accc48e1b0db55c56ba64abe0450d511c0fdf2a)

STABLE14-windows-library-cleanup-20060702

afs_winsockCleanup

(cherry picked from commit 20b4f785b56ff604bab72d3b624aae74a9e5ae5d)

STABLE14-windows-afsauth-krb-20060702

Build the afsuth.krb library

(cherry picked from commit e66467f382a31345f3ea9dd919ea703eccb623d9)

STABLE14-bucoord-64-bit-cleanup-20060701

However, in bucoord/commands.c, bc_VolRestoreCmd(), bc_DiskRestoreCmd(),
bc_VolsetRestoreCmd(), and bc_DumpCmd() were passing '0' as arguments
instead of NULL.  Because there is no prototype for bc_StartDmpRst() this
results in a crash on 64-bit Linux.  (when running 'volrestore' from an
interactive backup session)

STABLE14-butc-prototyping-20060701

have some prototypes
happy july 1
and what a crappy day it is

(cherry picked from commit 1947002d11358b9427adce1d6e03f5357a416ad8)

STABLE14-libcmd-support-version-switch-20060630

well, i guess adm will be sad until it exports AFSVersion, but, life is hard.

(cherry picked from commit 350c140d89198cb7f9f8b241188f9937d7da119e)

STABLE14-solaris-vfsops-20060629

fix symbol binding for modern solaris 10 patches (6/06)
which kills a few we didn't need anyway

(cherry picked from commit 8a1a06b5eb0e415859fd841c3c4a6b578eaea022)

STABLE14-salvager-20060628

make ClearROInUseBit actually do it

(cherry picked from commit 2dcf63ef4e044019bda5889f1de332432124bb71)

STABLE14-rxbind-typo-fix-20060629

FIXES 34634

kill this obvious typo

(cherry picked from commit bc9f256f1c6d2ac26a0b6bfc15de85de895ecf0b)

STABLE14-rxkad-rt34568-20060629

FIXES 34568

The use of strncmp() results in p->v5_str being treated as matching
prefix instead of as the complete string.  This causes "ftpa" to be
considered a match for "ftp".

(cherry picked from commit f0830f89a11a2fea06a26270a90c7ddf8365a533)

STABLE14-windows-uninitialized-vars-20060627

uninitialized pointers when dereferenced do bad things

(cherry picked from commit a8e481ddd92744944d2b06fd5c8ed28383297f6a)

STABLE14-afsutil-includes-20060623

Include <afs/afsutil.h> for prototypes of afs_inet_ntoa and afs_inet_ntoa_r.

(cherry picked from commit 2a1aff4104dd9a71bebf1a63ca85ac86338b3b1b)

STABLE14-windows-cleanup-20060625

cleanup AFS_component_version_number.h

(cherry picked from commit 0b9b02178e7a3d8df39fec9c83f8b49230ce30d9)

STABLE14-windows-nsis-help-files-install-location-20060625

place the help files in the correct location

(cherry picked from commit bfa26d1ddfc7727a9f8c2e79c8d608c586823a12)

STABLE14-windows-report-quota-diskfull-20060624

overquota and diskfull errors are detected while writing dirty buffers
to the file server.  A flag is set on the scp so that the error can
be returned to the caller.  cm_FSync was clearing the flag without
returning the error. This permitted calls to Win32 WriteFile to report
success even when they failed.

(cherry picked from commit 8e0033ba73ef9551cb24052ef90db677e4adad50)

STABLE14-rhel3-krb524-checking-20060624

FIXES 30589

Check for krb524 functions in libkrb524 if not found in libkrb5 so that the
Kerberos v5 support will build on RHEL3.  Thanks to Derek Atkins for the
debugging and testing.

(cherry picked from commit a4cb03504075771e0f2cfebf931f91f29553c73a)

STABLE14-windows-wix-help-files-install-location-20060623

place the help files in the correct location

(cherry picked from commit e264711295275b7b1372e701bf106cf75b342007)

STABLE14-obsd-39-141-20060623

By popular demand, merge OpenBSD 3.9 changes to 1.4.x branch

STABLE14-windows-cm_syncop-buflocked-20060622

If cm_SyncOp is called with the buffer locked, it must be returned
locked even on error.

(cherry picked from commit ae9a07a8e0a5d0e83e138ef9bbd71b3e7e0072ea)

STABLE14-windows-no-freelance-vldb-failover-20060622

When Freelance mode is not in use the cache manager must load the root.afs
volume of the default cell in order to be able to start.   Unfortunately,
if the first vlserver fails the request to obtain the volume information
will timeout and no retries will be performed.  This patch performs up
to 10 retries in the initialization code in order to enable the vlservers
to be marked down so that failover can occur.

(cherry picked from commit 4859232df55c94232a70d602cafe6a8bcc72d9dd)

STABLE14-aix-largefile-cleanup-20060620

some largefile cleanup and other minor aix tweaking

(cherry picked from commit 7826b313868ec8ce975b65e5c430a9503fca4dd5)

STABLE14-dont-override-default-fs-settings-for-aix5-20060620

this was just excessively clever. if aix 5 binaries "want" this, there are configure switches. laziness sucks.

STABLE14-windows-afsrpc-exports-20060620

New functions: rx_InitHost, rx_NewServiceHost

(cherry picked from commit 840ef4a80f3ec8378865232f23171c7c3485626b)

STABLE14-rxbind-for-servers-20060620

do not call parseNetFile on Windows where the function does not exist

(cherry picked from commit 7af6b06cb6611a8925e3fa887ddf0a03542ee4dc)

STABLE14-viced-findclient-do-not-return-wrong-client-20060620

In h_FindClient_r it was possible for the wrong client to be returned
and worse for it to be returned without the reference count being
incremented.

(cherry picked from commit 06fa1f9c40432747724e6e55154bf54e1e3e94c5)

STABLE14-viced-deadlock-20060620

FIXES 34073

* checks for client == oldClient

* drops the H_LOCK prior to obtaining the WriteLock on oldClient
  in order to prevent a deadlock

(cherry picked from commit efd74baa950133de6ed92b7395be6d14ef63a58e)

STABLE14-getaddr-avoid-looping-20060615

so it would be good to count these all off

oops

(cherry picked from commit 2e691074fb6ce70e8352da3e57a746a77f7489f1)

STABLE14-rxbind-for-servers-20060612

FIXES 32308

support -rxbind switch for servers

(cherry picked from commit 592d97f9472abcf77496cef07261fa8766f60b9f)

STABLE14-linux-enable-gcpags-20060612

enable gcpags

(cherry picked from commit 33418253e06534b43ad44eaa71b7c4a84be7cb49)

STABLE14-recvackpkt-count-acks-and-open-cong-window-20060612

"the non AFS_GLOBAL_RXLOCK_KERNEL version of
rxi_ReceiveAckPacket() wasnt counting 'hard' acks and not opening the
congestion window.  using rx_Readv() directs the incoming packets into
the target iov, and sends a hard ack (skipping soft acks).  "

(cherry picked from commit b365b06d34a9450dc53ad22e8ecb3f6d922af5f4)

STABLE14-windows-scache-recycle-fix-20060608

do not special case freelance entries

(cherry picked from commit 40b690803a928da8bf25b1e03745e95f14fd1743)

STABLE14-linux-enroll-locks-20060403

FIXES 25037

enroll our locks in the linux kernels' lock management

(cherry picked from commit 62e28d9a0a6adc6e1f3c2938bfc530b54bea4158)

STABLE14-getaddr-skip-loopbacks-version-for-macos-20060613

so rxbind does something sensible

(cherry picked from commit 3d1fc22c6ba5d07c831da79ffa01e3cf4d95322a)

STABLE14-namei-dec-drop-lock-20060612

FIXES 17638

unsure why yet, or even if, but this is all that makes sense

(cherry picked from commit 1e5f6a17fcf4f7039bae756c22735b77a31ceec5)

STABLE14-windows-cifs-std-info-take-two-20060607

The "is a directory" value was being set one field beyond the end of
the Standard Info reply structure.

(cherry picked from commit 2c20a3d603fcc654759b8c90924f36468f8aae81)

STABLE14-openafs-1-4-2-beta1-20060607

openafs 1.4.2-beta1i

STABLE14-no-client-tcon-20060607

remove client->tcon and avoid locking issues

STABLE14-openlog-consistency-20060607

open logs consistently so assert logging is appended

(cherry picked from commit ed9a2181217f19b7ddbfe28118f6fb2f700b08ad)

STABLE14-synchronize-xdr-20060606

Include the same set of xdr objects in the shared libafsrpc as are included
in the static libafsrpc.

(cherry picked from commit b8b1438716a14b596a68c2e77d54ca66009c01c2)

STABLE14-windows-smb-mountpt-is-dir-20060606

Be consistent about reporting mountpoints and invalid links as directories
to the CIFS client.

(cherry picked from commit 422f2d4523ae9be2ff3f1395db1dd62b0b3a3b4b)

STABLE14-windows-makefile-clean-20060606

cleanup additional files

(cherry picked from commit 81cd89320fe855d22f91bda461dd5cdba147b671)

STABLE14-windows-version-20060604

1.4.0151 == 1.4.2-beta-1

(cherry picked from commit 36da63e03de7230b8464a50a58acf732ce994ff2)