]> git.michaelhowe.org Git - packages/o/openafs.git/log
packages/o/openafs.git
16 years agoshutdown_icl returns void
Michael Meffie [Mon, 30 Nov 2009 23:15:59 +0000 (18:15 -0500)]
shutdown_icl returns void

Remove incorrect return statement from shutdown_icl,
introduced in b7198421a222a06cf5f34c53e516c31f5edf587c
Fix memory allocation warnings at shutdown.

Change-Id: Ib57161cb70152e08c1deba30bc0457efb98ebc97
Reviewed-on: http://gerrit.openafs.org/877
Reviewed-by: Marc Dionne <marc.c.dionne@gmail.com>
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
16 years agoModify Solaris Rx NetIfPoller to avoid Panic
Jeffrey Altman [Wed, 15 Jul 2009 07:58:02 +0000 (03:58 -0400)]
Modify Solaris Rx NetIfPoller to avoid Panic

This patch changes the common error reporting type from CE_PANIC
to CE_WARN and adds appropriate cleanup processing so that we do not
leak resources if an error occurs.

LICENSE MIT
FIXES 124498

Reviewed-on: http://gerrit.openafs.org/97
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit ccb3e316868f381831d9ddda530af83102f67697)

Change-Id: I8e289d256e5cbc3d3724cafe7114a77f0ddd900a
Reviewed-on: http://gerrit.openafs.org/878
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoAvoid IHandle leak when failing to attach volume
Rainer Toebbicke [Fri, 27 Nov 2009 14:03:25 +0000 (15:03 +0100)]
Avoid IHandle leak when failing to attach volume

When a volume fails to attach, ensure it's volume handles
get released before all traces of it are erased. Failure to do so
might cause problems once the volume is salvaged and re-attached.
VReleaseVolumeHandles_r is moved to FreeVolume and no longer called separately.

Change-Id: I2366f13fd9daa6eb8ce3b4240bb1534f14f706c4
Reviewed-on: http://gerrit.openafs.org/874
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoCheck for (hostFlags & HOSTDELETED) after h_Lock_r
Jeffrey Altman [Fri, 23 Oct 2009 14:54:35 +0000 (09:54 -0500)]
Check for (hostFlags & HOSTDELETED) after h_Lock_r

Many callers of h_Lock_r do not check if the HOSTDELETED flag is set,
even though it could have been set while waiting for the host lock. Add
checks for it everywhere we call h_Lock_r and we care if the host has
been deleted.

FIXES 125507

Reviewed-on: http://gerrit.openafs.org/716
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit cbe580fee176c9e9e083379129c309fd15d0f24c)

Change-Id: I5eacee7a33f7b1357ec13cfcc4e1b7b4538837f6
Reviewed-on: http://gerrit.openafs.org/776
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoAdd array bounds checking in h_Enumerate
Dan Hyde [Thu, 29 Oct 2009 16:07:47 +0000 (12:07 -0400)]
Add array bounds checking in h_Enumerate

When hostList is not properly NULL-terminated, the current code does
not protect from buffer overflow.  The following patch prevents buffer
overflow, prints a message, and asserts.

On our Linux hosts, we never reached the original assert, as there is
a problem handling the segfault the buffer overflow causes.

FIXES 125506

Reviewed-on: http://gerrit.openafs.org/757
Reviewed-by: Dan Hyde <drh@umich.edu>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 93d48b1735b88b3051357307bdbebf5da9872d69)

Change-Id: Idb920d2bd309b40ae9bebc6972e614ddf9ba8dcc
Reviewed-on: http://gerrit.openafs.org/862
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agovolser transaction object race conditions
Michael Meffie [Wed, 18 Nov 2009 15:07:14 +0000 (10:07 -0500)]
volser transaction object race conditions

Fix the transaction object races between VolMonitor and the
volume operation procedures which can cause the volume
server to crash.

Add a per transaction object mutex to safely set the
transaction call pointer and name. Fix VolMonitor to safely
traverse the transaction list and to access the call pointer
and last proc name while copying info to send to the vos
client.  Fix the sleep thread to safely access the last proc
name.

FIXES 125479

Change-Id: I6dffa31a84d98249712dd17aad353f99151b4fd5
Reviewed-on: http://gerrit.openafs.org/619
Reviewed-by: Alistair Ferguson <alistair.ferguson@mac.com>
Reviewed-by: Tom Keiser <tkeiser@sinenomine.net>
Reviewed-by: Andrew Deason <adeason@sinenomine.net>
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoMake ktc_curpag also detect ONEGROUP PAG gids
Andrew Deason [Wed, 11 Nov 2009 17:23:49 +0000 (11:23 -0600)]
Make ktc_curpag also detect ONEGROUP PAG gids

ktc_curpag falls back to looking at the group list if the VIOC_GETPAG
pioctl fails. If we're in AFS_LINUX26_ONEGROUP_ENV in the kernel,
though, ktc_curpag still looks for two groups, instead of the one
combined group. Add a check for the big one group in the fallback if
we're on LINUX26.

Reviewed-on: http://gerrit.openafs.org/815
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Marc Dionne <marc.c.dionne@gmail.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit c53abbb3e8d14643f3f82999cbe24fb12d358ce0)

Change-Id: I771d90546b1aec9f3e84a56a5b6a465e18b1241a
Reviewed-on: http://gerrit.openafs.org/853
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoClean up console message
Simon Wilkinson [Mon, 26 Oct 2009 12:49:11 +0000 (12:49 +0000)]
Clean up console message

Clean up the 'all buffers locked' console message so that it's
clear that it comes from AFS, and so it has a trailing newline.

From a bug report by Rainer Toebbicke to openafs-info

Change-Id: I3f69e57e37a91ae3e6e8822ccfe72dddc06b4359
Reviewed-on: http://gerrit.openafs.org/732
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 4a006f167bf6fe550eac9c96c6fce1f538760709)
Reviewed-on: http://gerrit.openafs.org/823
Reviewed-by: Dan Hyde <drh@umich.edu>
16 years agocm: address race condition in afs_QueueVCB
Michael Meffie [Mon, 9 Nov 2009 21:28:33 +0000 (16:28 -0500)]
cm: address race condition in afs_QueueVCB

Access the vcache callback member after taking
the xvcb lock to avoid the server object from
being freed in FlushServer on another thread.

Eventually, we should have a ref count on
avc->server.

FIXES 125596

Change-Id: I9328a0b3e372353b7812d6e9f2cbea44655a3d58
Reviewed-on: http://gerrit.openafs.org/817
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
16 years agoDo not check *aoutSize in PGetPAG
Andrew Deason [Wed, 11 Nov 2009 16:51:19 +0000 (10:51 -0600)]
Do not check *aoutSize in PGetPAG

*aoutSize is always zero in pioctls, since afs_HandlePioctl handles
checking the output buffer size, and sets outSize to 0 before calling
the pioctl. So, PGetPAG was always returning E2BIG; remove the check to
make it work.

Reviewed-on: http://gerrit.openafs.org/814
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 97d3fcfcd1617bafa6ee62d291f09b2e8bb6daf7)

Change-Id: Icc401904c26027e235ec869e834297861a640d39
Reviewed-on: http://gerrit.openafs.org/816
Reviewed-by: Dan Hyde <drh@umich.edu>
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoFix memory allocation warnings at shutdown
Marc Dionne [Sat, 31 Oct 2009 12:54:52 +0000 (08:54 -0400)]
Fix memory allocation warnings at shutdown

At shutdown we check for unfreed memory allocated with AllocSmallSpace
and AllocLargeSpace and complain in the syslog if there are dangling
pieces.  This patch takes care of a few cases that always showed up
as warnings, even after a simple start-stop of the client.

- The cacheInode file needs to be closed before the checks, since it
uses a large piece for its struct file.

- The ICL logging code allocates 6 small pieces that are never freed.
Add a shutdown_icl() function that releases everything.  While we're
at it, correct one place where we allocated with afs_osi_Alloc but
freed with osi_FreeSmallSpace, confusing our accounting.

Change-Id: I4c28c848f155dec9d89e6199cde34209227ca5c9
Reviewed-on: http://gerrit.openafs.org/758
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-on: http://gerrit.openafs.org/767

16 years agoLinux: Fix write_begin configure test for recent RHEL kernels
Marc Dionne [Sat, 31 Oct 2009 17:27:18 +0000 (13:27 -0400)]
Linux: Fix write_begin configure test for recent RHEL kernels

Recent RHEL kernels now define simple_write_begin, which was used as
a test for the write_begin address_space op.  This makes the test
succeed when it shouldn't, and breaks the build.
Rewrite the test to actually check the address_space operation.

Change-Id: Idac9b318ff716b61bf8ca4508d2dbdbfbad5b50d
Reviewed-on: http://gerrit.openafs.org/759
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Tested-by: Marc Dionne <marc.c.dionne@gmail.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-on: http://gerrit.openafs.org/766
Tested-by: Derrick Brashear <shadow@dementia.org>
16 years agomacos 10.6 64bit trace fixes
Derrick Brashear [Wed, 4 Nov 2009 00:25:23 +0000 (19:25 -0500)]
macos 10.6 64bit trace fixes

the 64 bit pointer massaging was incomplete for kernel
tracing on macos. add in the missing cases.

Change-Id: I38eb608c233819ab437adec25e2cd3afac3e234e
Reviewed-on: http://gerrit.openafs.org/771
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-on: http://gerrit.openafs.org/780

16 years ago.gitignore for src/pinstall
Russ Allbery [Wed, 4 Nov 2009 05:00:44 +0000 (21:00 -0800)]
.gitignore for src/pinstall

Ignore the pinstall binary in src/pinstall.  This is a 1.4-specific
change since pinstall has been removed from master.

Change-Id: I92ea9745174268e60ceb2b7aeff56d1263ff9072
Reviewed-on: http://gerrit.openafs.org/779
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoAlways use kbuild for all Linux kernel configure probes
Aaron M. Ucko [Tue, 7 Jul 2009 00:41:49 +0000 (17:41 -0700)]
Always use kbuild for all Linux kernel configure probes

Some Linux kernel probes for the existence of header files were done
with file existence checks (test -f).  This breaks if the kernel build
system is stacking multiple directories of headers together with
compile-time -I include path options, as is the case for the current
Debian Linux header packages.  Instead, always use kbuild to check
whether a kernel header is available.

Similarly, use AC_TRY_KBUILD instead of AC_TRY_COMPILE when checking
for an SELinux kernel, since AC_TRY_COMPILE doesn't call into kbuild
and won't get the correct kernel header paths.

This is part of the fix for Debian Bug#521745 and has been included in
the Debian package since 1.4.10+dfsg1-1.

Reviewed-on: http://gerrit.openafs.org/5
Verified-by: Russ Allbery <rra@stanford.edu>
Verified-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit b7942950c2d1fcaa1e94ebbe91c8d2a9a7670cdb)
Change-Id: If42353162712271d88ffc0d1c743b903c03c75c6
Reviewed-on: http://gerrit.openafs.org/777
Tested-by: Russ Allbery <rra@stanford.edu>
16 years agoLinux - Fix disk cache access for selinux/AppArmor constrained processes
Marc Dionne [Wed, 28 Oct 2009 21:54:32 +0000 (17:54 -0400)]
Linux - Fix disk cache access for selinux/AppArmor constrained processes

Preserve the credentials used for cache initialisation and use then
whenever disk cache files are opened.  This takes advantage of the
credentials separation work from David Howells available in kernels
2.6.29 and above.
Access to cache files was done under the security context of the
user process, causing processes constrained by selinux or AppArmor to
fail to access AFS cache files and causing the cache manager to panic.

Besides the RT tickets, should also fix the following Ubuntu bugs:
 415766 429260 457779 459299

FIXES 92944,125544

Change-Id: Ief8acd65c1a3e4d8c951f80bfd65f8340b8cec34
Reviewed-on: http://gerrit.openafs.org/752
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
Reviewed-on: http://gerrit.openafs.org/774

16 years agoBuild shadow header files when necessary on Linux
Aaron M. Ucko [Tue, 7 Jul 2009 00:51:00 +0000 (17:51 -0700)]
Build shadow header files when necessary on Linux

The current kernel module build infrastructure relies on the ability to
create symlinks from known directory names used in the AFS code to the
actual locations of the kernel header files.  This breaks if there is no
single kernel header tree and instead multiple trees are layered together
by kbuild using compile-time -I include paths.

Attempt to detect this case by seeing if linux/types.h is in the kernel
header directory where we expect it.  If not, rather than creating
symlinks for h, sys, and netinet, create directories and populate them
with single-line headers that just include the corresponding linux/*.h
header.  The list of headers for which to do this is generated dynamically
by analyzing the AFS kernel source code and looking for relevant #include
directives.

This patch has been part of the Debian OpenAFS package since
1.4.10+dfsg1-1.  The check for whether we have layered kernel header trees
may be specific to Debian and may require modification later if other
Linux distributions do something similar.

FIXES 124583

Reviewed-on: http://gerrit.openafs.org/6
Verified-by: Russ Allbery <rra@stanford.edu>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Verified-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 01d390d0d022309d06760aac1373829a35a6d79c)
Change-Id: I5a5ba79e88f8aafc07502c7d1c82058184c01da0
Reviewed-on: http://gerrit.openafs.org/775
Tested-by: Russ Allbery <rra@stanford.edu>
16 years agoMake afsd.pod reflect reality
Simon Wilkinson [Wed, 28 Oct 2009 11:12:18 +0000 (11:12 +0000)]
Make afsd.pod reflect reality

9d396c4916fdac64fcface30e6637ca6e2911203 (from 2005) introduced
autotuning for afsd, and changed some of the defaults which aren't
autotuned. Update the afsd man page to reflect the autotuning, and
the new defaults.

Change-Id: Iea2035743cb45cca1c249bc2e838405039ad7d3a
Reviewed-on: http://gerrit.openafs.org/744
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 7833e472b9585ee1a7085ca704e642a73ca969a5)
Reviewed-on: http://gerrit.openafs.org/773

16 years agoAvoid using released hosts
Andrew Deason [Wed, 28 Oct 2009 16:06:47 +0000 (11:06 -0500)]
Avoid using released hosts

Since h_Release_r has the possibility of freeing a host, we should not
be using a host after it has been released. A few places can still use a
released host, potentially causing heap corruption, double frees, and
generally weird behavior.

So either move calls of h_Release_r until after we finish using a host,
or make sure to set the pointer to NULL after it has been released.

Reviewed-on: http://gerrit.openafs.org/747
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Dan Hyde <drh@umich.edu>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 416e2f11c35f5d55f91090b30b4db1a9bf6d6e07)

Change-Id: I91bd09c3e6d87476de8c66c2eb710c0fd424cadd
Reviewed-on: http://gerrit.openafs.org/762
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoProtect rx_call iovq from simultaneous attempts to empty it
Jeffrey Altman [Mon, 27 Jul 2009 16:54:16 +0000 (12:54 -0400)]
Protect rx_call iovq from simultaneous attempts to empty it

The iovq queue is not safe to clear without holding the call lock.

FIXES 125110

LICENSE BSD

Reviewed-on: http://gerrit.openafs.org/242
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Hans-Werner Paulsen <hans@MPA-Garching.MPG.DE>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
(cherry picked from commit 4dadd24ba8ce72fd655e29d74801f27e9e148b01)
Change-Id: Ifdb1f420f07f9c55d56fe24899e3a040a77db6a9
Reviewed-on: http://gerrit.openafs.org/763
Tested-by: Andrew Deason <adeason@sinenomine.net>
16 years agoDec old special inodes in inode convertROtoRW
Andrew Deason [Mon, 26 Oct 2009 19:04:48 +0000 (14:04 -0500)]
Dec old special inodes in inode convertROtoRW

The convertROtoRW code for the inode fileserver makes copies of the
volume's special inodes, but leaves the old (RO) inodes around. If the
RO is created again, this will result in duplicate special inodes for
the same volume, which freaks out the salvager (and possibly other
things).

So IH_DEC the old RO special inodes after converting, so they go away.

Reviewed-on: http://gerrit.openafs.org/735
Tested-by: Andrew Deason <adeason@sinenomine.net>
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit dbe3b7b8eeb4a010f82248befc6167b3b5ed9606)

Change-Id: Id545befbb38f60c262ab23b5eb5dd54439008760
Reviewed-on: http://gerrit.openafs.org/761
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoLinux: Avoid deadlock in readdir - release GLOCK for filldir
Marc Dionne [Sun, 1 Nov 2009 21:03:17 +0000 (16:03 -0500)]
Linux: Avoid deadlock in readdir - release GLOCK for filldir

The GLOCK is held while calling the filldir function in afs_linux_readdir().
If this function causes a page fault, and in particular if this fault
involves AFS, we're in trouble as we'll eventually deadlock in the
readpage code.
A simple test case for this is to call the getdents syscall on an
AFS directory with a buffer that is part of an mmaped AFS file.

This is already the case in the master branch; the change was part of
the merge of the NFS translator code.

FIXES 125555

Change-Id: I829838e45f94921d22335154587216f7842e3955
Reviewed-on: http://gerrit.openafs.org/760
Tested-by: Marc Dionne <marc.c.dionne@gmail.com>
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoLinux: 2.6.32 - Adapt to writeback changes
Marc Dionne [Thu, 17 Sep 2009 20:57:52 +0000 (16:57 -0400)]
Linux: 2.6.32 - Adapt to writeback changes

Adapt to the writeback changes in kernel 2.6.32
- Since we define our own backing_dev, it needs to be registered with
the writeback code and attached to the super_block.  Otherwise it
might get ignored when writeback is needed.
- Each backing_dev now gets its own kernel thread.  The name of the
thread is based on the registered name - the openafs one will appear
as "flush-afs".

Reviewed-on: http://gerrit.openafs.org/574
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Change-Id: Ib1403a65991fc7c2cb73099feef7e0dbec02543d
Reviewed-on: http://gerrit.openafs.org/755

16 years agoAvoid prematurely destroying callback_rxcon
Andrew Deason [Thu, 22 Oct 2009 16:12:30 +0000 (11:12 -0500)]
Avoid prematurely destroying callback_rxcon

Currently, h_GetHost_r and removeAddress_r can destroy the
callback_rxcon of a host. Having a NULL callback_rxcon can cause
segfaults in code that does not properly check if a host has been
HOSTDELETED before trying to use it.

Although such code is incorrect and should be fixed, we can still avoid
a segfault in those situations by not destroying callback_rxcon until we
destroy the host itself. This just prevents destroying callback_rxcon in
h_GetHost_r and removeAddress_r, leaving it to h_TossStuff_r to destroy
when it destroys the host.

Reviewed-on: http://gerrit.openafs.org/717
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit bbcfbe1a04eda9e75b1643be88cf9d4842a8aa86)

Change-Id: I4d44fca9b79b656bc70f8108616f23f2dfaa353f
Reviewed-on: http://gerrit.openafs.org/749
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoHPUX: Do not sigwait on critical signals
Andrew Deason [Tue, 20 Oct 2009 17:43:42 +0000 (12:43 -0500)]
HPUX: Do not sigwait on critical signals

On HPUX, it is possible for 'critical' signals such as SEGV, ABRT, etc
to be delivered to the softsig thread when we sigwait(). The current
code marks these as 'fatal' and just exit(0)s when they are received,
preventing us from getting cores in the case of a SEGV, ABRT, etc.

To work around this and keep behavior on other platforms the same, just
do not wait on 'critical' signals on HPUX in the softsig thread.

Reviewed-on: http://gerrit.openafs.org/693
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit bf9c51a4e13b3e621b99866e9be53c8fe35a39fe)
Reviewed-on: http://gerrit.openafs.org/702

16 years agoUse f_bsize for ZFS afs_fsfragsize
Andrew Deason [Thu, 24 Sep 2009 17:02:55 +0000 (12:02 -0500)]
Use f_bsize for ZFS afs_fsfragsize

On ZFS, the disk space files can use up can be rounded up to the next
recordsize boundary if they've been truncated. This can cause the Unix CM
to mis-estimate cache usage, since it truncates files fairly often, and
assumes the disk space used is the file length rounded up to the next
f_frsize.

Since the ZFS recordsize is available via the statvfs f_bsize, just
round up to that instead. There is still some additional file metadata
that takes up some additional space on disk, but according to ZFS people
I've spoken to about this, it cannot be known in advance. In practice,
the additional metadata storage doesn't appear to exceed about 10% of
the data storage, which should be acceptable.

FIXES 125365

Reviewed-on: http://gerrit.openafs.org/650
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Andrew Deason <adeason@sinenomine.net>
(cherry picked from commit 78d8b8c5c19b8d6873b2d62beaa60cc7acba1840)
Reviewed-on: http://gerrit.openafs.org/655
Tested-by: Derrick Brashear <shadow@dementia.org>
16 years agoRemove extra arguments to afs_syscall_call
Andrew Deason [Tue, 13 Oct 2009 16:20:51 +0000 (11:20 -0500)]
Remove extra arguments to afs_syscall_call

Someone appears to have mistaken afs_syscall_call for
afs_syscall_pioctl, and passed rvp and CRED to afs_syscall_call on
solaris (afs_syscall_call always takes 6 arguments; afs_syscall_pioctl
takes various arguments depending on the platform). After
b1ff4a0b1115f5739c0365cc963189b1f931971f, this breaks the client build
on AFS_SUN5_ENV, but only because it added prototypes for
afs_syscall_call; it looks like it's always been wrong, but it was never
noticed.

Reviewed-on: http://gerrit.openafs.org/653
Tested-by: Andrew Deason <adeason@sinenomine.net>
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoMake namei_ops.c build again
Andrew Deason [Tue, 13 Oct 2009 15:56:47 +0000 (10:56 -0500)]
Make namei_ops.c build again

Commit 10441428634e6f474457a1fbf38ea693279d1ad1 used a 1.5-only
function, so did not build on 1.4 (sorry). Fix that.

Reviewed-on: http://gerrit.openafs.org/652
Tested-by: Andrew Deason <adeason@sinenomine.net>
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoMake namei convertROtoRW'd volumes usable
Andrew Deason [Wed, 7 Oct 2009 17:14:11 +0000 (12:14 -0500)]
Make namei convertROtoRW'd volumes usable

Right now, if you convertROtoRW a volume on namei, the converted volume
appears to need a salvage before it is usable, and the header of the old
(now empty) RO volume is kept around. Fix this:

 -- Set inUse = 0 on the converted volume, so the fileserver will be
    able to attach the volume when we give it back

 -- Unlink the RO header file, instead of trying to unlink the
    VI_VOLINFO file twice

 -- Log the actual error code (errno) in the error message for the last
    unlink

Reviewed-on: http://gerrit.openafs.org/591
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 906403c6d770026fd9da606ea468e47ee911f896)
Reviewed-on: http://gerrit.openafs.org/613

16 years agoviced cap fetchdata len to avoid negative
Derrick Brashear [Fri, 25 Sep 2009 07:15:40 +0000 (03:15 -0400)]
viced cap fetchdata len to avoid negative

when trying to read from a short or zero length file,
it's possible for sanity checking to knock the length below
zero. set a floor at zero.

Reviewed-on: http://gerrit.openafs.org/494
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 529d487d65d8561f5d0a43a4dc71f72b86efd975)
Reviewed-on: http://gerrit.openafs.org/500

16 years agoh_GetHost_r cleanup cases
Russ Allbery [Sat, 26 Sep 2009 00:50:54 +0000 (17:50 -0700)]
h_GetHost_r cleanup cases

avoid potentially leaking a cb_in or a capabilities blob;
use a common exit function and free when these were left set.

Reviewed-on: http://gerrit.openafs.org/496
Tested-by: Alistair Ferguson <alistair.ferguson@mac.com>
Reviewed-by: Alistair Ferguson <alistair.ferguson@mac.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
(cherry picked from commit 5d765426e597c60228e66bddaa21bf2453c74a85)
Reviewed-on: http://gerrit.openafs.org/497
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoImplement _PC_FILESIZEBITS for solaris pathconf
Andrew Deason [Mon, 21 Sep 2009 21:57:01 +0000 (16:57 -0500)]
Implement _PC_FILESIZEBITS for solaris pathconf

Using recent NFS clients and servers with the translator under Solaris
causes AFS to be queried for the _PC_FILESIZEBITS pathconf value. Right
now we don't implement it and return EINVAL, causing at least some
modern NFS clients to be unable to mount AFS via the translator on at
least some modern NFS servers.

So, return _PC_FILESIZEBITS as either 32 or 64, depending on whether we
are a 64-bit client or not.

Reviewed-on: http://gerrit.openafs.org/481
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 40b18318e4a04a323dbdd9e3c4a73d3b2e92104d)
Reviewed-on: http://gerrit.openafs.org/490

16 years agoAdd automatic sysname detection for ARM Linux
Russ Allbery [Sat, 22 Aug 2009 01:37:41 +0000 (18:37 -0700)]
Add automatic sysname detection for ARM Linux

Add arm*-linux* to the case statement that attempts to automatically
determine the AFS sysname, similar to the other Linux sysname
determination cases.

Reviewed-on: http://gerrit.openafs.org/334
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit a59df36211d6bc641eeec6d8fd8876a98ce3600a)
Reviewed-on: http://gerrit.openafs.org/462

16 years agofileserver should actually retry VL_RegisterAddrs on failure
Derrick Brashear [Mon, 3 Aug 2009 16:17:14 +0000 (12:17 -0400)]
fileserver should actually retry VL_RegisterAddrs on failure

currently the background daemon logic for retrying was never triggered.
easy enough fix: set FS_registered to 1 so the loop retries

Reviewed-on: http://gerrit.openafs.org/266
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 160aff88be210a8141b67a54583aed7387d8bffc)
Reviewed-on: http://gerrit.openafs.org/459

16 years agoInit the vrequest structure correctly
Rainer Toebbicke [Mon, 14 Sep 2009 12:53:56 +0000 (13:53 +0100)]
Init the vrequest structure correctly

FIXES 125355
(1.4.x only)

afs_InitReq fails to initialize the "flags" field of the vrequest structure.

Consequently the logic involving (flags & O_NONBLOCK) in afs_Analyze leads to
unpredictable results. afs_InitReq should initialize the complete vrequest
structure.

Reviewed-on: http://gerrit.openafs.org/457
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoUpdate accessDate on volume access
Andrew Deason [Thu, 3 Sep 2009 19:43:28 +0000 (14:43 -0500)]
Update accessDate on volume access

Right now accessDate is simply never updated, so the last access time
for a volume is never reported. Simply update the field in
VBumpVolumeUsage_r, so we track the last time the volume was accessed.

Note that this does not increase disk writes to the volume header; the
performance impact is effectively nil.

Reviewed-on: http://gerrit.openafs.org/385
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 5d9c767ac55b2d1b8f5b87dc9ced63e282af7bac)
Reviewed-on: http://gerrit.openafs.org/398

16 years agoFixed out-of-tree builds.
Felix Frank [Wed, 26 Aug 2009 18:34:01 +0000 (20:34 +0200)]
Fixed out-of-tree builds.

Fix a minor error that originates from replacing Makefile.version.CML
with Makefile.version.CML.in

(cherry picked from commit5bc9641620d8e1c56a4af6857cb4c79109c79194)

Reviewed-on: http://gerrit.openafs.org/342
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
16 years agoBuild on linux 2.4 again
Michael Meffie [Fri, 17 Jul 2009 15:06:34 +0000 (11:06 -0400)]
Build on linux 2.4 again

Add missing conditional to fix building on
linux 2.4.

FIXES 125120

Reviewed-on: http://gerrit.openafs.org/http://gerrit.openafs.org/148
Tested-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Tested-by: Russ Allbery <rra@stanford.edu>
Reviewed-by: Russ Allbery <rra@stanford.edu>
(cherry picked from commit f6e46babd8e3a0d51d31ae381748516330f5b84d)
Reviewed-on: http://gerrit.openafs.org/386
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agomacos 10.5 doesn't support compiler kext flag
Derrick Brashear [Fri, 28 Aug 2009 05:01:46 +0000 (01:01 -0400)]
macos 10.5 doesn't support compiler kext flag

-kext was added in later xcodes; it's safer to just assume we don't
have it.
(cherry picked from commit eefb4d83c5ee7fbf3b575773eeec2b126373557a)

Reviewed-on: http://gerrit.openafs.org/366
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoupdate decode-panic for 10.6
Derrick Brashear [Thu, 27 Aug 2009 19:47:58 +0000 (15:47 -0400)]
update decode-panic for 10.6

how panics get reported changed in 10.6. also, the 64 bit universal kext
means we don't know what the last boot was.

Reviewed-on: http://gerrit.openafs.org/353
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit a747f5951b920dfb6d77465de6ca8ff54bae1706)
Reviewed-on: http://gerrit.openafs.org/363

16 years agomacos 10.6 64 bit support
Derrick Brashear [Wed, 26 Aug 2009 21:34:50 +0000 (17:34 -0400)]
macos 10.6 64 bit support

support 64 bit userland with 32 bit kernel, and
32 and 64 bit userland with 64 kit kernel

Reviewed-on: http://gerrit.openafs.org/343
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 81099c479e148a103dcf9f4def125b28e989107c)
Reviewed-on: http://gerrit.openafs.org/362

16 years agomacos rc script server handling
Derrick Brashear [Wed, 26 Aug 2009 19:15:27 +0000 (15:15 -0400)]
macos rc script server handling

only start bosserver if its configured. shut it down if we (likely) started it

Reviewed-on: http://gerrit.openafs.org/344
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 9473089bdf1997b2f43e3ea368907c6cdc0fa582)
Reviewed-on: http://gerrit.openafs.org/361

16 years agomacos 10.6 updates
Derrick Brashear [Fri, 14 Aug 2009 16:06:03 +0000 (12:06 -0400)]
macos 10.6 updates

cope with mixed 32/64 environment provided by macos 10.6. build needed
stuff for audit logs in shlibafsauthent.

Reviewed-on: http://gerrit.openafs.org/310
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 3a48ae3e243b695ac5bb3b35aa3b09641db1af3b)
Reviewed-on: http://gerrit.openafs.org/360

16 years agomacos 10.6 package naming
Derrick Brashear [Thu, 30 Jul 2009 02:56:08 +0000 (22:56 -0400)]
macos 10.6 package naming

call the dmg Snowleopard for 10.6.

Reviewed-on: http://gerrit.openafs.org/252
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit e18216b91fa7171efdf796908ce2e98e7bc42f11)
Reviewed-on: http://gerrit.openafs.org/359

16 years agomacos 10.6 amd64 kmod build fix
Derrick Brashear [Thu, 30 Jul 2009 02:14:01 +0000 (22:14 -0400)]
macos 10.6 amd64 kmod build fix

make the ARCHFLAGS support for amd64 work correctly by grepping for the correct thing

Reviewed-on: http://gerrit.openafs.org/250
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 6bec05a874035b30a1b7d5119a053f550adeba4d)
Reviewed-on: http://gerrit.openafs.org/358

16 years agomacos 10.6 pam support
Derrick Brashear [Wed, 29 Jul 2009 21:23:14 +0000 (17:23 -0400)]
macos 10.6 pam support

Make pam module compile on 10.6. Not tested yet.

Reviewed-on: http://gerrit.openafs.org/249
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 5593ff360c7a8f9f69cebda0cb2e2909ff2f31b4)
Reviewed-on: http://gerrit.openafs.org/357

16 years agoOSX lock initialization cleanup
Derrick Brashear [Tue, 21 Jul 2009 21:09:39 +0000 (17:09 -0400)]
OSX lock initialization cleanup

Move GLOCK creation before other things at modload time. Also
remove unused rwlock support from rx kmutex package.

Reviewed-on: http://gerrit.openafs.org/http://gerrit.openafs.org/170
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 322a7a854b348e68b0bc49f5ad37049af188518f)
Reviewed-on: http://gerrit.openafs.org/356

16 years agoMacOS 10.6 support update
Derrick Brashear [Wed, 15 Jul 2009 18:57:56 +0000 (14:57 -0400)]
MacOS 10.6 support update

Make things build on MacOS 10.6 again. Don't break 10.5 build.
Fix an error in the MacOS port which was left from previous head.

Reviewed-on: http://gerrit.openafs.org/96
Verified-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Russ Allbery <rra@stanford.edu>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 1d1ae6195bd03c4348a57914cf892e06b507d855)
Reviewed-on: http://gerrit.openafs.org/355
Tested-by: Derrick Brashear <shadow@dementia.org>
16 years agoAdd additional vlprocs safety checks
Andrew Deason [Mon, 20 Jul 2009 17:31:44 +0000 (12:31 -0500)]
Add additional vlprocs safety checks

This adds additional safety checks to the vlserver's implementation of
the VL_CreateEntry, VL_ReplaceEntry, and VL_UpdateEntry RPCs. Now in all
three of these, any new volume ID that would be added to the VLDB or
that would be newly referenced in a VLDB entry is checked against
duplication in other entries. Additionally, any new volume names added
to the VLDB (either by creation, or modifying an existing volume) are
checked against duplication. This should make it impossible for clients
to make a volume ID or volume name correspond to multiple volume groups
(either conceptually or literally in the vldb).

This also alters the vlserver's implementation of the VL_GetNewVolumeId
RPC such that the vlserver increments maxvolid until the range of volume
IDs [*newvolumeid, *newvolumeid+bumpcount) is unused. 'vos' is modified
to only allocate one new volume id at a time, so we don't skip over
potentially-usable vol ids.

Reviewed-on: http://gerrit.openafs.org/158
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit ef7e0d64e7c29675405299f36ee3b7343b31cb35)
Reviewed-on: http://gerrit.openafs.org/280

16 years agoAllow specifying vos create/addsite volume IDs
Andrew Deason [Mon, 6 Jul 2009 15:29:20 +0000 (10:29 -0500)]
Allow specifying vos create/addsite volume IDs

This adds the -id option to 'vos create', and the -roid option to 'vos
create' and 'vos addsite'. This allows the user to manually specify the
volume IDs that a new RW or RO volume will get (or explicitly specify
that an RO volume ID should be unset), instead of always relying on the
volume IDs retrieved from the vlserver.

Reviewed-on: http://gerrit.openafs.org/157
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 29a5ddf1ddc05dd1ea826220a424f4f6650cc0c7)
Reviewed-on: http://gerrit.openafs.org/275

16 years agoIgnore SIGSYS when issuing pioctl syscall
Andrew Deason [Tue, 14 Jul 2009 16:29:01 +0000 (11:29 -0500)]
Ignore SIGSYS when issuing pioctl syscall

Ignore SIGSYS when we issue the pioctl syscall, so we don't dump core
when the kernel module hasn't yet been installed on several platforms.
Also, restore the old SIGSYS signal handler afterwards, so we don't
cause any side-effects.

Reviewed-on: http://gerrit.openafs.org/81
Verified-by: Andrew Deason <adeason@sinenomine.net>
Verified-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Russ Allbery <rra@stanford.edu>
(cherry picked from commit 4f36dd089a9c7187f94f77516a486245c057f7f4)
Reviewed-on: http://gerrit.openafs.org/274
Tested-by: Andrew Deason <adeason@sinenomine.net>
16 years agoCorrecting formatting typo in vos addsite manpage
Andrew Deason [Wed, 8 Jul 2009 21:15:37 +0000 (16:15 -0500)]
Correcting formatting typo in vos addsite manpage

Reviewed-on: http://gerrit.openafs.org/12
Verified-by: Russ Allbery <rra@stanford.edu>
Reviewed-by: Russ Allbery <rra@stanford.edu>
(cherry picked from commit 722070a5c6b1c87cb8c80e8938464183dab01c8f)
Reviewed-on: http://gerrit.openafs.org/273
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoAlways display vnode accesses in vos output
Andrew Deason [Tue, 7 Jul 2009 22:34:17 +0000 (17:34 -0500)]
Always display vnode accesses in vos output

Vos was not displaying the "N accesses in the past day (i.e., vnode
references)" message when updateDate was unset. Fix that.

FIXES 125064

Reviewed-on: http://gerrit.openafs.org/9
Verified-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit f52b97b76cba008c9aa6ae843869c95d83fbd9f9)
Reviewed-on: http://gerrit.openafs.org/272
Tested-by: Derrick Brashear <shadow@dementia.org>
16 years agoFixing manpage for vos addsite -valid
Andrew Deason [Wed, 8 Jul 2009 21:07:58 +0000 (16:07 -0500)]
Fixing manpage for vos addsite -valid

The -valid switch to vos addsite doesn't take an argument, and it is
optional. Correcting documentation to reflect that.

Reviewed-on: http://gerrit.openafs.org/11
Verified-by: Russ Allbery <rra@stanford.edu>
Reviewed-by: Russ Allbery <rra@stanford.edu>
(cherry picked from commit 7488aeef5cfd58935fe4a6a07810c8d757dd98a3)
Reviewed-on: http://gerrit.openafs.org/271
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoMake ktc_curpag generally available
Russ Allbery [Sun, 12 Jul 2009 18:26:03 +0000 (11:26 -0700)]
Make ktc_curpag generally available

ktc_curpag isn't specific to a Kerberos v4 environment, so move it outside
the AFS_KERBEROS_ENV #ifdef.  Add it to the auth.h header and to the
exports from the shared libafsauthent.

FIXES 125155

Reviewed-on: http://gerrit.openafs.org/64
Verified-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit cd37fd33f70c2196ff64188e32ff759590cb3748)
Reviewed-on: http://gerrit.openafs.org/208
Tested-by: Derrick Brashear <shadow@dementia.org>
16 years agoFix afs_GetVolume() for non-root dynroot FIDs
Jeffrey Hutzelman [Wed, 22 Jul 2009 14:19:28 +0000 (10:19 -0400)]
Fix afs_GetVolume() for non-root dynroot FIDs

Make afs_GetVolume() correctly handle requests for fids which are in
the dynroot cell but are not the root of the dynamic root volume.  This
is necessary to allow dynamic root mount points to be looked up and
followed in situations where the dynroot volume is not in the volcache,
but its root vnode is in the vcache.

Reviewed-on: http://gerrit.openafs.org/http://gerrit.openafs.org/180
Tested-by: Jeffrey Hutzelman <jhutz@cmu.edu>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoRemove the RCSID macro
Simon Wilkinson [Tue, 21 Jul 2009 23:04:24 +0000 (00:04 +0100)]
Remove the RCSID macro

The move to git means that we can no longer populate the RCSID
macro in the way that it was used with CVS. This patch simply
removes the macro from every file, except where it contains
information from upstream (and it's in a comment).

Reviewed-on: http://gerrit.openafs.org/14
Verified-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
(cherry picked from commit 3bb5fa5f1b3dc13ab20ed026c7ea64e11388b9c9)
Reviewed-on: http://gerrit.openafs.org/http://gerrit.openafs.org/174
Tested-by: Derrick Brashear <shadow@dementia.org>
16 years agoRevise git ignore files
Simon Wilkinson [Mon, 6 Jul 2009 13:38:42 +0000 (14:38 +0100)]
Revise git ignore files

Revise our git ignores to match the current state of the tree, and include
entires in the top level for all of the 'dest' directories for all of the
architectures we claim to support.

Reviewed-on: http://gerrit.openafs.org/2
Reviewed-by: Derrick Brashear <shadow@gmail.com>
Verified-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
(cherry picked from commit ba9ad6284528e669bfaacf0400b90fc08a5614e4)
Reviewed-on: http://gerrit.openafs.org/http://gerrit.openafs.org/172
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoRemove CVS ignore files
Simon Wilkinson [Mon, 6 Jul 2009 12:52:21 +0000 (13:52 +0100)]
Remove CVS ignore files

Now we're in git we don't need any cvsignore files any more...

Reviewed-on: http://gerrit.openafs.org/1
Verified-by: Derrick Brashear <shadow@gmail.com>
Reviewed-by: Derrick Brashear <shadow@gmail.com>
(cherry-picked from commit d5081f264dae17fed3daca5ed646f9c082557f0c)
Reviewed-on: http://gerrit.openafs.org/http://gerrit.openafs.org/173
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
16 years agoSTABLE14-openafs-1-4-11-20090710 openafs-stable-1_4_11
Derrick Brashear [Fri, 10 Jul 2009 19:00:15 +0000 (19:00 +0000)]
STABLE14-openafs-1-4-11-20090710

LICENSE IPL10

make 1.4.11

16 years agoSTABLE14-cellservdb-location-update-20090710
Derrick Brashear [Fri, 10 Jul 2009 19:26:48 +0000 (19:26 +0000)]
STABLE14-cellservdb-location-update-20090710

LICENSE IPL10

update cellservdb file location

16 years agoSTABLE14-openafs-1-4-11rc3-20090703 openafs-stable-1_4_11pre3
Derrick Brashear [Fri, 3 Jul 2009 14:20:02 +0000 (14:20 +0000)]
STABLE14-openafs-1-4-11rc3-20090703

LICENSE IPL10

make 1.4.11pre3

16 years agoSTABLE14-linux-vnop-return-value-cleanup-20090703
Simon Wilkinson [Fri, 3 Jul 2009 14:18:32 +0000 (14:18 +0000)]
STABLE14-linux-vnop-return-value-cleanup-20090703

LICENSE IPL10

rework all linux vnode ops so the vulnerability we previously had can't
recur later just because someone makes a change that would leak a negative
error

(cherry picked from commit ebc4f737ae8f8056de1a7f80828a6173ebb24292)

16 years agoSTABLE14-viced-ubik-clientdestroy-null-20090703
Andrew Deason [Fri, 3 Jul 2009 14:04:17 +0000 (14:04 +0000)]
STABLE14-viced-ubik-clientdestroy-null-20090703

LICENSE IPL10
FIXES 125020

null ubik client point when it's freed in hpr_End to avoid a null deref later

(cherry picked from commit 6ef1234402435306cb3935d1eadc5fcf332bd632)

16 years agoSTABLE14-redhat-spec-manpage-updates-20090703
Derrick Brashear [Fri, 3 Jul 2009 13:57:20 +0000 (13:57 +0000)]
STABLE14-redhat-spec-manpage-updates-20090703

LICENSE IPL10

add missing manpages to srpm

(cherry picked from commit 22196a7fd78dff342b2f9fca2b38d9a2dd04d42c)

16 years agoSTABLE14-linux-keyring-export-check-20090701
Marc Dionne [Wed, 1 Jul 2009 21:57:11 +0000 (21:57 +0000)]
STABLE14-linux-keyring-export-check-20090701

LICENSE IPL10
FIXES 125001

check for key_type_keyring being exported; only use it if so

(cherry picked from commit 4ca66112c2a012e04619445275507ae9cd6c11da)

16 years agoSTABLE14-rpmlint-try2-20090629
christof.hanke@rzg.mpg.de [Mon, 29 Jun 2009 21:24:47 +0000 (21:24 +0000)]
STABLE14-rpmlint-try2-20090629

LICENSE IPL10
FIXES 124951

remove unneeded return; fix iob initialization for curpag pioctl

(cherry picked from commit 58024c16844a7298f04495ff9023e396a5391aff)

16 years agoSTABLE14-revert-voldump-v-tag-generation-20090629
Derrick Brashear [Mon, 29 Jun 2009 16:28:22 +0000 (16:28 +0000)]
STABLE14-revert-voldump-v-tag-generation-20090629

LICENSE IPL10
FIXES 18349

generating v tags is currently incompatible. revert this for now
for 1.4, accept and discard 'V' tag

(cherry picked from commit f202b9778e4489fd80288c5be36e3c102b0cfba9)

16 years agoSTABLE14-allow-gcpags-disabling--at-compile-20090624 openafs-stable-1_4_11pre2
Derrick Brashear [Thu, 25 Jun 2009 13:13:34 +0000 (13:13 +0000)]
STABLE14-allow-gcpags-disabling--at-compile-20090624

LICENSE IPL10

put this back as was for now; if we want to undef, we can do it later, completel
y

(cherry picked from commit e43342e3778524a68219875b5fdc6fae0adbef1f)

16 years agoSTABLE14-openafs-1-4-11rc2-20090623
Derrick Brashear [Thu, 25 Jun 2009 00:16:58 +0000 (00:16 +0000)]
STABLE14-openafs-1-4-11rc2-20090623

LICENSE IPL10

make 1.4.11pre2

16 years agoSTABLE14-pag-get-lint-20090623
christof.hanke@rzg.mpg.de [Thu, 25 Jun 2009 00:15:06 +0000 (00:15 +0000)]
STABLE14-pag-get-lint-20090623

LICENSE IPL10
FIXES 124951

make things happy for rpmlint again (wrong initialization)

(cherry picked from commit 4ac7d59d464d66322a2925c86e955cd8d8067f75)

16 years agoSTABLE14-linux-rcu-read-lock-export-20090623
Marc Dionne [Wed, 24 Jun 2009 22:54:51 +0000 (22:54 +0000)]
STABLE14-linux-rcu-read-lock-export-20090623

LICENSE IPL10
FIXES 124986

deal with rcu_read_lock() becoming GPLONLY

(cherry picked from commit 0e718b642cedd1ebc799ada35960164c95c24280)

16 years agoSTABLE14-fbsd61-stdlib-proto-define-20090623
Harald Barth [Wed, 24 Jun 2009 22:45:14 +0000 (22:45 +0000)]
STABLE14-fbsd61-stdlib-proto-define-20090623

LICENSE IPL10
FIXES 124952

add define to make stdlib have malloc prototype

(cherry picked from commit a125b571a4d7c6c8c491bda079115a5eab835971)

16 years agoSTABLE14-osol-solookup-20090624
Andrew Deason [Wed, 24 Jun 2009 22:38:40 +0000 (22:38 +0000)]
STABLE14-osol-solookup-20090624

LICENSE IPL10
FIXES 124924

update for new opensolaris solookup function signature

(cherry picked from commit ee72f251049ca361b095e813adce63216343afd8)

16 years agoSTABLE14-rename-conn-to-afs-conn-20090121
Simon Wilkinson [Wed, 24 Jun 2009 22:30:13 +0000 (22:30 +0000)]
STABLE14-rename-conn-to-afs-conn-20090121

LICENSE IPL10
FIXES 124116

based on patch from daleg@elemental.org. rename to avoid conflict with opensolar
is

(cherry picked from commit a268678326f3713a0ba41f69a1432bd1fca2872f)

16 years agoSTABLE14-linux-call-inode-setattr-op-20090622
Marc Dionne [Mon, 22 Jun 2009 16:25:43 +0000 (16:25 +0000)]
STABLE14-linux-call-inode-setattr-op-20090622

LICENSE IPL10
FIXES 124942

call inode's setattr op instead of just inode_setattr, when one is available.
needed for xfs, notably also will cause truncates to be journalled for ext3,
which may solve some existing issues

(cherry picked from commit b02e22b5f0590929ef9120da4799ca9a47fa3aeb)

16 years agoSTABLE14-gitignore-20090615
Simon Wilkinson [Mon, 15 Jun 2009 23:37:24 +0000 (23:37 +0000)]
STABLE14-gitignore-20090615

LICENSE IPL10

make gitignore files

(cherry picked from commit c41a46774134aee6eaf6a41e9061386035ae6ebf)

16 years agoSTABLE14-vnode-disk-object-name-hi-length-field-20090611
Derrick Brashear [Thu, 11 Jun 2009 20:40:29 +0000 (20:40 +0000)]
STABLE14-vnode-disk-object-name-hi-length-field-20090611

LICENSE IPL10

reserved6 is used for the high portion of the 64 bit length. make the
structure reflect it.

(cherry picked from commit a7f9ce9ce1b2608937c5c3ffc73a2de07abc5fb1)

16 years agoSTABLE14-openafs-1-4-11rc1-20090611 openafs-stable-1_4_11pre1
Derrick Brashear [Thu, 11 Jun 2009 15:50:39 +0000 (15:50 +0000)]
STABLE14-openafs-1-4-11rc1-20090611

LICENSE IPL10

make 1.4.11pre1

16 years agoSTABLE14-bosserver-invoke-salvager-fully-specifying-force-20090611
Andrew Deason [Thu, 11 Jun 2009 15:19:35 +0000 (15:19 +0000)]
STABLE14-bosserver-invoke-salvager-fully-specifying-force-20090611

LICENSE IPL10
FIXES 124916

bos salvage can invoke salvager with -f (instead of -force) which is no
longer unique. fix that.

(cherry picked from commit 2c755529c79492738fec61a5fc311216678a52e7)

16 years agoSTABLE14-man-page-xstat-fs-test-20090608
Mike Meffie [Mon, 8 Jun 2009 23:41:37 +0000 (23:41 +0000)]
STABLE14-man-page-xstat-fs-test-20090608

LICENSE IPL10
FIXES 124905

Document xstat_fs_test collection 3, which returns callback statistics
from the file server (useful for tuning).

(cherry picked from commit 394243f3ae2dc5437b66d8f5afb393c2cfd1fb7b)

16 years agoSTABLE14-man-page-cellservdb-extras-20090605
Derrick Brashear [Fri, 5 Jun 2009 19:23:13 +0000 (19:23 +0000)]
STABLE14-man-page-cellservdb-extras-20090605

LICENSE IPL10

Document linked cells and non-voting replicas in the CellServDB man page
and note the need for better linked cell documentation.

(cherry picked from commit 6d973e499acd598bbc3a4e131d14f9801b94fe71)

16 years agoSTABLE14-aix51-no-userspace-pag-subsystem-20090604
Derrick Brashear [Thu, 4 Jun 2009 21:51:57 +0000 (21:51 +0000)]
STABLE14-aix51-no-userspace-pag-subsystem-20090604

LICENSE IPL10

getpagvalue() appeared post 5.1.

(cherry picked from commit 33e04c09a1f36281a6f37fdfd7b402426b9df5d4)

16 years agoSTABLE14-death-to-longc-procs-20060804
Russ Allbery [Thu, 4 Jun 2009 17:12:58 +0000 (17:12 +0000)]
STABLE14-death-to-longc-procs-20060804

longc_procs.h was only used on Ultrix, about which we no longer care.
Delete it, its installation rules, and the places where we were including
it.

(cherry picked from commit f3a77810db0044bd0fc3df1101269ab42b365fef)

16 years agoSTABLE14-ptserver-debug-flag-20090603
Andrew Deason [Wed, 3 Jun 2009 23:12:02 +0000 (23:12 +0000)]
STABLE14-ptserver-debug-flag-20090603

LICENSE IPL10
FIXES 124893

parse -d (debug) earlier in the ptserver argument parsing, so database
switch does not override it.

(cherry picked from commit 6124a81e791f54953ea4be3cf3588d9ccf5fb248)

16 years agoSTABLE14-ptserver-procs-loglevels-20090603
Andrew Deason [Wed, 3 Jun 2009 22:49:17 +0000 (22:49 +0000)]
STABLE14-ptserver-procs-loglevels-20090603

LICENSE IPL10
FIXES 124894

make loglevels for create/delete match those for other pts operations

(cherry picked from commit d9d50fea986569dfaa3c3b6aee2e8c99f0d280e8)

16 years agoSTABLE14-ptserver-vicelog-missing-newlines-20090603
Andrew Deason [Wed, 3 Jun 2009 22:44:14 +0000 (22:44 +0000)]
STABLE14-ptserver-vicelog-missing-newlines-20090603

LICENSE IPL10
FIXES 124894

some ViceLog calls in ptserver were missing newlines. that's fixed.

(cherry picked from commit 4310874d8f8b83b8a2c5840880d4f35059c67ea7)

16 years agoSTABLE14-klog-krb5-avoid-freeing-garbage-20090603
Douglas Engert [Wed, 3 Jun 2009 21:58:29 +0000 (21:58 +0000)]
STABLE14-klog-krb5-avoid-freeing-garbage-20090603

LICENSE IPL10

avoid freeing garbage due to an uninitialized variable

(cherry picked from commit edfb415915bff06c055f58867695cbb37df626e1)

16 years agoSTABLE14-aix-pag-complaints-20090603
Derrick Brashear [Wed, 3 Jun 2009 16:14:38 +0000 (16:14 +0000)]
STABLE14-aix-pag-complaints-20090603

LICENSE IPL10

make the new code work properly, thanks to aix's complaints

(cherry picked from commit ec51165b5134b406eecfbb2b7d53b2a094f30256)

16 years agoSTABLE14-curpag-via-pioctl-20090603
Derrick Brashear [Wed, 3 Jun 2009 06:41:55 +0000 (06:41 +0000)]
STABLE14-curpag-via-pioctl-20090603

LICENSE IPL10
FIXES 124709

curpag needs to know about kernel constructs (getpagvalue on AIX, onegroup
versus two group on linux) and on aix 5.1 simply can't work. add a new pioctl
and use it to simply ask the kernel what the current pag is

(cherry picked from commit 4af75fe96a9e4a985081d4af78439aeebabd6939)

16 years agoSTABLE14-linux-allow-disabling-syscall-probing-20090602
Simon Wilkinson [Tue, 2 Jun 2009 19:32:08 +0000 (19:32 +0000)]
STABLE14-linux-allow-disabling-syscall-probing-20090602

LICENSE IPL10

give a configure switch to allow disabling syscall probing

(cherry picked from commit 091139b3e2160e1f56b11820d59589de9ccd9c40)

16 years agoSTABLE14-doc-man-page-install-variables-20090602
Russ Allbery [Tue, 2 Jun 2009 18:34:06 +0000 (18:34 +0000)]
STABLE14-doc-man-page-install-variables-20090602

LICENSE BSD

We were horribly confused about variables when creating man page symlinks.
Be less confused by using variables less.

(cherry picked from commit 3ae35e12714bf4759d313086b17ef47bae1d7504)

16 years agoSTABLE14-auth-cellconfig-overflow-20090601
simon@sxw.org.uk [Tue, 2 Jun 2009 04:33:09 +0000 (04:33 +0000)]
STABLE14-auth-cellconfig-overflow-20090601

LICENSE IPL10
FIXES 124891

(cherry picked from commit e89f90f7f17efc80c984b2858cd85a87f8cdfbf3)

16 years agoSTABLE14-doc-admin-guide-updates-20090601
Jason Edgecombe [Mon, 1 Jun 2009 23:52:25 +0000 (23:52 +0000)]
STABLE14-doc-admin-guide-updates-20090601

LICENSE BSD
FIXES 124889

Updates to chapter one of the Admin Guide.  Remove references to the
Authentication Server, add references to a Kerberos server, revise ntpd
parts to reflect the fact that OpenAFS doesn't ship ntpd, and removed
the distinction between the US and non-US versions of the Update Server.

(cherry picked from commit d7b3953a530a49edfed203d59194b3a293ed80b8)

16 years agoSTABLE14-doc-xml-build-cvsignore-20090601
Russ Allbery [Mon, 1 Jun 2009 23:51:31 +0000 (23:51 +0000)]
STABLE14-doc-xml-build-cvsignore-20090601

LICENSE BSD

Ignore new generated files from the documentation build.

(cherry picked from commit e34fe0f9a85c0fa965fe30ead4c5ba14d077268f)

16 years agoSTABLE14-doc-xml-version-depends-20090601
Russ Allbery [Mon, 1 Jun 2009 23:41:04 +0000 (23:41 +0000)]
STABLE14-doc-xml-version-depends-20090601

LICENSE BSD

Make PDF manual generation also depend on version.xml, matching HTML
generation.

(cherry picked from commit 2520cc9aabef8877cdd99abb858dc011dfa5629f)

16 years agoSTABLE14-xml-versioning-take-two-20090601
Derrick Brashear [Mon, 1 Jun 2009 18:52:13 +0000 (18:52 +0000)]
STABLE14-xml-versioning-take-two-20090601

LICENSE IPL10

deal with this in a way which does not trip creation of CML state config files

(cherry picked from commit 0a617e33234f6104fdfb0d45593cf313c961f653)

16 years agoSTABLE14-vsprocs-log-format-bug-20090531
Simon Wilkinson [Mon, 1 Jun 2009 00:15:21 +0000 (00:15 +0000)]
STABLE14-vsprocs-log-format-bug-20090531

LICENSE IPL10

fix the logged data to avoid a format string issue

(cherry picked from commit 8e0b8243ea8b38e6ceab8f1dc6a0e032d6622568)

16 years agoSTABLE14-doc-doxygen-20090531
Jeffrey Altman [Sun, 31 May 2009 17:54:53 +0000 (17:54 +0000)]
STABLE14-doc-doxygen-20090531

LICENSE BSD

Doxygen converted versions of the IBM provided architecture and
protocol documentation.

Funded under an SBIR grant.

(cherry picked from commit 5526ba218f7f3653cabe5423b25b82f0241b0abc)

16 years agoSTABLE14-afsd-allow-maxmtu-override-20090530
Douglas Engert [Sun, 31 May 2009 02:23:11 +0000 (02:23 +0000)]
STABLE14-afsd-allow-maxmtu-override-20090530

LICENSE IPL10
FIXES 124880

rxi_Findcbi, rxi_FIndIfnet, rxi_FindIfMTU "failure" end up returning
the RX_REMOTE_PACKET_SIZE as the mtu to use unless we allow our override
to apply, so we do that. then, add an afsd switch to allow setting it.

afsd man page update required and will follow.

====================
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.
====================
LICENSE IPL10
FIXES 124880

man page update to document previous work

(cherry picked from commit 74df876dafb9cf80ab5a8099b4202f2205bd9791)