openafs.git
9 years agoConvert ubik_Call(x) calls to ubik_x()
Marc Dionne [Sat, 2 Apr 2011 13:43:26 +0000]
Convert ubik_Call(x) calls to ubik_x()

Convert remaining ubik_Call(function, ..) users to ubik_function(..).

Change-Id: I841d2f32d8fd6a5256e0a922827aad22d242d832
Reviewed-on: http://gerrit.openafs.org/4399
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agofsprobe: Fix usage message
Marc Dionne [Sat, 2 Apr 2011 13:03:55 +0000]
fsprobe: Fix usage message

Fix the usage message to display the correct command name.

Change-Id: I886ebd45554d7e7eb6254fa59fa010fd78e5952a
Reviewed-on: http://gerrit.openafs.org/4398
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agofsprobe: call RXAFS_GetTime directly
Marc Dionne [Sat, 2 Apr 2011 12:40:15 +0000]
fsprobe: call RXAFS_GetTime directly

Call RXAFS_GetTime directly instead of using ubik_Call which is
not prototyped and has logic that we don't need/want here since
we're targeting a specific server.

Change-Id: Ia71d1cb164d3fe71f01c88fc720837c69566baf5
Reviewed-on: http://gerrit.openafs.org/4397
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agolibwp: Tidy header includes
Simon Wilkinson [Fri, 25 Mar 2011 19:36:35 +0000]
libwp: Tidy header includes

Remove headers which are provided by libroken, and reorder header
includes so that they're a bit a more legible.

Change-Id: I4952f8d115d38a3bd1dfe86478cd4a30f533e86f
Reviewed-on: http://gerrit.openafs.org/4386
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>

9 years agoLinux: Fix USE_UCONTEXT detection
Simon Wilkinson [Fri, 1 Apr 2011 15:37:32 +0000]
Linux: Fix USE_UCONTEXT detection

On Linux, afs/param.h determines whether to use the ucontext()
function based on the glibc version. However, the glibc version
macros aren't available until a C library header has been included.

There have been a couple of attempts to fix this. The first included
afs_sysnames.h before the C library check, but this wasn't sufficient
to pull in a header for all builds. The second included stdio.h before
including afs/param.h in process.c.

Whilst the second approach works, it has the drawback that it breaks
our convention that afsconfig.h and afs/param.h must always be the first
includes in a file. More seriously, it means that the behaviour of
afs/param.h is different depending on where it appears in the include
order - which is especially dangerous for 3rd party user.

So, this patch tries to resolve all of this by explicitly including
features.h before doing the glibc version checks. The only danger I
can see here is that Linux platforms without glibc may lack a
features.h, but I suspect we have trouble building on such platforms
in any case.

Change-Id: Ief96d32b10372225335e2d2fbb7205da4bf5fc0a
Reviewed-on: http://gerrit.openafs.org/4389
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoafs: Avoid memory leak on recursive write flock
Andrew Deason [Fri, 1 Apr 2011 21:43:24 +0000]
afs: Avoid memory leak on recursive write flock

When a process requests an exclusive lock on a file on which it
already holds an exclusive lock, we basically form a no-op. However,
HandleFlock was allocating a new SimpleLocks and attaching it to
avc->slocks, without freeing the old SimpleLocks structure.

Since we don't need to do anything if we already hold an exclusive
lock, just break out of the loop right away when we detect that
scenario. Thus we avoid adding a new structure to avc->slocks, and we
avoid a memory leak.

Change-Id: I27c3df1d7807a0b74cba11d6e4a563df8232932a
Reviewed-on: http://gerrit.openafs.org/4395
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agocrypto: return a value from krb5_abortx
Marc Dionne [Fri, 1 Apr 2011 22:53:31 +0000]
crypto: return a value from krb5_abortx

The krb5_abortx stub is declared as returning a krb5_error_code,
so make it return something to avoid a warning.

Change-Id: Ib5b915c1c4e034b5c8393206c80596e5d6d5fc8b
Reviewed-on: http://gerrit.openafs.org/4396
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agomacos: no more startupitems
Derrick Brashear [Fri, 1 Apr 2011 16:17:13 +0000]
macos: no more startupitems

if we're upgrading, a "cruft" startupitems script should just
be nuked.

FIXES 129601

Change-Id: Ic9295b19c893f154fd98a6ad1e5822c8b2ef44df
Reviewed-on: http://gerrit.openafs.org/4387
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

9 years agolibafscp: code cleanup
Chaz Chandler [Sat, 3 Jul 2010 19:02:30 +0000]
libafscp: code cleanup

This patch is intended to bring libafscp into accordance with the
current OpenAFS coding standards while also fixing a few small
issues.  Apologies in advance for the numerous whitespace changes.

Change-Id: I606ed5024395319e12e8759f31494ebd27ff6112
Reviewed-on: http://gerrit.openafs.org/4380
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

9 years agoLinux: Fix fallout from path_lookup commit
Marc Dionne [Wed, 30 Mar 2011 22:32:04 +0000]
Linux: Fix fallout from path_lookup commit

Fix a few issues with the recent commit to deal withg the removal
of path_lookup, spotted on RHEL 5:
- the configure tests needs fs.h to be included before namei.h, to
get the definition of struct inode
- we need to avoid the use of struct path unless its needed; on
older kernels the structure doesn't exist

Change-Id: I6251a96a371a50548dcafc70d94e91b52fc2922a
Reviewed-on: http://gerrit.openafs.org/4382
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agolog: Tidy header includes
Simon Wilkinson [Fri, 25 Mar 2011 19:25:31 +0000]
log: Tidy header includes

Remove headers which are provided by libroken, and reorder header
includes so that they're a bit a more legible.

Change-Id: I41258d5b906350bcdf4854b244799fc8907b2bfc
Reviewed-on: http://gerrit.openafs.org/4385
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agolibadmin: Tidy header includes
Simon Wilkinson [Fri, 25 Mar 2011 10:12:08 +0000]
libadmin: Tidy header includes

Remove headers which are provided by libroken, and reorded header
includes so that they're a bit a more legible.

Add math.h to the list of headers that configure will check for

Change-Id: I530afa77866a0aa3a33f8684ce9cf630aa347812
Reviewed-on: http://gerrit.openafs.org/4384
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agolibacl: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 22:32:39 +0000]
libacl: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Change-Id: I67bb92abc985d1510d8b8a108f9b9e3a62525fd5
Reviewed-on: http://gerrit.openafs.org/4332
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agokopenafs: Tidy up headers
Simon Wilkinson [Thu, 24 Mar 2011 22:17:55 +0000]
kopenafs: Tidy up headers

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Change-Id: Ie4f0f87074b9f9c5a028098ea39e548d9932e54a
Reviewed-on: http://gerrit.openafs.org/4331
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoxdr symbol fallout
Derrick Brashear [Wed, 30 Mar 2011 13:46:43 +0000]
xdr symbol fallout

changes to xdr make xdr_vector and xdr_reference missing from libuafs.
fix it.

Change-Id: I2c48716dc16c86ce91cfdd9deb61cc778f94275a
Reviewed-on: http://gerrit.openafs.org/4379
Tested-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoUnbreak make dest for FBSD
Ben Kaduk [Wed, 30 Mar 2011 02:26:50 +0000]
Unbreak make dest for FBSD

It turns out that we do need an afs.rc.fbsd that is set up for
transarc paths in this directory.  To get it to work properly
will require the user to symlink to it from a dir that gets
checked by rcorder, but them's the breaks.

Change-Id: I1786e4862768127f29a6d309097536829da3e029
Reviewed-on: http://gerrit.openafs.org/4378
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>

9 years agoOpenBSD: Complete implementation of afs_osi_TimedSleep
Antoine Verheijen [Mon, 28 Mar 2011 16:05:40 +0000]
OpenBSD: Complete implementation of afs_osi_TimedSleep

The OpenBSD version of afs_osi_TimedSleep() is missing the required
afs_event structure and afs_getevent routine. This update adds them
(by borrowing a copy of the code from the FreeBSD implementation).

Change-Id: Ie470e83a56caecb9494bb86dc3ff0a18420bb56e
Reviewed-on: http://gerrit.openafs.org/4373
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agovol: Add timeouts to SYNC server select() calls
Andrew Deason [Tue, 29 Mar 2011 17:28:46 +0000]
vol: Add timeouts to SYNC server select() calls

Normally *SYNC server processes wait indefinitely for activity to
occur on one of the SYNC sockets. On some Linux kernels, there exists
a race condition where data can come in on a socket, but the select()
call continues to wait. To ensure that we do not hang forever in such
a scenario, add a timeout to the select() call, which will ensure we
notice the new data within 10 seconds. Raise the timeout on non-Linux
to reduce impact elsewhere.

The Linux kernel bug is tracked in
<https://bugzilla.redhat.com/show_bug.cgi?id=494404>, though that bug
report may not represent all affected kernels.

Change-Id: I3250eb53d59610ccbcffe9e8e283984d5ae0e2b4
Reviewed-on: http://gerrit.openafs.org/4377
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoAlways set LIB_roken when we find libroken
Andrew Deason [Tue, 29 Mar 2011 18:04:48 +0000]
Always set LIB_roken when we find libroken

If we found an external libroken when configuring, we were not setting
LIB_roken if --with-roken=DIR was not specified. Set LIB_roken when we
find a libroken without --with-roken=DIR.

Change-Id: I927647a14f4c5cd8925f0e719b0a9b4164584571
Reviewed-on: http://gerrit.openafs.org/4376
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoNEWS updates for 1.6.0pre4
Russ Allbery [Mon, 28 Mar 2011 19:56:10 +0000]
NEWS updates for 1.6.0pre4

Add additional 1.6.0 NEWS entries based on the release notes for
1.6.0pre4.

Change-Id: Ibd681058f9534c9c270e6e0f127d3beb463c6a95
Reviewed-on: http://gerrit.openafs.org/4374
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

9 years agorx: always use/protect the xdr routines in the kernel
Chas Williams (CONTRACTOR) [Wed, 16 Mar 2011 14:32:48 +0000]
rx: always use/protect the xdr routines in the kernel

This clears up some warnings about duplicate symbols with Solaris 11
since the Solaris kernel already has these routines.  Since we never
use stock kernel version of the xdr routines perhaps we should always
use/protect our version of the symbols.

Change-Id: I4b5b4a691fb838093aff728469d17c28dccbaaea
Reviewed-on: http://gerrit.openafs.org/4252
Reviewed-by: Andrew Deason <adeason@sinenomine.net>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agokauth: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 21:58:29 +0000]
kauth: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

This adds arpa/inet.h to the list of headers that configure checks for

Change-Id: I1792ede4be761238bb05567c1d763ed63f50051e
Reviewed-on: http://gerrit.openafs.org/4330
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agogtx: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 21:08:35 +0000]
gtx: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Change-Id: I444a8a325cca8da76f410185949103ccd873c743
Reviewed-on: http://gerrit.openafs.org/4329
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agofsprobe: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 21:03:00 +0000]
fsprobe: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Change-Id: I0eee5dd0656e32617855ccf4a30d4c4be47b7da2
Reviewed-on: http://gerrit.openafs.org/4328
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agofinale: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 20:58:46 +0000]
finale: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Also, tidy up translate_et_nt.c by removing the AIX specific code which
will never be built, as this file is only compiled on Windows!

Change-Id: Iec7a811f02eaf1a2ad573bbadc30d782426231d0
Reviewed-on: http://gerrit.openafs.org/4327
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agodir: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 20:56:11 +0000]
dir: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Change-Id: I355cf33fdf72fed640c17ff5c2dfca1dd6f93d6f
Reviewed-on: http://gerrit.openafs.org/4326
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agocom_err: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 20:50:21 +0000]
com_err: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Change-Id: I167b0fa7f48d7dd71efb43a4f137655b38137314
Reviewed-on: http://gerrit.openafs.org/4325
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agovol: remove flock emulation now provided by roken
Jeffrey Altman [Mon, 28 Mar 2011 12:29:25 +0000]
vol: remove flock emulation now provided by roken

flock emulation is provided by rk_flock from roken.   No need
to provide a local emulation function in namei_ops.c.

This change removes a macro redefinition warning.

Change-Id: Id8b7cf030d8653ceb979db60c20ed28c03bd9737
Reviewed-on: http://gerrit.openafs.org/4365
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agoAIX: Add flock to roken
Simon Wilkinson [Mon, 28 Mar 2011 13:52:33 +0000]
AIX: Add flock to roken

Add the rk_flock() emulation function to roken, primarily for use
on the AIX build, but it will also be used for WIN32

Change-Id: Ia8f11e19bbe9fdf721a7b72660d860624139d9b1
Reviewed-on: http://gerrit.openafs.org/4368
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agocmd: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 20:46:54 +0000]
cmd: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Change-Id: I424a0df98ff9334d9e46a9af2ba9d7f1861f501e
Reviewed-on: http://gerrit.openafs.org/4324
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agobu*: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 20:42:37 +0000]
bu*: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Change-Id: If20a2ef67d4cbc70156c1707264a5b7360cfa11a
Reviewed-on: http://gerrit.openafs.org/4323
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agobozo: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 20:07:37 +0000]
bozo: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Change-Id: Ie4f1b955c111f30b98aa8793b3e3d15dd0bb9275
Reviewed-on: http://gerrit.openafs.org/4322
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoImport of code from heimdal
Heimdal Developers [Mon, 28 Mar 2011 13:44:20 +0000]
Import of code from heimdal

This commit updates the code imported from heimdal to
987658325eb7bd6b10882cc4cdd224499e7f7a03 (switch-from-svn-to-git-2052-g9876583)

New files are:
roken/flock.c

Change-Id: I7b9470cfdf99e41ea460772025767c47f6ad26f0
Reviewed-on: http://gerrit.openafs.org/4367
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoImport flock.c from roken
Simon Wilkinson [Mon, 28 Mar 2011 13:43:28 +0000]
Import flock.c from roken

Add flock.c to our list of roken imports, so that we can use it on
AIX

Change-Id: I1f7b953c8853f4b021a486bb93ddc392c2c08e07
Reviewed-on: http://gerrit.openafs.org/4366
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

9 years agoLinux: 2.6.39: replace path_lookup with kern_path
Marc Dionne [Sun, 27 Mar 2011 14:59:55 +0000]
Linux: 2.6.39: replace path_lookup with kern_path

path_lookup is no longer available, use kern_path instead.

Change-Id: I42ae43114fe257fc65452f1b0a35d43595b0044b
Reviewed-on: http://gerrit.openafs.org/4360
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoihandle release locking simplification
Derrick Brashear [Tue, 22 Mar 2011 16:47:16 +0000]
ihandle release locking simplification

several callers drop locks so ih_release can be called unlocked,
then relock. simply allow a locked call (via _ih_release_r).

side effect: we had races before on refcnt check versus cleanup style.
this addresses it, but only removes lock contention.

Change-Id: Id2d132baa170894ba3ab0e1e8d0bcf9cf6c0c712
Reviewed-on: http://gerrit.openafs.org/4271
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoauth: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 19:56:35 +0000]
auth: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Change-Id: I2afd32d86b40b84c14f8a6021a583b6a4888780c
Reviewed-on: http://gerrit.openafs.org/4321
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agoautoconf: Add tests for resolver headers
Simon Wilkinson [Mon, 28 Mar 2011 09:24:14 +0000]
autoconf: Add tests for resolver headers

Add tests for arpa/nameserv.h and resolv.h, so that roken will look
after including these files for us.

Change-Id: Ifa6b7d877d67c967811017a25c98fab6bd571d4c
Reviewed-on: http://gerrit.openafs.org/4364
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agofssync: Rename state
Simon Wilkinson [Mon, 28 Mar 2011 09:23:22 +0000]
fssync: Rename state

The structure 'state' in fssync-debug conflicts with a structure
of the same name in AIX's resolv.h header. Renaming the structure here
to fssync_state to avoid the conflict.

Change-Id: I15df31a48369f76ef927a5153d3415e821bc0c51
Reviewed-on: http://gerrit.openafs.org/4363
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agoaudit: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 19:27:21 +0000]
audit: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Change-Id: I5543c5c9c22828ee19460c2ab24065f508a6353f
Reviewed-on: http://gerrit.openafs.org/4320
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agovol: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 19:23:34 +0000]
vol: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Change-Id: Ifd88310dc26ebfd995b4a25b2984ddb01606264f
Reviewed-on: http://gerrit.openafs.org/4319
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agovolser: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 19:11:15 +0000]
volser: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Also, add sys/uio.h to the list of libraries that we check for.

Change-Id: I03d9c143db42a1ec415ab5624e2cbede0d34e310
Reviewed-on: http://gerrit.openafs.org/4318
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agovolser: More renaming of putshort and putint32
Simon Wilkinson [Fri, 25 Mar 2011 00:37:52 +0000]
volser: More renaming of putshort and putint32

Commit 2d6bc153689f14f8690878b15d9fb711a56b0546 renamed the putshort
and putint32 macros in dumpstuff.c to avoid conflicts with macros
that are defined in resolv.h. However, an almost identical copy of
that code also exists in vol-dump.c

Rename putshort and putint32 there too.

Change-Id: I9b98f99fb5c2b24b57e2bc480c891aa86b6df344
Reviewed-on: http://gerrit.openafs.org/4317
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agoafs_snprintf is dead, long live rk_snprintf
Simon Wilkinson [Thu, 24 Mar 2011 15:36:49 +0000]
afs_snprintf is dead, long live rk_snprintf

We were shipping our own implementation of snprintf and friends, as
afs_snprintf. Now that we're using roken everywhere, we can make use
of roken's rk_snprintf, and no longer need to ship our own.

As Window's snprintf isn't C99 compliant, roken always uses its own
on this platform. The effect of this is that we can no longer use
AFS_UINT64_FMT and AFS_INT64_FMT for snprintf calls (and the Log
functions that call them). Instead, we need to always use the
Unix format specifiers when calling these functions.

Make thse changes across the whole tree.

Change-Id: I3fffef97566f239ad639f15c4decd136d5bbd765
Reviewed-on: http://gerrit.openafs.org/4316
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

9 years agoWindows: Don't use %I for IP addresses
Simon Wilkinson [Sun, 27 Mar 2011 20:06:27 +0000]
Windows: Don't use %I for IP addresses

%I isn't supported by the roken printf functions. In preparation
for switching over to them, replace our use of asprintf to convert
IP addresses to strings with the afs_inet_ntoa_r function.

Change-Id: Id38dc56405071d62fb2b12e4f69905aa7a2026b0
Reviewed-on: http://gerrit.openafs.org/4362
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

9 years agoLinux: 2.6.39: deal with BKL removal
Marc Dionne [Sun, 27 Mar 2011 15:20:17 +0000]
Linux: 2.6.39: deal with BKL removal

For 2.6 kernels the OpenAFS code is already BKL free, but the
corresponding header file smp_lock.h is still included in various
places.  This header no longer exists, so remove all occurrences
outside of LINUX24.

Change-Id: Ib659fa787eab4b39efce5b0a62f5723950e79703
Reviewed-on: http://gerrit.openafs.org/4361
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agovlserver: Rationalise usage message
Simon Wilkinson [Thu, 24 Mar 2011 10:57:47 +0000]
vlserver: Rationalise usage message

Rather than having 2 copies of the usage message, just have one
with #ifdef switches to disable options that are not available in
that build.

Change-Id: I5b2135fba3d42fdbb879d5e8cb61933f4929d598
Reviewed-on: http://gerrit.openafs.org/4305
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoptserver: Rationalise usage message
Simon Wilkinson [Thu, 24 Mar 2011 10:28:59 +0000]
ptserver: Rationalise usage message

Rather than having 4 copies of the usage message, just have one
with #ifdef switches to disable options that are not available in
that build.

Change-Id: If888a64307b78189fcf617bbe0dc4ca9e55e8d2b
Reviewed-on: http://gerrit.openafs.org/4304
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoHP/UX: Make more use of roken
Simon Wilkinson [Fri, 25 Mar 2011 17:14:26 +0000]
HP/UX: Make more use of roken

Make more use of roken on HP/UX by letting it provide definitions
for random and srandom, and using its implementation of
getdtablesize

Change-Id: I1212c77ea9cc6ef436cfc5c16e893aefbc33d31b
Reviewed-on: http://gerrit.openafs.org/4341
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agocrypto: Fixes for recent Heimdal changes
Simon Wilkinson [Fri, 25 Mar 2011 17:34:51 +0000]
crypto: Fixes for recent Heimdal changes

The last Heimdal import changed the name of a number of structures
so that they are private to the krb5 code. Mirror this by changing our
prototypes to match the new names.

This solves the key_type problem that we previously fixed by #defining
key_type to hc_key_type, so remove that define.

Change-Id: I9b3f1712b0bf6641a780cc4002e73f8a511ff080
Reviewed-on: http://gerrit.openafs.org/4340
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoImport of code from heimdal
Heimdal Developers [Fri, 25 Mar 2011 17:18:37 +0000]
Import of code from heimdal

This commit updates the code imported from heimdal to
987658325eb7bd6b10882cc4cdd224499e7f7a03 (switch-from-svn-to-git-2052-g9876583)

New files are:
roken/getdtablesize.c

Upstream changes are:

Love Hornquist Astrand (2):
      prefix symbols that are _krb5_ structures
      prefix internal structures with _krb5_

Change-Id: I36aae13d877b6b40b8d2db47217884dc24f3f9c8
Reviewed-on: http://gerrit.openafs.org/4339
Tested-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoAdd getdtablesize to the roken import
Simon Wilkinson [Fri, 25 Mar 2011 17:17:47 +0000]
Add getdtablesize to the roken import

Change-Id: I089e3976e90429e39102eeb82a74ece97e87bd25
Reviewed-on: http://gerrit.openafs.org/4338
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoaklog: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 15:58:19 +0000]
aklog: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Also add pwd.h to the list of headers that we check for in configure,
and that roken will include if they are available

Change-Id: I61ab95eeca11127a33bb668dddfc24ec6ce7f8f1
Reviewed-on: http://gerrit.openafs.org/4303
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoNO_DES_H_INCLUDE is no longer necessary
Simon Wilkinson [Thu, 24 Mar 2011 15:50:47 +0000]
NO_DES_H_INCLUDE is no longer necessary

Now that we're using hcrypto for DES, rather than rolling our own,
we don't need to specific NO_DES_H_INCLUDE, so get rid of it.

Change-Id: Id370ddacb0a57264cc8f5cdda3c59fbb627708af
Reviewed-on: http://gerrit.openafs.org/4302
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoptserver: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 10:01:00 +0000]
ptserver: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Also reorder our headers so that they go

 #include <afsconfig.h>
 #include <afs/param.h>
 #include <afs/stds.h> (if required)

 #include <roken.h>

 [ ... out of directory headers ... ]

 [ ... local headers ... ]

Change-Id: I9214f6bf65085947cfc588b47484e1b022ffc5b1
Reviewed-on: http://gerrit.openafs.org/4299
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agonew files for FreeBSD packaging at 1.6.0pre3
Ben Kaduk [Fri, 17 Dec 2010 06:16:30 +0000]
new files for FreeBSD packaging at 1.6.0pre3

Packaging for use with FreeBSD's Ports Collection.
The directory layout is flattened, and a couple of files that do not
make sense here are not included.
The afsd rc script lives in packaging/ as it is not runnable in
its committed form, it requires variable substitution that occurs
at install-time from the Ports Collection.

Change-Id: I72d6c2ee7106611b0ec9bfe3e47a471ed66b7255
Reviewed-on: http://gerrit.openafs.org/4351
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

9 years agoRemove outdated rc file for afsd on FBSD
Ben Kaduk [Mon, 22 Nov 2010 04:32:37 +0000]
Remove outdated rc file for afsd on FBSD

The semantics of rc scripts have changed drastically since this
was written.  It will need to be rewritten from scratch; the
new version will live in src/packaging/FreeBSD as afsd.in, as it
is included with the port packaging in that form.

Change-Id: I397f8d5a9dc0430f1a9e8f2f0f378585cee4bd3b
Reviewed-on: http://gerrit.openafs.org/4350
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

9 years agoRemove stale FreeBSD packaging
Ben Kaduk [Sat, 6 Nov 2010 04:30:12 +0000]
Remove stale FreeBSD packaging

We no longer use the Transaarc paths, and the rc environment
and packaging methods are quite different now.

Change-Id: I6915f41455dcf0518472b0a81fe7e5c559d4a994
Reviewed-on: http://gerrit.openafs.org/4349
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

9 years agoWindows: export roken symbols
Jeffrey Altman [Sat, 26 Mar 2011 20:50:52 +0000]
Windows: export roken symbols

rk_vsnprintf
rk_vasnprintf

Change-Id: I39344d10a599e7bdc761ae03853b679ab6edcb4f
Reviewed-on: http://gerrit.openafs.org/4348
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

9 years agoWindows: change log for 1.5.9904 (1.6.0pre4)
Jeffrey Altman [Sat, 26 Mar 2011 17:05:26 +0000]
Windows: change log for 1.5.9904 (1.6.0pre4)

Change-Id: I0c302ef20bf245f927f3167d3c662334e71d81df
Reviewed-on: http://gerrit.openafs.org/4347
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

9 years agoviced: REALLYCLOSE origfdP after CoW
Andrew Deason [Fri, 25 Mar 2011 20:54:01 +0000]
viced: REALLYCLOSE origfdP after CoW

In StoreData_RXStyle, we save a pointer to the original FdHandle_t if
we need to CopyOnWrite the target vnode, for the purposes of possibly
copying additional data later on.

After the CopyOnWrite call, this points to an inode that is not in the
current volume, and is thus less likely to be accessed. In CopyOnWrite
itself, the original file handle is FDH_REALLYCLOSE'd because of this,
so the handle does not remain in the cache. Do the same in
StoreData_RXStyle, so origfdP is always FDH_REALLYCLOSE'd after we
have called CopyOnWrite.

Change-Id: Ibe33d26120eb354318147f3a31fe32d38c6a70c6
Reviewed-on: http://gerrit.openafs.org/4344
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

9 years agoRevert "ihandle: Ensure FDH_REALLYCLOSE really closes"
Andrew Deason [Fri, 25 Mar 2011 20:44:57 +0000]
Revert "ihandle: Ensure FDH_REALLYCLOSE really closes"

This reverts commit a50f4c8d802c7b1c258a9cbe388dfab50425596b.

The only difference between FDH_CLOSE and FDH_REALLYCLOSE has to do
with ihandle cache behavior. FDH_REALLYCLOSE just indicates that the
reference we are closing is not likely to be used by other code, and
should be kicked out of the cache immediately (thus closing the file
descriptor). So, if there is another reference to that FdHandle_t, and
it issues an FDH_CLOSE, it thinks that the reference is likely to be
used again soon, and so the handle should not be kicked out of the
cache. Thus, the behavior before a50f4c8d was already correct.

Change-Id: I5a5c0c18b459f50e16c1d901ecc5f4eb18e8ef62
Reviewed-on: http://gerrit.openafs.org/4343
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

9 years agovol: avoid MAXINT redefinition warning in vnode.c
Jeffrey Altman [Tue, 8 Mar 2011 16:04:07 +0000]
vol: avoid MAXINT redefinition warning in vnode.c

Change-Id: Ifeed132ac477a83f21af5b52cba7b658fb083221
Reviewed-on: http://gerrit.openafs.org/4179
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoSOLARIS: Perform daemon syscalls as kernel threads
Andrew Deason [Tue, 8 Mar 2011 22:59:32 +0000]
SOLARIS: Perform daemon syscalls as kernel threads

Add AFS_SUN5_ENV to the list of platforms where AFS_DAEMONOP_ENV is
defined. Implement the necessary functionality so we spawn kernel
threads when a daemon syscall is called. Remove the rxk_Listener
wrapper, since it will be called in a separate thread via the
afs_DaemonOp interface.

Change-Id: I3c2570696a83f1837d08522fdd9dfc30dfefda4b
Reviewed-on: http://gerrit.openafs.org/4189
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agolibafs: Consolidate afs_DaemonOp code
Andrew Deason [Tue, 8 Mar 2011 21:37:17 +0000]
libafs: Consolidate afs_DaemonOp code

Create the AFS_DAEMONOP_ENV define to simplify the logic of when we
perform afs_DaemonOp-y code paths. Also create the daemonOp_common
function, to perform common pre-fork operations that are common
between platforms.

Change-Id: Id93e1ffd7b8f4c0029cf4632239a519a619e9f9f
Reviewed-on: http://gerrit.openafs.org/4188
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoSOLARIS: Correct ioctl syscall error handling
Andrew Deason [Fri, 25 Mar 2011 21:37:30 +0000]
SOLARIS: Correct ioctl syscall error handling

Do not use the return code from ioctl for errno. Ioctl itself will set
the errno and return value correctly, so don't mess with them.

Change-Id: I26c3afb5c77a3fce57e0d53d8671541805b32e08
Reviewed-on: http://gerrit.openafs.org/4345
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agolibafs: Indent afs_call.c ifdef maze
Andrew Deason [Tue, 8 Mar 2011 20:59:44 +0000]
libafs: Indent afs_call.c ifdef maze

Change-Id: I46198de875c16e221d4b75bb45afc9e2fd09dde1
Reviewed-on: http://gerrit.openafs.org/4187
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoihandle: One more indentation fix
Andrew Deason [Fri, 25 Mar 2011 19:11:16 +0000]
ihandle: One more indentation fix

Change-Id: Icfed36ba0087d82706aea88e083576480244a9ad
Reviewed-on: http://gerrit.openafs.org/4342
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

9 years agosalvager: Stop asking for DAFSness on SYNC_FAILED
Andrew Deason [Thu, 24 Mar 2011 22:57:32 +0000]
salvager: Stop asking for DAFSness on SYNC_FAILED

Break out of the loop so we stop asking the same thing repeatedly.

Change-Id: Ifcb2cd62589867927ab9f8be03bab7b981e5f761
Reviewed-on: http://gerrit.openafs.org/4315
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoDAFS: Allow the volumeSalvager to checkout volumes
Andrew Deason [Wed, 23 Mar 2011 22:38:35 +0000]
DAFS: Allow the volumeSalvager to checkout volumes

Change-Id: Ia255c866bbd4f6cb883383650373223f09875374
Reviewed-on: http://gerrit.openafs.org/4298
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoWindows: trailing dot not part of volume name
Jeffrey Altman [Fri, 25 Mar 2011 12:58:20 +0000]
Windows: trailing dot not part of volume name

When dynamically adding new cell mount points to the Freelance
directory do not include a trailing dot when specifying the root.cell
volume name.

Change-Id: Idfae9e3f44380cbce49d77fff22bee15dc4856ce
Reviewed-on: http://gerrit.openafs.org/4333
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

9 years agoadd AFS_DEMAND_ATTACH_UTIL
Derrick Brashear [Thu, 24 Mar 2011 20:11:17 +0000]
add AFS_DEMAND_ATTACH_UTIL

use this for non-pthreads "support code" e.g. salvager.

Change-Id: I17825fa90de33a7a4d17f364713fd4602dae497b
Reviewed-on: http://gerrit.openafs.org/4312
Reviewed-by: Andrew Deason <adeason@sinenomine.net>
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoutil: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 15:26:38 +0000]
util: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times.

Change-Id: I5e5e67b695f07bf121b0543c7f936278009eff15
Reviewed-on: http://gerrit.openafs.org/4301
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agovlserver: Tidy header includes
Simon Wilkinson [Thu, 24 Mar 2011 10:54:44 +0000]
vlserver: Tidy header includes

Since we adopted libroken, we get a lot of header includes automatically
from roken.h. Use these, instead of including lots of things multiple
times

Change-Id: I8b304bcd0e376601fbe622f7453d0295b188b697
Reviewed-on: http://gerrit.openafs.org/4300
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoDAFS: Correct FSYNC_VOL_QUERY_VOP checks
Andrew Deason [Thu, 24 Mar 2011 15:22:52 +0000]
DAFS: Correct FSYNC_VOL_QUERY_VOP checks

Check that the given partition matches the vp partition, and ensure
the vp is not in an exclusive state when we check the state.
Otherwise, we may return pending vol ops for a volume on a different
partition, or we may incorrectly return that there is no pending vol
op when in fact the volume does not exist at all.

Change-Id: I3e28c0b7b372360d181a3310eb1fb7fce223ae59
Reviewed-on: http://gerrit.openafs.org/4308
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoCorrectly document the AFS client setuid defaults
Russ Allbery [Thu, 24 Mar 2011 19:56:56 +0000]
Correctly document the AFS client setuid defaults

AFS no longer honors setuid status by default.  Update the admin
guide documentation appropriately and add a warning recommending
against enabling setuid status given the limitations of the current
AFS protocol.

Reformat this section of the admin guide to make it easier to
maintain.

Change-Id: I6ea5859037d7d194df801f1a76583257cfc8bbe8
Reviewed-on: http://gerrit.openafs.org/4311
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

9 years agoWindows: avoid recursive cm_CheckOfflineVolume
Jeffrey Altman [Wed, 16 Mar 2011 06:17:38 +0000]
Windows: avoid recursive cm_CheckOfflineVolume

Add a new cm_req_t flag CM_REQ_OFFLINE_VOL_CHK which is used
to prevent cm_Analyze() from performing recursive
cm_CheckOfflineVolume operations that will exhaust the stack.

Change-Id: Ia4ee14307bf812cc2208482a19c1a914aca3e447
Reviewed-on: http://gerrit.openafs.org/4240
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

9 years agosalvager: Give back volumes when exiting early
Andrew Deason [Wed, 23 Mar 2011 22:25:03 +0000]
salvager: Give back volumes when exiting early

Sometimes the salvager exits a bit earlier than normal. For instance,
when no applicable inodes are found for a volume group, or if the
-inodes command line option was given. In these cases, we have already
checked out singleVolumeNumber from the fileserver (if we're salvaging
a single VG), so we need to give it back. So, give it back in those
instances.

Change-Id: I3ab732d3b640b76d3bdec7ac9d01e57dc5a54ade
Reviewed-on: http://gerrit.openafs.org/4297
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoDAFS: Do not record vol ops for DELETED vols
Andrew Deason [Wed, 23 Mar 2011 21:46:47 +0000]
DAFS: Do not record vol ops for DELETED vols

When a volume is VOL_STATE_DELETED, it effectively does not exist, so
there is little point in recording a vp->pending_vol_op structure for
it. Just let callers checkout the volume as they would a nonexistent
volume: without recording anything about the operation.

This just reduces some edge cases and confusing debugging info, so we
don't have to worry about cleaning up pending_vol_op structures for
nonexistent volumes.

Change-Id: I56b92c3f2548ea48f9c61100be07edf9f7277fee
Reviewed-on: http://gerrit.openafs.org/4296
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agosalvager: Do not AskOnline nonexistent volumes
Andrew Deason [Wed, 23 Mar 2011 21:12:20 +0000]
salvager: Do not AskOnline nonexistent volumes

If singleVolumeNumber is not in our volume summary list, then the
singleVolumeNumber volume does not exist. So, don't try to bring it
back online. Still do try to make sure we don't have the volume
checked out, though, so issue an AskDelete, so ensure that it's not
checked out and that the fileserver does not think it exists.

Change AskDelete so we don't care if we tried to delete a volume that
the fileserver thinks already doesn't exist. Change the FSYNC_VOL_DONE
handler so it does not complain about already-deleted volumes.

Change-Id: I33dd876c0db91a3c8a5210fca329d63bf4b3d212
Reviewed-on: http://gerrit.openafs.org/4295
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agonamei: Log ListViceInodes write failures
Andrew Deason [Wed, 23 Mar 2011 20:19:03 +0000]
namei: Log ListViceInodes write failures

If we fail to write to the inodeFile given by ListViceInodes, we
should log the failure. Otherwise, no indication is given as to what
failed, when a salvage gathers the inode list.

Change-Id: Ia54105e1043969781711b9168c6787d07d1d319f
Reviewed-on: http://gerrit.openafs.org/4294
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agosalvager: Fix conversion from stdio calls
Andrew Deason [Wed, 23 Mar 2011 19:30:18 +0000]
salvager: Fix conversion from stdio calls

Commit 5247fa38a4faebfdffba178ca01e5b419f034d5d converted some I/O
calls in the salvager from stdio to OS_OPEN/OS_WRITE/etc. Fix some of
the conversions, including:

 - We need to pass O_CREAT to OS_OPEN calls, since we are creating
   these files

 - OS_WRITE returns the number of bytes written on success, not the
   number of "elements" like in stdio

Change-Id: I789b7549c1eabeb821591bf9f42d810252fb11e1
Reviewed-on: http://gerrit.openafs.org/4293
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoihandle: Ensure FDH_REALLYCLOSE really closes
Andrew Deason [Tue, 22 Mar 2011 21:36:47 +0000]
ihandle: Ensure FDH_REALLYCLOSE really closes

If FDH_REALLYCLOSE is given an FdHandle_t that has more than one user,
currently it does effectively nothing. Ensure that the file descriptor
actually gets closed on a subsequent FDH_CLOSE, but setting the new
fd_needs_rclose field.

Change-Id: I04794662ca64e6be718da82e10994a4a7bc0b39a
Reviewed-on: http://gerrit.openafs.org/4274
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agomacos next support
Derrick Brashear [Wed, 23 Mar 2011 18:16:10 +0000]
macos next support

first brush at supporting next macos. totally not lion about it.

Change-Id: I4d2e05f68266ea82de710717c5340f16425f897e
Reviewed-on: http://gerrit.openafs.org/4290
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoosconf: reduplicate systype stuff
Derrick Brashear [Wed, 23 Mar 2011 18:11:24 +0000]
osconf: reduplicate systype stuff

in some cases there is more to do, and in some cases we probably
already didn't build anymore, but, let's simplify.

Change-Id: I779c3ba8f3f02e6202700d96ee92cd473de66af2
Reviewed-on: http://gerrit.openafs.org/4288
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

9 years agoDAFS: Do not VDeregisterVolOp_r while exclusive
Andrew Deason [Wed, 23 Mar 2011 16:07:03 +0000]
DAFS: Do not VDeregisterVolOp_r while exclusive

We should wait for a volume to transition out of an exclusive state
before calling VDeregisterVolOp_r on a volume, since some code may be
examining the vol op outside of VOL_LOCK in an exclusive state. We
should be doing this anyway before performing volume state checks,
since we may be trying to e.g. attach the volume at the same time.

Change-Id: I3a5b78b7c0033887a973ce941ced592df8c4bf3d
Reviewed-on: http://gerrit.openafs.org/4287
Tested-by: Andrew Deason <adeason@sinenomine.net>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agorpm: Build srpms without %dist in their name
Simon Wilkinson [Wed, 23 Mar 2011 15:35:35 +0000]
rpm: Build srpms without %dist in their name

On RHEL6, %dist is set for all RPMS and SRPMS built on the system. This
leads to makesrpm.pl building SRPMS with .el6 in their release. As we
endeavour to use the same SRPM on all of our Fedora platforms, this is
less than ideal, so change makesrpm.pl to squash the setting of %dist
before building the RPM.

Change-Id: I5b5b9db7654ce21a269233e8950d4f4d1df81ad7
Reviewed-on: http://gerrit.openafs.org/4285
Reviewed-by: Alexander Ivan Redinger <aredinger@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

9 years agoFix ihandle.c indents
Andrew Deason [Tue, 22 Mar 2011 21:18:17 +0000]
Fix ihandle.c indents

ihandle.c had some blocks that were not indented. Indent them.

Change-Id: I05442c6fabc19fab314f515fa950159de05b8c48
Reviewed-on: http://gerrit.openafs.org/4273
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>

9 years agovclosevnodefiles-ihandle-leak-20090216
Jeffrey Altman [Tue, 17 Feb 2009 04:37:27 +0000]
vclosevnodefiles-ihandle-leak-20090216

LICENSE IPL10
FIXES 124359

don't leak ihandles on close. this isn't a complete fix for the issue
(cherry picked from commit b9816e12f7ed8213c9c4eaea09e992e69ce4ee05)

This reverts the part of 12e85227c5dbfdb1258718ee3360bffacc4f96ac that
reverted b9816e12f7ed8213c9c4eaea09e992e69ce4ee05.

Change-Id: If4ac661a3a0f8abf7f6a79fa9d72d715861bcf51
Reviewed-on: http://gerrit.openafs.org/4272
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoDAFS: fix forceDAFS support in salvager
Derrick Brashear [Wed, 8 Dec 2010 02:26:59 +0000]
DAFS: fix forceDAFS support in salvager

we need to get the shared partition lock for the per-volume
forceDAFS case

Change-Id: I7be91420bb0df5963e789c932e5c469fe02ec108
Reviewed-on: http://gerrit.openafs.org/3487
Reviewed-by: Andrew Deason <adeason@sinenomine.net>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoDAFS: allow salvager to detect whether FSYNC server is DAFS
Derrick Brashear [Wed, 8 Dec 2010 02:23:53 +0000]
DAFS: allow salvager to detect whether FSYNC server is DAFS

in order to enable the right commands and print the right errors,
query for DAFS-only FSYNC op.
doesn't use bozo InstanceInfo as it's possible you can run fileserver
outside of bos.

Change-Id: Ib94f7b3a2960cd493f110af4d5dee7e97a6493c9
Reviewed-on: http://gerrit.openafs.org/3486
Reviewed-by: Andrew Deason <adeason@sinenomine.net>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoUpdate NEWS for 1.6.0pre3 release
Russ Allbery [Tue, 22 Mar 2011 09:04:46 +0000]
Update NEWS for 1.6.0pre3 release

Add additional NEWS entries from the 1.6.0pre3 release notes.

Change-Id: I46b306a7fc85f95be87c2b9eeaefae6b9816be8f
Reviewed-on: http://gerrit.openafs.org/4270
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoviced: Check vnode length on dir ops
Andrew Deason [Tue, 15 Mar 2011 19:24:01 +0000]
viced: Check vnode length on dir ops

The commit aadf69eabb1962496fa93745ab560a5b48cacd61 added checks on
vnode length whenever we read or write from a vnode. Add the same
check on directory vnodes when we modify the directory (whenever
entries are added or deleted).

Change-Id: I8aa438941f840019bc541d5a978610c4f78330c8
Reviewed-on: http://gerrit.openafs.org/4233
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agovol: Correct VolumeNumber for large volume IDs
Andrew Deason [Fri, 11 Mar 2011 00:37:44 +0000]
vol: Correct VolumeNumber for large volume IDs

VolumeNumber was using atoi to convert a volume header name to a
volume ID. This can return just -1 for volume IDs larger than 2^31-1,
though, so use strtoul instead.

Change-Id: Ibae2d29f196646154fcc4f6ff174246e37b3d735
Reviewed-on: http://gerrit.openafs.org/4199
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agovol: Handle large volume IDs in VLockFile
Andrew Deason [Thu, 10 Mar 2011 23:59:39 +0000]
vol: Handle large volume IDs in VLockFile

VLockVolumeByIdNB currently cannot handle volume IDs larger than
2^31-1. Fix this by using struct flock64, F_SETLKW64, and F_SETLK64 in
the VLockFile functions where possible.

Thanks to Simon Wilkinson for pointing out F_SETLK64.

Change-Id: I422c685aec035716e2f42d13bd97541425ead6a2
Reviewed-on: http://gerrit.openafs.org/4198
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agovol: Make VLockFile arguments consistent
Andrew Deason [Thu, 10 Mar 2011 22:41:02 +0000]
vol: Make VLockFile arguments consistent

The Windows implementation of the VLockFile functions had the handle
incorrectly declared in a few places (they accepted a struct VLockFile
instead of an FD_t). Correct them, and make all VLockFile
implementation functions declare 'FD_t's instead of 'HANDLE's or
'int's.

Change-Id: I2bc414712599cb8f2a2f0a5034b583a15b2c3ad8
Reviewed-on: http://gerrit.openafs.org/4197
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agolibafs: Do not osi_FlushPages for dirs
Andrew Deason [Thu, 17 Mar 2011 21:32:00 +0000]
libafs: Do not osi_FlushPages for dirs

Directory contents are never mapped or stored in pages, so dealing
with page invalidation on directories is just overhead. So make
osi_FlushPages a no-op when we're given a directory, which can avoid a
lot of locks and other processing (particularly when we are called in
afs_getattr in BOZONLOCK_ENV).

Change-Id: I56fbd08c74d01a5fcb45a57e08bb31afcec9331d
Reviewed-on: http://gerrit.openafs.org/4259
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoDAFS: DFlushVolume outside of vol glock
Andrew Deason [Wed, 16 Mar 2011 16:48:08 +0000]
DAFS: DFlushVolume outside of vol glock

DFlushVolume may traverse a long list of directory objects, and can
even hit the disk, so we should drop the glock for it. This should be
safe in DAFS, since we already transition the volume to an exclusive
state before doing this, and DFlushVolume only deals with structures
internal to the directory package and maintains its own locking.

Change-Id: I6493042c8f36614b55d4dc40d97639e56b34fabd
Reviewed-on: http://gerrit.openafs.org/4242
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agoutil: Fix exec_alt
Simon Wilkinson [Sun, 20 Mar 2011 21:13:09 +0000]
util: Fix exec_alt

exec_alt was failing its tests on Mac OS X (but passing them on Linux).

It turns out that this is because it was failing to NULL terminate the
string that it creates in construct_alt(), which copies in the
characters from argv0, prefix and suffix, but never copies in a trailing
NULL.

Amend the code so that the trailing NULL from suffix is used to
terminate the string.

Change-Id: I286604fb09193367c4e7b2d80051832080362f79
Reviewed-on: http://gerrit.openafs.org/4267
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

9 years agolwp: override dbg and optmz settings
Simon Wilkinson [Sun, 20 Mar 2011 23:23:29 +0000]
lwp: override dbg and optmz settings

we were wiping out fun things like archflags. shouldn't be needed.
let's just do this instead.

Change-Id: I80b3453bb237b4995d642b947d03c17d82cda5e7
Reviewed-on: http://gerrit.openafs.org/4266
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>