openafs.git
10 years agoAdd printf format checks to afs_com_err()
Simon Wilkinson [Sat, 7 Nov 2009 22:31:08 +0000]
Add printf format checks to afs_com_err()

Add gcc printf format checks to the afs_com_err() functions

Deal with the fallout, in particular change callers which pass
an empty format string to pass NULL instead - the com_err functions
already permit this alternate use.

There's a couple of real bugs here - in one case, we attempt to
print a NULL pointer, rather than a security index, and in the other
we supply a NULL format string, rather than the string we meant to
print.

Change-Id: Icd48f92a4447d4af3dba9a4caa2ff73c1657ad47
Reviewed-on: http://gerrit.openafs.org/794
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoAdd printf format checks to util's log functions
Simon Wilkinson [Sat, 7 Nov 2009 21:21:00 +0000]
Add printf format checks to util's log functions

Add gcc printf format checks to all of the logging functions
defined in src/util/afsutil.h

Deal with the fallout, in particular create cast functions to go
from VolumeId and VnodeId to (unsigned int)

Change-Id: I28181b73cfeac22cf869d34e75a8335e65125e36
Reviewed-on: http://gerrit.openafs.org/793
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoDon't pass tokens around the backup system
Simon Wilkinson [Fri, 29 Jan 2010 15:12:42 +0000]
Don't pass tokens around the backup system

The backup system has a global ktc_token, which is used
to work out when its credentials are about to expire. This leads to
an unfortunate dependency throughout the code on the format of this
token.

Replace this with a global time_t which stores the expiry time, and
copy the required field from the token into this when we get the
token. This limits the exposure of the token, and simplifies the code.

Change-Id: Ia2929c2c0a4c1ba9ca5db881865f33af5a732d2f
Reviewed-on: http://gerrit.openafs.org/1218
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoCommon interface for server security objects
Simon Wilkinson [Thu, 28 Jan 2010 23:50:54 +0000]
Common interface for server security objects

Add an interface for creating arrays of server security objects.
This collects all of the various daemon code for initialising
security objects into one place, reducing duplication and making
it easier to add new objects in the future.

Change-Id: I9df20afccdbac4a610cb5bde32f01069272f27a2
Reviewed-on: http://gerrit.openafs.org/1177
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoDocument the extra options to fs getcacheparms
Simon Wilkinson [Tue, 2 Feb 2010 22:13:45 +0000]
Document the extra options to fs getcacheparms

Write some documentation for the -files and -excessive options to
fs getcacheparms

Change-Id: I769f8c0cf6d9d100a1687ae73a337132befb2449
Reviewed-on: http://gerrit.openafs.org/1217
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoAdd a set of strings for the InstallationCheck
Simon Wilkinson [Tue, 2 Feb 2010 15:59:12 +0000]
Add a set of strings for the InstallationCheck

InstallationCheck needs a strings dictionary to turn its errors
into human readable content. Add such a dictionary, and use the
correct product name and version for each Mac OS X release.

FIXES 126322

Change-Id: I8401a1163ea078e690152215c20db6970969ee04
Reviewed-on: http://gerrit.openafs.org/1208
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoSquash pthreaded ubik warnings
Andrew Deason [Mon, 25 Jan 2010 22:56:13 +0000]
Squash pthreaded ubik warnings

Fix/ignore some warnings with --enable-pthreaded-ubik:

 - ubik/ubik.c: move rx_stackSize decl to non-pthread code, since it's
   only used there

 - budb/db_dump.c: move 'code' decl to non-pthread code, since it's only
   used there

 - ubik/recovery.c: move return to outside the ifdef, so we still have a
   return statement in the pthreaded case

 - Add -Wno-error to beacon.c in tubik, to match the ubik case

Change-Id: I614f2425c36e77bb1a08838b45166d8edb080c73
Reviewed-on: http://gerrit.openafs.org/1211
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoUse -A and -u in pthreaded ubik rxgen
Andrew Deason [Mon, 25 Jan 2010 23:05:09 +0000]
Use -A and -u in pthreaded ubik rxgen

Use -A for rxgen in tvlserver and tptserver to generate ANSI code, so we
don't generate warnings. Also use -u in tubik rxgen, so we generate the
ubik_* functions that new code uses.

Change-Id: Id13168e65e36aeebf48610c28e7b3df2164fc073
Reviewed-on: http://gerrit.openafs.org/1210
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoBuild utst_client for pthreaded ubik
Andrew Deason [Mon, 25 Jan 2010 23:08:36 +0000]
Build utst_client for pthreaded ubik

Build utst_client in src/tubik, as it is in src/ubik.

Change-Id: I655445905efbaaa9e578a9e2061091679813a768
Reviewed-on: http://gerrit.openafs.org/1209
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoDon't echo from the InstallationCheck script
Simon Wilkinson [Tue, 2 Feb 2010 15:47:02 +0000]
Don't echo from the InstallationCheck script

Return codes from InstallationCheck are turned into errors by
the GUI - don't echo them out to stdout as well.

Change-Id: Iee9c1ff4687ea42ab8c18566493a8a92b6d51fed
Reviewed-on: http://gerrit.openafs.org/1206
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoRemove internal vldbClientInit prototype
Simon Wilkinson [Fri, 29 Jan 2010 14:17:46 +0000]
Remove internal vldbClientInit prototype

vldbClientInit is used by other modules in the backup suite, and is
already prototyped in bucoord_prototypes.h. So, remove it from the
internal header.

Change-Id: I2688426189c3aeaacc8db4f815ede466caa5dde0
Reviewed-on: http://gerrit.openafs.org/1205
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoicl 64 bit platform rationalization
Derrick Brashear [Tue, 2 Feb 2010 05:36:58 +0000]
icl 64 bit platform rationalization

be consistent about how we apply 64 bit longness. right now we were
doing 3 different things. oops.

Change-Id: Ie48816852d5db916e3023ae6e3cd2f6298802668
Reviewed-on: http://gerrit.openafs.org/1200
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

10 years agoicl trace code deduplication
Derrick Brashear [Tue, 2 Feb 2010 05:16:54 +0000]
icl trace code deduplication

right now we have 2 copies of the same code for dealing with traced objects.
it's the same code! make a function, and call it 4 times.

Change-Id: I1c8b6d7b2cff59108eada1522ba1ec76efeba421
Reviewed-on: http://gerrit.openafs.org/1199
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

10 years agoFix USS to use volser prototypes
Simon Wilkinson [Fri, 29 Jan 2010 14:13:47 +0000]
Fix USS to use volser prototypes

Fix the uss_vol.c file to use the volser prototypes. This clears up the
list of the warnings in this file, so drop the warning supression from
the Makefile, and remove the entry from README.WARNINGS

Change-Id: I9e4ec8bdfa39e3a7adafea454a4c6d96c5fc9d0f
Reviewed-on: http://gerrit.openafs.org/1204
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoxdr_proc_t really is different on linux26_i386
Simon Wilkinson [Tue, 2 Feb 2010 13:14:28 +0000]
xdr_proc_t really is different on linux26_i386

i386 Linux uses a register based calling convention within the
kernel, but uses the stack for va_args based functions. This means
that the hack of prototyping xdrproc_t as a va_args function doesn't
work, as you end up with arguments being in the wrong place.

Restore the Linux only xdrproc_t prototype that
167e1aa21f5bbea1272b239dc6518a7bdbfc3ee6 removed, add a warning to
explain why its there, and modify xdr_free() so that it works
without error.

Change-Id: I789d387b01fcb892b187fe05f961f01c2c1f55e4
Reviewed-on: http://gerrit.openafs.org/1202
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoDarwin: Stop CM builds when errors occur
Simon Wilkinson [Mon, 1 Feb 2010 14:12:05 +0000]
Darwin: Stop CM builds when errors occur

Make the kernel module build on Mac OS stop when errors are
encountered, instead of just ploughing on regardless

Change-Id: I489357c04cb7c78663f593af2c2adbc50ebfffff
Reviewed-on: http://gerrit.openafs.org/1203
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoLinux: warning fix in osi_file.c
Marc Dionne [Tue, 2 Feb 2010 00:31:05 +0000]
Linux: warning fix in osi_file.c

Cast printed variable to avoid a warning about a mismatch between
%d and long int.

Change-Id: I3245cbb5e4780a258e8756b4210c1078f4fbc51c
Reviewed-on: http://gerrit.openafs.org/1195
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoLinux: Fix breakage in llseek error handling
Simon Wilkinson [Tue, 2 Feb 2010 00:03:04 +0000]
Linux: Fix breakage in llseek error handling

Commit 7a5cee30cc5f0e6d5780387633ce2b46608fd5fb changed the way
that errors from llseek are dealt with. Unfortunately, it is missing
some all important bracing, and so we end up going down the error
path, even when the llseek succeeds.

My fault. Sorry.

Change-Id: I03061ba0663b610a8fb73a08d257f6d786795076
Reviewed-on: http://gerrit.openafs.org/1194
Reviewed-by: Marc Dionne <marc.c.dionne@gmail.com>
Tested-by: Marc Dionne <marc.c.dionne@gmail.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agocall afs_osi_suser correctly in PNewUuid
Jeffrey Hutzelman [Mon, 1 Feb 2010 21:59:14 +0000]
call afs_osi_suser correctly in PNewUuid

we were using the wrong level of indirection on acred. fixed.

Change-Id: I00ab1a97b42717adc2d9d26e885b91c44103ab1d
Reviewed-on: http://gerrit.openafs.org/1190
Reviewed-by: Jeffrey Hutzelman <jhutz@cmu.edu>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agomake 1.5.71 for unix openafs-devel-1_5_71
Derrick Brashear [Mon, 1 Feb 2010 05:00:39 +0000]
make 1.5.71 for unix

version updates for unix

Change-Id: Ia7901e3969b3afb72b870e7bfaf1be931c2fb0c5
Reviewed-on: http://gerrit.openafs.org/1185
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoWindows: ChangeLog and Version Number for 1.5.71
Jeffrey Altman [Mon, 1 Feb 2010 04:27:32 +0000]
Windows: ChangeLog and Version Number for 1.5.71

Change-Id: I0969e2b3fcff98489aef1b0e660ef12149a4b96c
Reviewed-on: http://gerrit.openafs.org/1184
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoUnix CM: Always use ->storeproc()
Chas Williams [Wed, 6 Jan 2010 16:48:41 +0000]
Unix CM: Always use ->storeproc()

Always use ->storeproc() and avoid branching

"do or do not -- there is no try"

Change-Id: Idbcf0c80930d6a433b4c921b44db28cb6fbbf9bb
Reviewed-on: http://gerrit.openafs.org/1073
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agorx idledeadtime don't track window wait
Derrick Brashear [Sat, 30 Jan 2010 18:54:20 +0000]
rx idledeadtime don't track window wait

if we're in writer mode and waiting for more window, don't consider it idle.

Change-Id: Id1e80f297929eef363ffe88bfe181b74bb331515
Reviewed-on: http://gerrit.openafs.org/1182
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoDarwin: Don't use ARCHFLAGS to determine arch
Simon Wilkinson [Fri, 29 Jan 2010 17:47:12 +0000]
Darwin: Don't use ARCHFLAGS to determine arch

The Darwin build uses ARCHFLAGS to control the architectures it
builds binaries for. However, this causes problems for standalone
builds which just do ./configure && make. So, if ARCHFLAGS is
undefined, ask gcc which architecture it thinks that it is
building for, and use that.

Change-Id: If9e7dc3c41ded5ac0593ae20ddd741690914a699
Reviewed-on: http://gerrit.openafs.org/1181
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agomacos prefpane fat binary arch selection correction
Derrick Brashear [Fri, 29 Jan 2010 17:59:59 +0000]
macos prefpane fat binary arch selection correction

don't choose 64 bit arches for the prefpane if we're not on macos 10.6

sadly, this is the best i can do to conditionalize

Change-Id: I31fb2e10fbef5c66a27e0170e8925513a1b63050
Reviewed-on: http://gerrit.openafs.org/1180
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agolinux cache file open fail print error
Derrick Brashear [Thu, 28 Jan 2010 16:35:56 +0000]
linux cache file open fail print error

if we're going to oops, print a little more info about why. in this case,
knowing why dentry_open failed would help

Change-Id: I78a3c97d67c7c29110de975465cce0edd6367496
Reviewed-on: http://gerrit.openafs.org/1173
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoxdrproc_t probably not really different on linux26
Derrick Brashear [Fri, 29 Jan 2010 05:03:07 +0000]
xdrproc_t probably not really different on linux26

for some reason we prototype it differently. makes xdr_free sad.

Change-Id: Icd6060acd922314fc400aa2fe360c662ec705fa6
Reviewed-on: http://gerrit.openafs.org/1179
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agofurther irix updates
Chaz Chandler [Fri, 29 Jan 2010 04:59:12 +0000]
further irix updates

updating for non-c99 and for no vprintf in the irix kernel

Change-Id: Id7239dae35ff195007c615319522827d06395f8f
Reviewed-on: http://gerrit.openafs.org/1178
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

10 years agoafsd missing close brace
Derrick Brashear [Thu, 28 Jan 2010 23:12:44 +0000]
afsd missing close brace

fallout from the open-by-path-only unification

Change-Id: I06f6a7bbb45368b5ab8c336e93ffae11af6b3fac
Reviewed-on: http://gerrit.openafs.org/1176
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoignore generated macos packaging description file
Derrick Brashear [Thu, 28 Jan 2010 05:35:36 +0000]
ignore generated macos packaging description file

added ignore for OpenAFS.Description.plist as it is now generated

Change-Id: Icbad9f8bbbdd5be6604838366d14f648eec05ee4
Reviewed-on: http://gerrit.openafs.org/1171
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoirix cc defaults to pre-c99
Derrick Brashear [Thu, 28 Jan 2010 19:22:53 +0000]
irix cc defaults to pre-c99

no xdr labels for irix

Change-Id: I4e1ca001d62a0a3d1f2f6ddef4b6419d1b2b2c13
Reviewed-on: http://gerrit.openafs.org/1174
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoirix can't have vprintf in the kernel
Derrick Brashear [Thu, 28 Jan 2010 19:25:32 +0000]
irix can't have vprintf in the kernel

vprintf doesn't exist for irix either. deal

Change-Id: I60f78feaf36dbbb086e2ea9e6f7a839c76074647
Reviewed-on: http://gerrit.openafs.org/1175
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoWindows: drop cm_Freelance_Lock before call to cm_FreelanceAddMount
Jeffrey Altman [Tue, 26 Jan 2010 15:23:03 +0000]
Windows: drop cm_Freelance_Lock before call to cm_FreelanceAddMount

The cm_Freelance_Lock must not be held before calling
cm_FreelanceAddMount().  cm_InitLocalMountPoints() was violating
this requirement.   It is safe to drop the lock because the only
time the lock would be dropped is if the "Freelance" registry
key does not exist.  This should only be true during the initial
startup of the application the first time afsd_service.exe
executes on the machine.   At this point in time there is only
one thread that is executing that could be initializing the
Freelance mount point list.

LICENSE MIT

Change-Id: Ib49cf8bc830836c8e08446073e64ccf30086d354
Reviewed-on: http://gerrit.openafs.org/1164
Reviewed-by: Asanka Herath <asanka@secure-endpoints.com>
Tested-by: Asanka Herath <asanka@secure-endpoints.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoWindows: export xdr_serverList and xdr_Capabilities from afsrpc.dll
Jeffrey Altman [Wed, 27 Jan 2010 19:09:19 +0000]
Windows: export xdr_serverList and xdr_Capabilities from afsrpc.dll

Required by new xdr_free() usage.

Change-Id: I585947f110204c9043f92fc573d5d7263a46fd64
Reviewed-on: http://gerrit.openafs.org/1170
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoLinux: don't count pag keys against root's keyring quotas
Marc Dionne [Wed, 27 Jan 2010 00:48:24 +0000]
Linux: don't count pag keys against root's keyring quotas

Keys associated with PAGs are created with root ownership for
security reasons, which means that they count agains root's
keyring quotas.  The default configuration used by most distros
restricts root to the same quotas as a regular user, so a single
user can potentially fill up the quota and prevent new pags from
getting created system-wide.
This can also be an issue for busy multi-user systems where the
default maximum number of keys (200 currently) can easily be reached.

Change-Id: I37e3f3d881facff9b14f2057b404d4b441e72305
Change-Id: Id1ef2d5a7109d8ee7c08aa81dfd98ecf951c8ba8
Reviewed-on: http://gerrit.openafs.org/1167
Reviewed-by: Marc Dionne <marc.c.dionne@gmail.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoinit pthread mutexes only if an initializer exists
Derrick Brashear [Wed, 27 Jan 2010 17:58:49 +0000]
init pthread mutexes only if an initializer exists

and windows has none

Change-Id: Iaff3047abde197791e3fbb42c12066f896e540d1
Reviewed-on: http://gerrit.openafs.org/1169
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoswitch to always attempting cache-config-by-path
Derrick Brashear [Wed, 20 Jan 2010 16:38:43 +0000]
switch to always attempting cache-config-by-path

solve the "are we built for linux fh or inode" problem: just always use path.

Change-Id: I076ee495b29e4a2dc964ce04001f86e100528fb6
Reviewed-on: http://gerrit.openafs.org/1133
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoXDR memory management fixes
Simon Wilkinson [Wed, 27 Jan 2010 02:30:09 +0000]
XDR memory management fixes

Change the xdr_free function so that it matches in signature and
behaviour the xdr_free that is available in other xdr packages, and
update all callers.

Don't use xdr_alloc where the data we're allocating isn't freed by
xdr. The only place where xdr_alloc is required is in server stubs
when filling OUT parameters.

Change-Id: Ice231f163726afc374cb25a89302e3d82fa1e093
Reviewed-on: http://gerrit.openafs.org/1168
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoinitialize pthread mutexes to avoid compiler common variable stupidity
Derrick Brashear [Tue, 26 Jan 2010 18:32:03 +0000]
initialize pthread mutexes to avoid compiler common variable stupidity

instead of worrying about ranlib versus ranlib -c, -fno-common, etc,
just initialize the variables if an initializer exists

Change-Id: I6ed28df4d2134f84359a1bc55706e75d54e6e083
Reviewed-on: http://gerrit.openafs.org/1166
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agorx ResetCall should wait if it says it will
Derrick Brashear [Mon, 25 Jan 2010 21:49:53 +0000]
rx ResetCall should wait if it says it will

We set TQ_WAIT... and then don't bother to wait. uh.

Change-Id: Ia7d3e52c8880db53fcf26b7d79ad349720018701
Reviewed-on: http://gerrit.openafs.org/1160
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

10 years agorx avoid discarding packets while tq is busy
Derrick Brashear [Mon, 25 Jan 2010 21:02:23 +0000]
rx avoid discarding packets while tq is busy

we don't check here that no one else is working the tq *after* getting the
lock. do so.

Change-Id: I91d1ac020334b038a6a045734911335299613875
Reviewed-on: http://gerrit.openafs.org/1159
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

10 years agoviced detailed stats should record stored bytes correctly
Derrick Brashear [Tue, 26 Jan 2010 17:27:11 +0000]
viced detailed stats should record stored bytes correctly

if we get an error reading data, don't subtract it from the length
stored statistic

Change-Id: I2fec260a9f677c0747ff2c6f975745ccb1e58447
Reviewed-on: http://gerrit.openafs.org/1165
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agodarwin80 if changes cleanup
Derrick Brashear [Mon, 25 Jan 2010 17:36:25 +0000]
darwin80 if changes cleanup

fix the indirection here to be correct.
ifnet_list_get needs a pointer to a pointer for the ifnet structure
Change-Id: I81bd7c383c63cc1e312fcfd482d7162d50f6f1d3
Reviewed-on: http://gerrit.openafs.org/1158
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agodarwin rx if structure fix
Derrick Brashear [Mon, 25 Jan 2010 15:44:54 +0000]
darwin rx if structure fix

link kernel to correct ifaddr symbol for withnet

Change-Id: I384974bef73b9cc44eba7ddba43f28f65c03f39b
Reviewed-on: http://gerrit.openafs.org/1157
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agopagsh dependencies don't get to include system libs
Derrick Brashear [Sun, 24 Jan 2010 16:25:00 +0000]
pagsh dependencies don't get to include system libs

we'll never be able to "build" -lresolv

revert this part

Change-Id: I2b1caa5bffc3d826412967fb70d9eadf8223f1f8
Reviewed-on: http://gerrit.openafs.org/1154
Reviewed-by: Marc Dionne <marc.c.dionne@gmail.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoLinux: Handle llseek failure
Simon Wilkinson [Mon, 25 Jan 2010 00:02:11 +0000]
Linux: Handle llseek failure

If llseek fails, then we return immediately, potentially leaving the
wrong address space configured for this process, and without resetting
the processes saved limits.

Fix this by creating a common exit point for osi_rdwr, and using it.

Change-Id: Ifff8cb0155f92a4130d21828d5ee2d2a55c09195
Reviewed-on: http://gerrit.openafs.org/1155
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agounix 1.5.70 openafs-devel-1_5_70
Derrick Brashear [Sun, 24 Jan 2010 15:09:41 +0000]
unix 1.5.70

make 1.5.70 for unix

Change-Id: I84ed384e9cd38e49a2c002a4ffe19a6ff5ba633d
Reviewed-on: http://gerrit.openafs.org/1153
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoWindows: 1.5.70
Jeffrey Altman [Fri, 22 Jan 2010 19:24:17 +0000]
Windows: 1.5.70

ChangeLog updates

Version number updates

Change-Id: I16fd2d2ef507cc862417eef8d221ea5ba58567a1
Reviewed-on: http://gerrit.openafs.org/1146
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoLinux Keyrings: don't ignore error code from session keyring creation
Marc Dionne [Sat, 23 Jan 2010 01:03:58 +0000]
Linux Keyrings: don't ignore error code from session keyring creation

Creating a session keyring can fail, for instance if the user's
keyring quotas are reached (number of keys/keyrings or data size).

Instead of ignoring errors, return them so they can be passed back
to the caller.

FIXES 126230

Change-Id: I745abeef4b3b8e4c3ab1b90667c6a5478c1e0ad2
Change-Id: I4b0ce6423ee493fbbcbdac1c580e3a157d3c11f4
Reviewed-on: http://gerrit.openafs.org/1151
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoAdd missing dependencies for pagsh in Makefile
Marc Dionne [Sat, 23 Jan 2010 01:05:34 +0000]
Add missing dependencies for pagsh in Makefile

Make pagsh and pagsh.krb depend on the libraries that are used to
link them, and on the source file pagsh.c
Without this dependency, modifying pagsh.c would not trigger a
rebuild of the executable.

Change-Id: I9ed20ce69f10916f97d6c335be9bb8374fb9ecd1
Reviewed-on: http://gerrit.openafs.org/1150
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoDon't throw data away in afs_StoreMini
Simon Wilkinson [Fri, 22 Jan 2010 20:11:21 +0000]
Don't throw data away in afs_StoreMini

afs_StoreMini had some interesting error handling. Instead of returning
the error code from StoreData, it would return the error from EndCall,
potentially masking the StoreData error. When it encountered an error
it would discard all of the cached data.

StoreMini's only caller is afs_StoreAllSegments. If StoreAllSegments is
called from DoPartialWrite, then it squashes the error code. This
combination could lead to the user's data being disposed of, without an
error being reported.

Fix all of this by not invalidating segments in StoreMini. Make
StoreMini static to make it clear its only used by StoreAllSegments, and
fix the error handling in StoreMini so that StoreData errors always take
priority.

Change-Id: I41f0c753655fac343485d2a473ad70b6ae96bb78
Reviewed-on: http://gerrit.openafs.org/1147
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoRevert "Fix afs_AccessOK for dropbox case"
Simon Wilkinson [Fri, 22 Jan 2010 21:50:44 +0000]
Revert "Fix afs_AccessOK for dropbox case"

This reverts commit 3f89c0feae89e9a255afb8a7f08995412a3f1b79.

That change broke the expected semantics of a drop box, by
opening up read() access to users to whom it would usually be
denied. The expected dropbox behaviour is that whilst anyone
can write a file there, only those who are specifically given
r permissions may read it.

Change-Id: I025cef7cad1031fbed1e914d58a0a515561f3f78
Reviewed-on: http://gerrit.openafs.org/1149
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoabstract rx if structure access
Derrick Brashear [Thu, 14 Jan 2010 20:41:23 +0000]
abstract rx if structure access

change things to use accessor macros styled after the mac KPI
where feasible, and make access consistent.

Change-Id: I284101b370666e1aeb8e42bc1a651a2781a9af78
Reviewed-on: http://gerrit.openafs.org/1101
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoEnable weak enctypes for klog.krb5 if supported by Kerberos
Russ Allbery [Fri, 22 Jan 2010 18:59:14 +0000]
Enable weak enctypes for klog.krb5 if supported by Kerberos

The same as cb4b62a40352ccebae3a299f4327fa70fc7a0c5c, but for klog.krb5.

Current versions of both MIT Kerberos and Heimdal disable DES enctypes
by default, but DES enctypes are still required for AFS service tickets.
Probe for either krb5_allow_weak_crypto() (MIT Kerberos 1.8) or
krb5_enctype_enable() (Heimdal) and, if found, call them to enable DES
enctypes.  If neither is found, assume that the Kerberos libraries are
old enough that DES is enabled by default.

Change-Id: I99c93621c847f6edcc485207f5b6b99b2370b347
Reviewed-on: http://gerrit.openafs.org/1144
Tested-by: Russ Allbery <rra@stanford.edu>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

10 years agoFix afs_AccessOK for dropbox case
Andrew Deason [Fri, 22 Jan 2010 01:42:33 +0000]
Fix afs_AccessOK for dropbox case

afs_AccessOK did not check for if we have 'i' rights on a directory, nor
if we were the owner of a file, if we were only checking PRSFS_READ or
PRSFS_WRITE. Thus, it does not correctly address the 'dropbox'
permissions case (if you are the owner of a file and have insert
permissions, you get implicit 'r' and 'w').

Fix this to check for 'i' and ownership when we need to, so the dropbox
special-case check works as intended.

FIXES 126216

Change-Id: Ib629bc30b73bc3965e777747067259451f0adff6
Reviewed-on: http://gerrit.openafs.org/1143
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoEnable weak enctypes for aklog if supported by Kerberos
Russ Allbery [Thu, 21 Jan 2010 23:41:49 +0000]
Enable weak enctypes for aklog if supported by Kerberos

Current versions of both MIT Kerberos and Heimdal disable DES enctypes
by default, but DES enctypes are still required for AFS service tickets.
Probe for either krb5_allow_weak_crypto() (MIT Kerberos 1.8) or
krb5_enctype_enable() (Heimdal) and, if found, call them to enable DES
enctypes.  If neither is found, assume that the Kerberos libraries are
old enough that DES is enabled by default.

Change-Id: Ib5231bb7c2fe88f4c424628394ed08122d5710a1
Reviewed-on: http://gerrit.openafs.org/1141
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoafs_vnop_attrs: syntax error in conditional expression
Marc Dionne [Thu, 21 Jan 2010 22:53:42 +0000]
afs_vnop_attrs: syntax error in conditional expression

Add a missing && in a 2-line condition.
Not sure how this could have compiled successfully on the affected
platforms.

Change-Id: I2e8b791c6d8cd09795be1ef52c078ebfa5a3e53b
Reviewed-on: http://gerrit.openafs.org/1140
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoWindows: AFS_AFSDB_ENV preprocessor symbol no longer exists
Jeffrey Altman [Fri, 22 Jan 2010 01:05:25 +0000]
Windows: AFS_AFSDB_ENV preprocessor symbol no longer exists

The decision to add AFS_AFSDB_ENV to make code consistent was
a mistake.  The preprocessor symbol no longer exists.  AFSDB
is broken in 1.5.69.

Change-Id: I7ea147b824f3a67e2039abee8b2513af175fcc5d
Reviewed-on: http://gerrit.openafs.org/1142
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoWindows: add build and clean rule for bc.h
Jeffrey Altman [Thu, 21 Jan 2010 22:38:43 +0000]
Windows: add build and clean rule for bc.h

Add a dependency rule to build bc.h before it is installed

Add cleanup rules to delete bc.h and other installed files.

Change-Id: I9a40fd64aff048bfc87b9fd823ae7499816cb3e5
Reviewed-on: http://gerrit.openafs.org/1139
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoWindows: Windows v6.0 SDK does not define __RPC__out
Jeffrey Altman [Thu, 21 Jan 2010 21:43:18 +0000]
Windows: Windows v6.0 SDK does not define __RPC__out

The Windows v6.0 SDK does not define __RPC__out in the rpcsal.h
header file.  The preprocessor symbol was added in the v6.0A SDK
update.  In case the v6.0 SDK is used with the 6000 WDK, add a
symbol declaration.

Change-Id: Iecce566e906977f41be00efb458753acd09f7949
Reviewed-on: http://gerrit.openafs.org/1138
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoUndo labeling of elements of rx structures for Windows
Jeffrey Altman [Thu, 21 Jan 2010 20:34:49 +0000]
Undo labeling of elements of rx structures for Windows

Change I006bbbdb8923dbb72a97fde931a1b23e303375c2 broke
the Windows build.  Restore the unlabled assignments for
Windows since Windows always uses the AFS xdr implementation.

Change-Id: I2eadc624d84f4100281424993b7287a330d6c753
Reviewed-on: http://gerrit.openafs.org/1136
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoInstall bucoord/bc.h to permit Windows to build
Jeffrey Altman [Thu, 21 Jan 2010 20:30:23 +0000]
Install bucoord/bc.h to permit Windows to build

Change  I0a0ae92d3c7ba75e4a331bf59ea9c031763cc704
broke the Windows build.   Fix it.

Change-Id: I08db225790a37eee3e4fd778eacabc7168abee77
Reviewed-on: http://gerrit.openafs.org/1135
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoWarning cleanup: implicit function declarations in butc and bucoord
Marc Dionne [Thu, 21 Jan 2010 02:03:08 +0000]
Warning cleanup: implicit function declarations in butc and bucoord

Cleanup some prototype related warnings under bucoord and butc:
- Copy a few prototypes from bucoord_internal.h to bucoord_prototypes.h,
since they are used in other source directories
- Install bc.h under include/afs, and include it in tcmain.c
- Include bucoord_prototypes.h in tcmain.c
- Add a prototype for GetResponseKey
- Remove warning inhibition for tcmain.c, and entry in README.WARNINGS
- Remove warning inhibition for bucoord/dump.c (no entry in README)

Change-Id: I0a0ae92d3c7ba75e4a331bf59ea9c031763cc704
Reviewed-on: http://gerrit.openafs.org/1134
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agosimplify ufs cache mechanisms available
Derrick Brashear [Wed, 20 Jan 2010 16:15:29 +0000]
simplify ufs cache mechanisms available

instead of having use_fh for linux, vnode_path for solaris and darwin,
use the same mechanism for both. for darwin/solaris we cache the path
(tested via gop_lookupname) instead of an fh, freeing the path at
shutdown.

Change-Id: Id22052db0803264bdaba6b870d9e511953e348bc
Reviewed-on: http://gerrit.openafs.org/1132
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

10 years agobatch panic decoding for MacOS
Derrick Brashear [Tue, 19 Jan 2010 06:29:11 +0000]
batch panic decoding for MacOS

add support for decoding (nominally) any panic for MacOS. Limited testing.
Requires hdutil from http://www.dementia.org/~shadow/dmgutil-0.1.tar.gz
to extract files from DMGs.

now with support for cross-version debugging, when run on 10.6 (e.g. with kextutil)
Change-Id: I5d9db005e3014e22f916070f8af25271a28615ea
Reviewed-on: http://gerrit.openafs.org/1125
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agomacos set don't backup attribute on cache files
Derrick Brashear [Tue, 19 Jan 2010 00:25:02 +0000]
macos set don't backup attribute on cache files

hint to time machine that cache files should not be backed up

Change-Id: I098c9231139b91ff093be74d8a39f94c185d478c
Reviewed-on: http://gerrit.openafs.org/1122
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoFix so that UKERNEL can keep using system xdr
Simon Wilkinson [Wed, 20 Jan 2010 13:13:58 +0000]
Fix so that UKERNEL can keep using system xdr

This fix modifies the way that the RX XDR operations are declared
so that UKERNEL can continue using the system provided XDR glue,
rather than using our own (as is the case in userspace, and in
normal kernel modules)

Change-Id: Ib6abad376c7f7a7f0f5c2314efe80f5b7b44f842
Reviewed-on: http://gerrit.openafs.org/1131
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoLabel elements of rx structures
Simon Wilkinson [Sun, 17 Jan 2010 18:34:03 +0000]
Label elements of rx structures

Label all of the assignments to the xdr_ops structure. This allows
the rec, stdio and mem variants to continue to function, even when
the ordering of the elements has changed in the header.

Change-Id: I006bbbdb8923dbb72a97fde931a1b23e303375c2
Reviewed-on: http://gerrit.openafs.org/1126
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoAvoid executable headers in local include tree
Russ Allbery [Wed, 20 Jan 2010 01:43:35 +0000]
Avoid executable headers in local include tree

sys_prototypes.h, afs_AdminErrors.h, and afs_Admin.h were being installed
into include/ in the build tree via ${INSTALL} instead of ${INSTALL_DATA}.
rx_prototypes.h was executable in the repository.  Install header files
non-executable and make rx_prototypes.h non-executable for cleanliness.

Discovered by Lintian because the executable permissions were duplicated
by make libafs_tree.

Change-Id: I609b62f69d1fecc2cdbba7e7442d93e9675c2fa0
Reviewed-on: http://gerrit.openafs.org/1129
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoWindows: Help text update for Network Identity Manager Provider openafs-devel-1_5_69
Jeffrey Altman [Tue, 19 Jan 2010 03:23:54 +0000]
Windows: Help text update for Network Identity Manager Provider

Update the NetIdMgr provider help text to better describe its
functionality.  (Or at least not be incorrect.)

Change-Id: I5d6b18f15cab7c48751666e754223ee68fb18d11
Reviewed-on: http://gerrit.openafs.org/1124
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoWindows: Add a context menu for the AFS icon
Asanka Herath [Mon, 18 Jan 2010 21:47:31 +0000]
Windows: Add a context menu for the AFS icon

The OpenAFS plug-in for Network Identity Manager displays an icon
in the notification area that displays status information about
the AFS service and tokens.  Add a context menu to the icon so
that users can open the NIM application and get help for the
plug-in from the notification icon.

Change-Id: I704934b4f4eab72a18c54ef2a7a7dc3ca248ba78
Reviewed-on: http://gerrit.openafs.org/1121
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoWindows: Version number update for 1.5.69
Jeffrey Altman [Sat, 16 Jan 2010 16:07:23 +0000]
Windows: Version number update for 1.5.69

Change-Id: I8af0167649cfeec95e5a5e5a72faeaccdeb92102
Reviewed-on: http://gerrit.openafs.org/1115
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoWindows: ChangeLog for 1.5.69
Jeffrey Altman [Sat, 16 Jan 2010 16:05:57 +0000]
Windows: ChangeLog for 1.5.69

Change-Id: I078f719c6e769cfcb7997bb8b67403de480255f9
Reviewed-on: http://gerrit.openafs.org/1114
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agomake 1.5.69 for unix
Derrick Brashear [Tue, 19 Jan 2010 00:59:28 +0000]
make 1.5.69 for unix

set version number for unix

Change-Id: I2273e97ec43fd172e6f486ab6226cc6376885163
Reviewed-on: http://gerrit.openafs.org/1123
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agocreate debugging kext package for MacOS
Derrick Brashear [Sun, 17 Jan 2010 06:10:28 +0000]
create debugging kext package for MacOS

this creates and installs a debugging kext package for macos. it also
always installs the decode-panic script (which can be used even without
the debug kext)

Change-Id: Iff03de66cd3df2690f03333e6629d21660364cd1
Reviewed-on: http://gerrit.openafs.org/1120
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoMore pam warning cleanups - test_pam.c
Marc Dionne [Sat, 16 Jan 2010 22:40:52 +0000]
More pam warning cleanups - test_pam.c

Clear up a few more warnings in the pam code.  The changes here
are very similar to the ones in the previous patch and use
the same configure test:
- use PAM_CONST to conditionally declare pam_message as const
- cast a few arguments to putenv, which expects a non-const pointer

Change-Id: I6c98623c35f4453f34c1d48b8b7d6ff1bfbc1e0c
Reviewed-on: http://gerrit.openafs.org/1116
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoup.c: remove unused variable pageSize
Marc Dionne [Sat, 16 Jan 2010 21:29:10 +0000]
up.c: remove unused variable pageSize

Variable pageSize is declared and set to the system page size
using getpagesize(), but the value is not used anywhere.

getpagesize() doesn't necessarily have a prototype declared in
unistd.h with recent glibc, which can generate a new warning and
trip up an enabled-checking build.

Instead of trying to put in a workaround, just remove the unused variable.

Change-Id: I6a9519eb2642bb0479edd1cf1195ce3f751c72bb
Reviewed-on: http://gerrit.openafs.org/1117
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agouss: Fix IP address parsing and cleanup warnings
Marc Dionne [Sat, 16 Jan 2010 23:57:24 +0000]
uss: Fix IP address parsing and cleanup warnings

In uss_vol_GetServer(), the IP address octets are scanned into "char"
variables, and later cast to afs_int32 to build the address.
sscanf gives warnings because it's expecting an int pointer, and
testing shows that this code doesn't work as expected - the first
3 octets are always parsed as 0.

Use afs_int32 variables instead, which works, eliminates warnings and
simplifies the code.

Note that this code does not seem to be reachable currently.
It was probably meant to be used by planned additional uss commands.

Change-Id: I646d4cbfa8ac0d0c50f98a334ac3fe387d6d361b
Reviewed-on: http://gerrit.openafs.org/1119
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoWindows: Query the MTU value from the registry
Asanka Herath [Fri, 15 Jan 2010 20:39:12 +0000]
Windows: Query the MTU value from the registry

The syscfg_GetIFInfo() call uses the interface info returned by
GetAdaptersAddresses().  Also manually query the MTU value in the
registry for any interfaces bound to the same adapter and use the
smallest such MTU as the value returned by syscfg_GetIFInfo().

Change-Id: I8a1d487b51f6674ff6ebbf6b1f20d37b1e278279
Reviewed-on: http://gerrit.openafs.org/1105
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoDon't double free call structure
Simon Wilkinson [Sat, 16 Jan 2010 15:35:34 +0000]
Don't double free call structure

If the rx_Read() of the number of bytes in the FetchData64 response
fails, then it sets code, and disposes of the call structure. However,
the length safety check that was added in
c7b92a3018044f7aca4d9a77644e5c06ef64d1e9 executes regardless of whether
code is set, and the call has already been freed. So we end up calling
rx_Error with a NULL call structure, and panic.

Change-Id: Ia2e341b7a9a2ddc1d656e8b8a31698c0d1771d5e
Reviewed-on: http://gerrit.openafs.org/1112
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoLinux: Add missing prototype
Simon Wilkinson [Sat, 16 Jan 2010 02:15:36 +0000]
Linux: Add missing prototype

6e65ffbb6c520bf1c28356206410363d8eb36d78 added osi_get_group_pag.
On most platforms this is static, but on Linux it's provided by
the OSI layer. Add a prototype so the compiler doesn't go boom.

Change-Id: I106fa5d5299f8ed486e70cea2ea8681b43a81a33
Reviewed-on: http://gerrit.openafs.org/1110
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoUnix CM: Fix negative file length case, again
Simon Wilkinson [Sat, 16 Jan 2010 02:09:34 +0000]
Unix CM: Fix negative file length case, again

As originally noted in f6f9ee5402f1718f330a00ec89fb34b05c3cd360
some fileservers return a negative length, typically when a client
is attempting to fetch data that is past the extents of the file,
the CM needs to retain this negative length, and handle it correctly.

c7b92a3018044f7aca4d9a77644e5c06ef64d1e9 added safety checks for
the fileserver returning a length larger than that asked for by
the client. Sadly, this check does a comparison between a
signed, and an unsigned, variable. This leads to it incorrectly
classifying negative responses as being too large.

Add a cast, and a comment, to fix this.

Change-Id: I2ca67f55204c565667d5cd91cde3d520f8d9b10c
Reviewed-on: http://gerrit.openafs.org/1109
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoRx: Correct AFS_NT40_ENV rx_GetIFInfo max MTU assignments
Jeffrey Altman [Fri, 15 Jan 2010 14:06:05 +0000]
Rx: Correct AFS_NT40_ENV rx_GetIFInfo max MTU assignments

On UNIX, the rx library values for rx_maxReceiveSize and
rx_MyMaxSendSize are sent by the cache manager directly.
In Windows, they are set by rx_GetIFInfo() which had two
errors.

(1) The comparison of rx_maxReceiveSize and maxsize were
    reversed which prevented rx_maxReceiveSize from ever
    being set to the interface MTU.

(2) rx_MyMaxSendSize was never assigned a value.

As a result, two problems occurred.

(1) The remote peer was never told about the local MTU.

(2) The local peer ignores the MTU.

From 1.3.60 to 1.5.33, OpenAFS for Windows installers
provided a registry default RxMaxMTU of 1260.  This caused
the cache manager to call rx_SetMaxMTU() which in turn set
both rx_maxReceiveSize and rx_MyMaxSendSize in effect
masking these errors.

Change-Id: Ib05927d7985052e233ff6f4bd170a939eb05c320
Reviewed-on: http://gerrit.openafs.org/1107
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoRx: Do not drop call lock in rx_WriteProc* and rx_ReadProc*
Jeffrey Altman [Fri, 15 Jan 2010 14:18:50 +0000]
Rx: Do not drop call lock in rx_WriteProc* and rx_ReadProc*

rx_WriteProc and rx_ReadProc has special fast logic that
handles the most frequent case.  This code was called
without obtaining the call lock. However, each of these functions
must obtain the call lock for the queue_IsNotEmpty() test and
must re-obtain the call lock if the rxi_XXX variant is required.
Dropping the lock and re-obtaining it is more expensive than
holding it across the memcpy.  Therefore, we shouldn't drop the
lock until we are done.

Change-Id: Icca679567994e91bbbf3afec72b863d986f86582
Reviewed-on: http://gerrit.openafs.org/1108
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Tested-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoRx: Remove last remnants of DJGPP support
Jeffrey Altman [Fri, 15 Jan 2010 14:00:41 +0000]
Rx: Remove last remnants of DJGPP support

DJGPP code just clutters the AFS_NT40_ENV specific code
but is never built.

Change-Id: I90192bb5cf35239fdbbeaa28f85d1381162f3bae
Reviewed-on: http://gerrit.openafs.org/1106
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agosrc/pam warning fixes
Marc Dionne [Thu, 14 Jan 2010 23:56:49 +0000]
src/pam warning fixes

The second parameter to pam_conv() should be a const pointer on
recent systems.  Make it so to eliminate a couple of warnings.
A configure test is added to deal with some systems where pam_conv()
might not be const.

Cast a few assignments to cell_ptr in afs_auth.c and afs_setcred.c
since the argv parameter is const.

Change-Id: I5757310c94a6f26ca7dab656edaa416d16e32e2a
Reviewed-on: http://gerrit.openafs.org/847
Tested-by: Marc Dionne <marc.c.dionne@gmail.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoFix typo in AdminGuide
Asanka Herath [Fri, 15 Jan 2010 00:15:03 +0000]
Fix typo in AdminGuide

</pare> -> </para>

Change-Id: If0f8183f59f038b1b5e033d5ebd2607b5091f6da
Reviewed-on: http://gerrit.openafs.org/1103
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>

10 years agovolmonitor keep vtrans lock
Dan Hyde [Wed, 13 Jan 2010 19:38:47 +0000]
volmonitor keep vtrans lock

VolMonitor must maintain the VTRANS lock the whole time it is walking
the allTrans list.  Failure to do so can cause core dumps trying to
access memory that has already been free'd.

Two versions of this change were coded and tested.  The other version
used the VTRANS lock only around a THOLD, but needed a TRELE, too.
Timing tests were run counting the number of vos status, vos listvol,
and vos backupsys operations that could be performed during a fixed
number of vos status operations.  The THOLD/TRELE version caused other
vos operations to run about 5% slower.

FIXES 126110

Change-Id: I7e749d30c955867faacafa978d7d643dee648ca3
Reviewed-on: http://gerrit.openafs.org/1098
Reviewed-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoKernel is always defined
Simon Wilkinson [Tue, 17 Nov 2009 20:46:17 +0000]
Kernel is always defined

When we build the kernel module, either in the real or the ukernel
case, KERNEL is defined. So, remove the #ifdefs around our locking
primitives, so that things are a little bit clearer.

Change-Id: I60893cc630d8cd4dd6cb11cdc44e8bf4f2f9f863
Reviewed-on: http://gerrit.openafs.org/838
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoMove GLOCK initialisation to platform directories
Simon Wilkinson [Tue, 17 Nov 2009 19:52:15 +0000]
Move GLOCK initialisation to platform directories

Rework the GLOCK initialisation code so that it's moved out into
platform directories, rather than all being done in osi_Init.

Change-Id: I1aae76ba12cd4e45f54881f5573ed1713159b64b
Reviewed-on: http://gerrit.openafs.org/837
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agorewrite afs_MemWriteBlk() using afs_MemWritevBlk()
Chas Williams [Tue, 5 Jan 2010 00:22:19 +0000]
rewrite afs_MemWriteBlk() using afs_MemWritevBlk()

this avoids the code duplication of extending/shrinking cache entries

Change-Id: Ieeae668caa506e934f92edf193b99dabdba2eabe
Reviewed-on: http://gerrit.openafs.org/1062
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoQuick fix for readpages when using memcache
Chas Williams [Tue, 5 Jan 2010 00:18:36 +0000]
Quick fix for readpages when using memcache

Just don't do readpages if we're using memcache

Change-Id: I6b8b45f291d106ad1303d5bea1caf7f910b7ecf4
Reviewed-on: http://gerrit.openafs.org/1061
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoafsd: decprecate -mem_alloc_sleep flag (and related code)
Chas Williams - CONTRACTOR [Thu, 7 Jan 2010 23:07:43 +0000]
afsd: decprecate -mem_alloc_sleep flag (and related code)

On most platforms, afs_osi_Alloc_NoSleep() is already afs_osi_Alloc().
I have tested this on Solaris (the major exception) and haven't seen
any problems.

Change-Id: I0e8bfd76b2bd25889c1491633a3a6d50aceac510
Reviewed-on: http://gerrit.openafs.org/1079
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoAlways check VInitVolumePackage2 return code
Andrew Deason [Mon, 11 Jan 2010 19:36:32 +0000]
Always check VInitVolumePackage2 return code

Some programs were not checking the return code of VInitVolumePackage2.
Although some programs do not care so much if we fail to properly init
the volume package (SYNC debug tools), at the very least log/print an
error, so it doesn't silently fail.

Other programs are changed to exit when VInitVolumePackage2 fail, so
e.g. salvages don't accidentally cause corruption on 'logging'-mounted
UFS partitions.

In any case, ensure the return code is always checked anywhere it is
called.

Change-Id: I164d5920a5ea2dd7e5ed9ad4ccc578e9bdf0db0b
Reviewed-on: http://gerrit.openafs.org/1090
Tested-by: Andrew Deason <adeason@sinenomine.net>
Reviewed-by: Derrick Brashear <shadow@dementia.org>

10 years agoRemove weekly bosserver restarts
Simon Wilkinson [Wed, 13 Jan 2010 17:28:24 +0000]
Remove weekly bosserver restarts

Change the default so new installations of the bosserver have
no weekly restarts. Update the manpage and XML documentation to
reflect this change.

FIXES 126138

Change-Id: Ic22b750a602f6d2a22be881f5e1b04cd4fa132ae
Reviewed-on: http://gerrit.openafs.org/1097
Reviewed-by: Russ Allbery <rra@stanford.edu>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoLinux: Rework PAG to group conversions
Simon Wilkinson [Sat, 19 Dec 2009 23:58:54 +0000]
Linux: Rework PAG to group conversions

The change to using a single group for PAGs on Linux led to a load
of special cases throughout the code. This patch moves the OS
dependent parts of this into the LINUX/ platform directory, and
rearranges that file so that there is a clear distinction between
the one and two group sections of the code.

Change-Id: Ib6e6f081519b23494dcf731b4be689c006180c68
Reviewed-on: http://gerrit.openafs.org/1007
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>

10 years agoWindows: remove signed vs unsigned error in smb_Init
Jeffrey Altman [Mon, 11 Jan 2010 19:35:12 +0000]
Windows: remove signed vs unsigned error in smb_Init

Use UINT_PTR instead of INT_PTR.

LICENSE MIT

Change-Id: Id10cd7b7a18688c9eee68e3eac9c20348519fd37
Reviewed-on: http://gerrit.openafs.org/1089
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoWindows: Protect buffers in smb_WriteData from simultaneous writes
Jeffrey Altman [Mon, 11 Jan 2010 19:21:11 +0000]
Windows: Protect buffers in smb_WriteData from simultaneous writes

smb_WriteData does not properly use CM_SCACHESYNC_WRITE to
protect buffers from simultaneous writes.  Instead of simply
testing CM_SCACHESYNC_WRITE at the top of the while loop,
the flag must remain set until the entire write completes.
cm_SyncOp is now called once and cm_SyncOpDone is only called
upon final success or error.

In addition, as 'count' is unsigned, the test for count < 0
is replaced with count != 0.

LICENSE MIT

Change-Id: I82c8dc20e62079b13bf305e906f4744756aa0ac2
Reviewed-on: http://gerrit.openafs.org/1087
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoWindows: fs examine owner and group are signed
Jeffrey Altman [Mon, 21 Dec 2009 17:35:12 +0000]
Windows: fs examine owner and group are signed

Change the owner array to be signed for displaying owner
and group within fs examine output.

LICENSE MIT

Change-Id: I807e3f1b3c07eb26c1782d6b76cd49977467ce33
Reviewed-on: http://gerrit.openafs.org/1017
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>

10 years agoWindows: do not call cm_SearchCellByDNS if AFS_AFSDB_ENV is not defined
Jeffrey Altman [Mon, 11 Jan 2010 18:59:12 +0000]
Windows: do not call cm_SearchCellByDNS if AFS_AFSDB_ENV is not defined

cm_SearchCellByDNS is only available when AFS_AFSDB_ENV is
defined.  Do not call it when it isn't.

LICENSE MIT

Change-Id: I203ab775513bd75ff6a547c557579c0bc095d555
Reviewed-on: http://gerrit.openafs.org/1086
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Asanka Herath <asanka@secure-endpoints.com>
Tested-by: Asanka Herath <asanka@secure-endpoints.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>