when 37fc2dfbb9 was merged, the close
bracket went in the wrong place, meaning for non-remote calls
we don't re-enable keepalives. fix it.
Change-Id: I163db9b3b2f42f629d8a5ff22a75902a61a3138b
Reviewed-on: http://gerrit.openafs.org/7548
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Reviewed-by: Simon Wilkinson <simonxwilkinson@gmail.com>
Reviewed-by: Marc Dionne <marc.c.dionne@gmail.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
Since commit 5fc2365f, a VNOVOL error early in attach2 meant that we
skipped changing the volume state to anything, and just returned
instead. When we do this, the volume is in VOL_STATE_ATTACHING for
DAFS, and so if we return, the volume will forever be in
VOL_STATE_ATTACHING. The next thing that tries to access the volume
will wait forever for the volume to come out of that state.
So, revert half of 5fc2365f, and transition to ERROR state instead.
This code path should not be hit during normal usage, since a
nonexistant volume access for the fileserver will be detected earlier.
If the volume does not appear to exist at this stage of attachment,
something is wrong with the volume, so this warrants the ERROR state.
For the volserver and other volume utilities, we may hit this when a
request just plain references a nonexistant volume for whatever
reason, but in that case the vp should go away soon. For non-DAFS,
this commit does not change much, since the difference between
error_notbroken and unlocked_error is very small.
The other half of 5fc2365f is not changed, since it is correct. For
VOFFLINE errors at this point, the volume has already been
transitioned to VOL_STATE_UNATTACHED, so it is okay to return. Add a
comment to help make this more explicit.
Change-Id: I63853f6e1721020cd204639c6e3ce4b59de08f75
Reviewed-on: http://gerrit.openafs.org/7516
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
FSYNC_Drop currently attaches volumes that were checked out by the
dropped fssync handler, but not checked back in, in order to make the
volume available again. For DAFS, however, a full attachment is
unnecessary; just preattach instead.
Change-Id: I2c9d0b8cf0ddc7f4182347706b2c26d86f69725e
Reviewed-on: http://gerrit.openafs.org/7515
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
Normally, volume locks acquired by an exiting salvaging process would
be automatically given up when the process exits, since our FDs are
closed. However, if we exit by calling Exit() or Abort(), we
gracefully shutdown our SYNC channels before exiting. For FSSYNC, this
can result in the fileserver trying to online the volumes we had
checked out but had not yet checked back in, so the fileserver may try
to online a volume we have locked, before the locks have been
released.
To avoid this, unlock all volume locks for all partitions before we
shutdown SYNC channels on exit.
Change-Id: Idcda5fcf4639da7d4b951bea36606ea8daf68f3c
Reviewed-on: http://gerrit.openafs.org/7514
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
Tested-by: Derrick Brashear <shadow@dementix.org>
reuse the disk cache dcache creation code for memcache, then
call it to prime the list
Change-Id: If63268aa1a89c9642236f944e1ac8328d68aeba5
Reviewed-on: http://gerrit.openafs.org/7446
Reviewed-by: Chas Williams - CONTRACTOR <chas@cmf.nrl.navy.mil>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
Tested-by: Derrick Brashear <shadow@dementix.org>
we try to free the string if there was a value, so
if there's none we should probably ensure we know it
Change-Id: I20fc535e56d65c82f26852ba59528d5660c98e20
Reviewed-on: http://gerrit.openafs.org/7546
Reviewed-by: Derrick Brashear <shadow@dementix.org>
Tested-by: Derrick Brashear <shadow@dementix.org>
stealing the krb5 parser and linking with static heimdal can cause problems.
remap things we have which duplicate heimdal's private namespace.
Change-Id: I89030967a32997f00a21e97258056c26f2a4cbf6
Reviewed-on: http://gerrit.openafs.org/7545
Reviewed-by: Derrick Brashear <shadow@dementix.org>
Tested-by: Derrick Brashear <shadow@dementix.org>
The encode_fh export operation now expects two inode arguments
instead of a dentry and a "connectable" flag. Use the inode of
the dentry we're interested in, and NULL as the parent inode which
is the same as passing a 0 flag in the previous API.
Change-Id: I05cf146fb2a4bacdca20a9f108d04ccb11530804
Reviewed-on: http://gerrit.openafs.org/7523
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
With a recent glibc update, sys/wait.h no longer includes
sys/resource.h unless __USE_SVID, __USE_XOPEN or __USE_XOPEN2K8
are set.
Don't rely on the indirect inclusion to get the bits we need;
include it directly in afsd_kernel.c. This include used to be
there but was dropped when afsd_kernel.c was split off.
Change-Id: I1bfd2e0c340e15ca44472aa89614f088e0c0ce0c
Reviewed-on: http://gerrit.openafs.org/7522
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
Protect against NULL FileObject or ObjectInformation.
Log additional information when an exception is caught in
AFSQueryVolumeInfo.
Change-Id: I6111345e53edc641f4d809839c1666f7a11a93d3
Reviewed-on: http://gerrit.openafs.org/7525
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Tested-by: Jeffrey Altman <jaltman@secure-endpoints.com>
commit 56a2cbb5fb added a
GetFileAttributes query to the pioctl interface to protect
against printing pages on printers with a broken smb protocol
implementation. This patchset adds support for GetFileAttribute
on _._AFS_IOCTL_._ files to the SMB implementation.
Change-Id: I9d60dc77b7048892eb816e9d1c6e94033711ae7b
Reviewed-on: http://gerrit.openafs.org/7524
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Tested-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Consolidate all operations necessary to free an extent in
a single function, AFSFreeExtent().
Change-Id: Id0edd2009136c49cbafb9374e373749d6a906d40
Reviewed-on: http://gerrit.openafs.org/7521
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Tested-by: Jeffrey Altman <jaltman@secure-endpoints.com>
All extents were flushed whenever AFSReleaseExtentsWithFlush was
executed. This included a call at the completion of each
NonCached Read operation which could result in heavy thrashing
as the data would be released prior to it being needed by the
application.
This patchset makes the following adjustments. First,
AFSReleaseExtentsWithFlush() has been modified to release all
but 1024 extents belonging to the file. Second, NonCached Reads
only execute AFSReleaseExtentsWithFlush() when there are more
than 4096 extents associated with the file. Third,
AFSReleaseExtentsWithFlush() now has a 'bReleaseAll' parameter
which is used for calls from AFSCleanup() and AFSFlushExtents()
which need to be able to flush all extents attached to a FCB.
Change-Id: Id8b05f02c59eb46b1881e4d905a511a2597455e8
Reviewed-on: http://gerrit.openafs.org/7520
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Tested-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Before freeing the memory for an Extent object, the extent must
be removed from all of the extent skip lists. Otherwise, the
lists will be corrupted.
Change-Id: I891321477b9f0e194544e9e73dad53473f0865b4
Reviewed-on: http://gerrit.openafs.org/7518
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Tested-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Both AFSExtentsSupport.cpp and AFSIoSupport.cpp provided
NextExtent() implementations. The AFSExtentsSupport version was
more general. Use that one for both modules.
Change-Id: I19d17c989b4eadd40dcbb3d6d4186c316c26577e
Reviewed-on: http://gerrit.openafs.org/7517
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Tested-by: Jeffrey Altman <jaltman@secure-endpoints.com>
In cm_RecycleSCache cm_GetVolumeByFID can fail leaving 'volp'
a NULL pointer. Do not call cm_PutVolume() if cm_GetVolumeByFID
fails.
Change-Id: I57eaf5d40130ecad75a312ef5487cdfa6e7367e3
Reviewed-on: http://gerrit.openafs.org/7506
Reviewed-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Tested-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Add a NonPaged ERESOURCE object to the AFS_CCB structure for
use in synchronizing access to the Ccb->Flags field and the
construction of Directory Snapshots.
Change-Id: I10ecc95619d0e5bd1ca02b044c2f1541a1b9a9f0
Reviewed-on: http://gerrit.openafs.org/7478
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Tested-by: Jeffrey Altman <jaltman@secure-endpoints.com>
AFSPrimaryVolumeWorkerThread() holds VolumeCB->VolumeLock SHARED
across the call to AFSCleanupFcb() -> CcPurgeCacheSection().
If a filter driver such as Sophos (savonaccessfilter.sys ->
savonaccesscontrol.sys) triggers an AFSCreate() in response to
the cache section being purged that will force the evaluation of
the file path by AFSLocateNameEntry(). If the path contains a
mount point that requires validation, AFSBuildMountPointTarget()
is called which in turn required the VolumeCB->VolumeLock EXCL.
AFSBuildMountPointTarget() only requires the VolumeCB->VolumeLock
if the VolumeCB->RootFcb == NULL. That should only be true if
the VolumeCB was allocated by AFSInitVolume() or under very rare
race conditions.
This patchset refactors AFSInitVolume() to ensure that it holds
an extra VolumeCB->VolumeReferenceCount reference. This reference
is used to assist in the refactoring of AFSBuildRootVolume() and
AFSBuildMountPointTarget() to avoid races with volume root object
invalidation as well as permitting the VolumeCB->VolumeLock to
be ignored in the common case.
Avoiding the acquisition of VolumeCB->VolumeLock during mount
point target evaluation has the additional benefit of reducing
lock contention during path evaluation.
FIXES 130812
Change-Id: Id9b0dcc2bfd91277d522f3724893b60ce4d947f5
Reviewed-on: http://gerrit.openafs.org/7474
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Peter Scott <pscott@kerneldrivers.com>
Reviewed-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Tested-by: Jeffrey Altman <jaltman@secure-endpoints.com>
Commit 9efc255a makes the numBuffBytes variable unnecessary.
Remove it to clear up an "unused but set" warning.
Change-Id: Ifc50c40542e008d76007c76a9ce146fe6ebab4be
Reviewed-on: http://gerrit.openafs.org/7504
Reviewed-by: Derrick Brashear <shadow@dementix.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
end_writeback() is renamed to clear_inode(). Add a configure test
and cope.
Change-Id: Icaf5b6b54d0ee377fabcf0b295d690eaa6b4be5e
Reviewed-on: http://gerrit.openafs.org/7503
Reviewed-by: Derrick Brashear <shadow@dementix.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
In afs_FlushVCache, when we QueueVCB, we might drop the afs_xvcache
lock (as of 76158df491). The vcache may
still be on the DNLC, so a lookup while xvcache is dropped can cause
someone else to grab a reference to the vcache while it is being
flushed. This can cause panics and failed assertions, since someone
will have a reference to the flushed vcache, which is effectively
freed and many of the structure fields are no longer valid.
So instead, do not call QueueVCB until we have purged the vcache from
the DNLC.
Change-Id: Ia6b51b4a2bb26903346217fbd6fc7d959d87eae7
Reviewed-on: http://gerrit.openafs.org/7430
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it is
unnecessary noise.
Change-Id: Idc5f098a20809a5f869c4ea9f3970e598520c0e9
Reviewed-on: http://gerrit.openafs.org/7468
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it is
unnecessary noise.
Change-Id: Ie328ebd2d31160bf07b86b760fbb5ce7108a532d
Reviewed-on: http://gerrit.openafs.org/7473
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it is
unnecessary noise.
Change-Id: I8287709413fe0e34f417936d1fc64c421fea6d28
Reviewed-on: http://gerrit.openafs.org/7472
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it is
unnecessary noise.
Change-Id: I6a33ab25b092faa96c764f0a469d052c181344ee
Reviewed-on: http://gerrit.openafs.org/7471
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it is
unnecessary noise.
Change-Id: I58b7695e3cf4b99f17869a46e1efca5c7daf6ff7
Reviewed-on: http://gerrit.openafs.org/7470
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it is
unnecessary noise.
Change-Id: Iaddc6c44041dc77a576f32133b2aa80aefce5cb2
Reviewed-on: http://gerrit.openafs.org/7469
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it is
unnecessary noise.
Change-Id: Ie41f7b831f0ba70796649e2493e014fe44f1c39e
Reviewed-on: http://gerrit.openafs.org/7467
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() and osi_Alloc both return (void *). So, don't bother
casting the return value before assigning it - its unnecessary
noise.
Change-Id: I71a66a8fa0d9f49f4833dd77bbb55422e6f20d6f
Reviewed-on: http://gerrit.openafs.org/7466
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it's
unecessary noise.
Change-Id: I270b81c239afb9ac4bd8121adfd5a48fcfac9c3b
Reviewed-on: http://gerrit.openafs.org/7465
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it's
unecessary noise.
Change-Id: I74b1f3ba3000ecaeb3da8eb3201f6050c9972c4e
Reviewed-on: http://gerrit.openafs.org/7464
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it's
unecessary noise.
Change-Id: I5a7c800e2836e7401f5f8bccf1aa2a1b223100b7
Reviewed-on: http://gerrit.openafs.org/7463
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it's
unecessary noise.
Change-Id: I9a3387c07c6ee02efbaa1ec0035fcea8691f171c
Reviewed-on: http://gerrit.openafs.org/7462
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it's
unecessary noise.
Change-Id: I9d41f2f1aee03d9317b4c5780fb00e13d30b0821
Reviewed-on: http://gerrit.openafs.org/7461
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it's
unecessary noise.
Change-Id: I6a878d53007e27179c11d8f745b09dacade4c83d
Reviewed-on: http://gerrit.openafs.org/7460
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it's
unecessary noise.
Change-Id: If6f0dbc00b31f5d9a1622984c6ac9eba83c04900
Reviewed-on: http://gerrit.openafs.org/7459
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
malloc() returns a (void *) on all of our current platforms. So,
don't bother casting the return value before assigning it - it's
unecessary noise.
Change-Id: I534d424da13e03d3c0f0de3dddf1dd19088d1659
Reviewed-on: http://gerrit.openafs.org/7458
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
All of our allocators return (void *) values, so there is no need
to cast the returns - doing so is just noise.
Change-Id: I6b1009efbb5034225a4af0e19d03b74c628dfb82
Reviewed-on: http://gerrit.openafs.org/7457
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
calloc returns a (void *) argument. So we don't need to cast its
output before storing it.
Change-Id: Ibdfc61a8a2cf9779a0bfb17ff8ebf1f3bef3ea0e
Reviewed-on: http://gerrit.openafs.org/7456
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
realloc takes, and returns a (void *) argument. So we don't need to
cast these parameters into, or out of, realloc. Doing so is just
noise, so don't bother.
Change-Id: I64e721703536515b2e315e3b033eb2be14f7e18c
Reviewed-on: http://gerrit.openafs.org/7455
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
Rather than doing
a = malloc(sizeof(me));
memset(a, 0, sizeof(me));
Just use
a = calloc(1, sizeof(me));
This is simpler, shorter, and removes the potential for the size of
the memset not matching the size of the malloc (or the target of the
memset being wrong!)
Where the size is of the form (n * sizeof(me)), we also use
calloc(n, sizeof(me));
Change-Id: Ia0f75665c1031fd2982eee0e1d8c8ebe23d7fbc0
Reviewed-on: http://gerrit.openafs.org/7454
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
free(3) is defined as:
void free(void *ptr);
so there is no need to cast the pointer argument being past to it to
(char *), (void *), or whatever other randomness happened to be floating
through the author's mind. Remove all such casts, as they are just
noise.
Change-Id: Ib7c8c259d53e9ce74fc486cb0997360044795395
Reviewed-on: http://gerrit.openafs.org/7453
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
Instead of allocing a maximal string, and using snprintf to
construct each possible DNS search string, just use asprintf to
construct each string. This greatly simplifies the code, and makes
it much less likely that maths errors can creep in causing buffer
overflows in the future. The downside is that we have more round
trips to the allocator, but that shouldn't matter in this context.
Change-Id: Iae9ab7d45f454c90a937354c71c9ec7fa2c55b69
Reviewed-on: http://gerrit.openafs.org/7452
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementix.org>
