Add some documentation explaining some of the minor quirks in
AFS::ukernel; specifically how some functions look a bit different
than in plain libuafs.
Change-Id: Ib8b18720c3fa6087de98d58ba8381f1ae4fb2f10
Reviewed-on: http://gerrit.openafs.org/3902
Tested-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Add trace logging to cm_StoreMini which never had it before.
Improve the logging of cm_BufWrite by adding the trucPos value
which is the new length of the file that is being reported to
the file server.
Remove cm_buf_t data references when cm_BufWrite is performing
operations using rx_Writev and iovec structures.
Change-Id: I28798eb2232e739f665eb51499305a7a1898f824
Reviewed-on: http://gerrit.openafs.org/4924
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
cm_Unlock() and cm_UnlockByKey() duplicate a significant amount
of code. Refactor it into a new static function, cm_IntUnlock()
which handles the process of downgrading or releasing a file
server lock depending upon the lock state of the cm_scache_t
object.
Change-Id: Ic5db7b3928fc0477f155183326321717ea04ace0
Reviewed-on: http://gerrit.openafs.org/4923
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
cm_NewServer() can result in a call to cm_UpdateVolumeLocation()
if a server probe is performed. In order to avoid recursive
calls to cm_UpdateVolumeLocation() do not probe new servers from
within cm_UpdateVolumeLocation().
Change-Id: Icdb8efe030ae3e1f714ca72ad741bd38c692697f
Reviewed-on: http://gerrit.openafs.org/4922
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
When we have a 64-bit ino_t, use the full 64 bits, instead of always
limiting ourselves to 32 bits.
Change-Id: I8f9f552b230e1723c8b77bfe92213ca43816240c
Reviewed-on: http://gerrit.openafs.org/4921
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
Instead of having two separate afs_calc_inum functions, just have one
afs_calc_inum, and split off the md5 inode code into its own function
under a LINUX20 ifdef.
Change-Id: I400d2e70403905da40055238ffc2bb785659e9e0
Reviewed-on: http://gerrit.openafs.org/4920
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
The algorithm for calculating inode numbers was copied in several
places in libafs. Make them all use afs_calc_inum instead.
Change-Id: Ie835a0b92bf940b78090fd7ca4e36aace1df9767
Reviewed-on: http://gerrit.openafs.org/4919
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
When calculating the inode number for a file with md5 inodes, include
the cell number in the calculation, in order to reduce collisions
between cells.
Change-Id: I4b939042dd993419f785a78e87e68cf346b56e26
Reviewed-on: http://gerrit.openafs.org/4902
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Specify -Wold-style-definition when compiling lex.yy.c. This allows us
to compile when --enable-checking is specified and our lex generates
code with old-style function definitions.
Change-Id: I09c87fd6274b5850952a8aaf63f0e3358100d567
Reviewed-on: http://gerrit.openafs.org/4905
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
afs_calc_inum can currently yield an inode of 0 if MD5-based inode
numbers are turned on. Some userspace applications (and for some
platforms, maybe even the kernel) make certain assumptions about the
inode number for a file; in particular for example, 'ls' will not
display a file with inode 0 in a normal directory listing.
So, read the md5 digest until we get a non-zero result. Fall back to
the non-md5 calculation if we still somehow end up with a 0.
While this case may at first glance seem to be extremely rare, in
practice it can occur, as the current calculation for volume
538313506, vnode 26178 does actually yield a 0.
Change-Id: Iee1ef4cc2ad66269f2c677e29d586ef0964d7c70
Reviewed-on: http://gerrit.openafs.org/4901
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Linux 3.0 is just a rebranded Linux 2.6.40, so we can use the same
sysname and the same code.
Change-Id: I154334dc5b73980a0d9b8abd9fd6645994f63829
Signed-off-by: Anders Kaseorg <andersk@mit.edu>
Reviewed-on: http://gerrit.openafs.org/4843
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
If we dec a linktable entry or get a free tag from the link table,
there is no reason to FDH_REALLYCLOSE the linktable fd handle.
FDH_REALLYCLOSE is the same as FDH_CLOSE, except that it tells the
ihandle package that the file handle will not be used again soon. If
we dec a linktable entry or get a free tag, there is no reason to
think that, so just FDH_CLOSE the handle instead.
Change-Id: I5f48a78528a75b984201f9ee0563b142ef59baf9
Reviewed-on: http://gerrit.openafs.org/4903
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
When a volume is put into an error state via the FSYNC_VOL_FORCE_ERROR
command, we clear the salvage state informaton on it, since we're
forcing it offline and thus inaccessible. However, if we are forcing
it to an error state because the volume needs salvaging, we just
salvage it. In this case, do not clear the salvage state, since we
need to know if we've already requested or scheduled a salvage so we
can correctly keep track of the number of salvages performed.
Change-Id: Ic4efd7a78bfb1b99a5308f0c67e81f4779dfe545
Reviewed-on: http://gerrit.openafs.org/4900
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
.epub is generated using dbtoepub which is still considered alpha
software apparently and installed in a non-standard place. for now,
use the docbook stylesheet location to find it. .mobi is generated
using kindlegen from the .epub in order to have a real toc.
there is some preprocessing with a custom stylesheet to make
things "look right". see mobi-fixup.xsl.in
Change-Id: Ice92e701e2f921e70c0f98683b5e9ab44a347e3b
Reviewed-on: http://gerrit.openafs.org/4887
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Currently, a struct vcache has a multiPage counter, indicating how
many afs_getpage requests are in-flight for that vcache that involve
retrieving multiple pages. Any dcache associated with such vcaches are
then avoided when choosing dcache entries to evict from the cache,
since we may deadlock when trying to evict a dcache entry from one of
the earlier afs_GetOnePage calls in a particular afs_getpage request.
This behavior can cause the client to become unusable if the cache
becomes full, and the only items in the cache are dcache entries in a
file that has an in-flight multi-page afs_getpage request. Since, in
that case, we cannot kick out any entries from the cache, and so we
wait forever to wait for the cache utilization to go down.
To prevent this from occurring, record exactly which ranges in the
file have in-flight multi-page afs_getpage requests, and just avoid
dcache entries in those ranges. This way afs_GetDownD can evict dcache
entries in the same file, but still avoid entries that would cause a
deadlock.
Also add some comments explaining this situation a bit more.
Change-Id: Idb305c8b7511065301739542772d16d4fe8cd574
Reviewed-on: http://gerrit.openafs.org/4896
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Currently there are two preprocessor defines related to bozon locks:
AFS_BOZONLOCK_ENV, and AFS_NOBOZO_LOCK. The former creates the pvnLock
member of a struct vcache, and controls calls to e.g. afs_BozonLock in
cross-platform code. The latter, if defined, turns calls to e.g.
afs_BozonLock into no-ops.
It doesn't make any sense to have both of these, since if
AFS_BOZONLOCK_ENV and AFS_NOBOZO_LOCK are defined, the pvnLock member
exists but is never used, since afs_BozonLock &co are no-ops. On
Solaris, the only platform where AFS_NOBOZO_LOCK is currently defined
(DUX used to define it before DUX was dropped), this is the case.
So to make things a bit more clear, get rid of the AFS_NOBOZO_LOCK
define, and just use AFS_BOZONLOCK_ENV to dictate whether we do
anything with bozon locks (ppc_darwin_80 appears to be the only
platform at this time).
Remove AFS_BOZONLOCK_ENV from Solaris param files, since it doesn't
use bozon locks. Remove all references to pvnLock in Solaris-specific
code.
Change-Id: Id72c14ec5485d35b853d38e2ea1a944a385f2b5b
Reviewed-on: http://gerrit.openafs.org/4889
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
Remove support for all Solaris and SunOS platforms prior to Solaris 8,
since Solaris 7 reached end-of-life in August of 2008. Remove all
non-documentation references to sunx86_57 and earlier, sun4x_57 and
earlier, and AFS_SUN57_ENV and earlier.
References to AFS_SUN58_ENV have been changed to AFS_SUN5_ENV where
appropriate, and AFS_SUN5_ENV now implies Solaris 8.
AFS_SUN57_64BIT_ENV has been renamed to AFS_SUN5_64BIT_ENV.
Change-Id: Ia64ce7da7bfc685fa28a5119c51ec740625456e3
Reviewed-on: http://gerrit.openafs.org/4888
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Introduces the first batch of man page testing as part of
the TAP tests. We would like to fail, for example, when
someone has added a new command to vos but not AHEM documented
it.
For now, the tests consist of checking to ensure that for
every subcommand listed in the output of "command help"
(e.g. vos help), fail the test if there is not a man page
for those (e.g. vos_delentry.1 etc).
Copy any of the -man-t tests and edit to make a new one
All tests make use of a simple new Perl library stored
in tests-lib/perl5 (a new area, not just named 'lib'
because I didn't want it to be confused with a s test
for a 'lib' in the src).
Change-Id: I1e07adafe718c4549f1855c5e5b0d10dd9ab5f00
Reviewed-on: http://gerrit.openafs.org/4846
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Prefer "Georgia" as a typeface over the less readable
Times New Roman, but with Times New Roman then "serif"
as fallbacks. Georgia is available everywhere.
Provide 10px top/bottom and 30px left/right margin on
the main body for readability. Margins are good.
For review:
Old:
http://docs.openafs.org/Reference/8/kas_apropos.html
New:
http://kickflop.net/openafs-man-20110617/kas_apropos.html
Change-Id: I687b2007ef56019c5b683a43ac3811426ec825b0
Reviewed-on: http://gerrit.openafs.org/4850
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Simon Wilkinson <sxw@inf.ed.ac.uk>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
multi_Init() registers an arrival procedure which is called when
the first response packet for the call arrives. If the call times
out the multi_Body loop will call rx_EndCall() and then set
multi_h->calls[multi_i] to NULL. If the first data packet of the
call arrives before rx_EndCall() is executed, then the arrival
procedure, multi_Ready(), will be executed adding the call to the
firstNotReady list. When the multi_Body loop attempts to process
the call from the firstNotReady list it attempts to dereference
the NULL multi_call. This race was introduced by
be4abb4ec8.
multi_h->calls[multi_i] is set to NULL as an indicator to
multi_Finalize() that rx_EndCall() has already been processed
for the call. When rx_EndCall() is executed the arrival
procedure is cleared.
If rx_EndCall() has already been processed, the fact that
the arrival procedure has been executed must be ignored. Add
an additional check in multi_Body for a non-NULL call pointer
to skip the startProc and rx_FlushWrite processing on the
no longer existent call.
Note that it is not safe to hold onto the call reference after
rx_EndCall() has been processed since the call slot may be
reused for a new RPC before the multi processing on all calls
is complete.
Change-Id: Ib4694a7e1d133f621d15e79534a42f780b141e34
Reviewed-on: http://gerrit.openafs.org/4890
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Change -n to -dryrun for "don't do it, show it though" operation
to be in line with agreement on -dryrun in place of -noexecute
or -n. Updated man page POD sources to reflect the changes
and updated README to remove these specific todo line items.
Change-Id: I5c1361dcc866ea3d1efbb0c0ddcd4a7fe513c816
Reviewed-on: http://gerrit.openafs.org/4827
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
refactor ListAddrs to be more readable. Clarify that -uuid and
-host cannot be issued at the same time. Rename 'nentries' to
'max_index' so it is clear that ubik_VL_GetAddrs() is issued
to set an upper-bound for the number of subsequent ubik_VL_GetAddrsU()
calls that are issued when neither -host nor -uuid are specified.
Change-Id: Icbd511722728396fda837abba39a3e6809f4d26c
Reviewed-on: http://gerrit.openafs.org/4754
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
cm_SyncOp/cm_SyncOpDone is used to synchronize the RPC processing
to ensure that calls which are in conflict cannot occur at the
same time but also to ensure that the ordering of operations
is consistent. cm_MergeStatus() was in many cases executed after
cm_SyncOpDone() removed the synchronization barrier which in turn
permitted status information to be applied out of order. Side
effects could have included data loss due to client side file
truncation. More commonly two StoreData RPCs would have their
status information applied out of order forcing the cache manager
to invalidate all of the cached data for the file.
Change-Id: I8073da701cf11cd0df2b901e81180de7c193ae23
Reviewed-on: http://gerrit.openafs.org/4891
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
smb_T2SearchDirSingle() must not fail directory search requests
for the _._AFS_IOCTL_._ file. Although this file does not actually
exist, it is successfully processed by CreateFile operations.
Therefore, an explicit search for it should return a valid answer.
Change-Id: I9df3443f0bdf0dd05caac74f4de3a0cc64d26800
Reviewed-on: http://gerrit.openafs.org/4884
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
MS11-043 adds response validation for SMB_COM_NEGOTIATE messages
received by the SMB Redirector. OpenAFS failed to properly specify
a Challenge and DomainName in the response when the security mode
is SMB_AUTH_NONE (or share with password). This patchset corrects
smb_ReceiveNegotiate() so that it adheres to the protocol specification.
FIXES 130033
Change-Id: I3dc6e571326c7259a39d30bd80b5986ff35c743c
Reviewed-on: http://gerrit.openafs.org/4886
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
This reverts commit 3cd3715e60,
which replaces an ack with an ackall; ackall processing does
not actually mark all packets acked when it is received, so
it is insufficient.
Change-Id: I7ee90e1190688570cf3b268229972064480283cc
Reviewed-on: http://gerrit.openafs.org/4837
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
index.html might not be built if you are building anything other than
the html/chunk.xsl but in general this is better than before.
Change-Id: I2e1b98c8a3fce65cb702274d5e276989031e338b
Reviewed-on: http://gerrit.openafs.org/4822
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
It would seem xsltproc -> fop -> pdf is the "modern" way to generate
pdf from docbook now. The hard part is finding the stylesheets.
This should work for fedora, sles and debian. Additionally, it brings
some consistency--xsltproc for all the conversions. You can still
override via configure options if you prefer something else.
Change-Id: Id779e9473a6759daddc9a61be714109b27da980e
Reviewed-on: http://gerrit.openafs.org/4821
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
"local" links to section heads inside the same pod page should be written
L</OPTIONS> instead of L<OPTIONS>. the other broken links are assorted
typos and capitalization changes.
Change-Id: I18b57a81d984501309e96dcd4c04582e744a6db2
Reviewed-on: http://gerrit.openafs.org/4831
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Ken Dreyer <kdreyer@usgs.gov>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
If we encounter a preattached volume during GetVolume, we currently
ignore vp->specialStatus before trying to attach. However, we will
generally always fail to attach due to a conflicting vol op, but even
if we don't, GetVolume always returns an error later on if
vp->specialStatus is set. So, same some processing and attempted
attachments by bailing out sooner if vp->specialStatus is set.
Change-Id: Id5df08fe31073485e4d733c42310b42090ee8e7b
Reviewed-on: http://gerrit.openafs.org/4874
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Not every field in the summary header in RecordHeader is set, leaving
some used uninitialized when we copy to the given volumeSummaryp (like
'deleted'). Zero out the header before we do anything.
Change-Id: I94f6e5c8b8c60675de25eb6f8290de9545ff4dd0
Reviewed-on: http://gerrit.openafs.org/4876
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Currently dasalvager links to vlib.a. But vlib.a is built without any
DAFS defines, and so the size of a struct DiskPartition64 is different
(since dasalvager is built with AFS_DEMAND_ATTACH_UTIL). Build our own
copies of the volume package files instead, with
AFS_DEMAND_ATTACH_UTIL defined.
Change-Id: I7b9f965fc18b6e6b71275912684ad041aaef19f4
Reviewed-on: http://gerrit.openafs.org/4875
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
DRead was missing a return statement in one of the cases where we
found the buffer we were looking for, so we locked the buffer but kept
looking. Return it instead.
Change-Id: If72a0ba3ce60a847f2796b51a82f0f473bbc608a
Reviewed-on: http://gerrit.openafs.org/4877
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
attach2 wants to set specialStatus to VBUSY in certain conditions
(such as, it discovers a conflicting vol op where VVolOpSetVBusy_r is
true). However, specialStatus may already be set to something else,
like VMOVED if the volume is being moved off of the server. This can
happen if the volserver has checked out and FSYNC_VOL_MOVE'd a
preattached volume but hasn't deleted or checked the volume back in
yet.
So, if specialStatus is already set, don't touch it, so we don't start
reporting VBUSY errors to clients when we should be reporting VMOVED,
or some other error code previously set.
Change-Id: Icb2895036620f186230e1558b8bc04d18cc45c86
Reviewed-on: http://gerrit.openafs.org/4873
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
The current code only exits fast restart when we receive an ACK
packet that contains no missing chunks at all. On a network that is
dropping a reasonable chunk of its packets, this means that we spend
most of the call in fast recovery. (I originally found this by running
with the intentionally drop packets feature set to 10%)
TCP's fast retransmit behaviour is that we stay in fast recovery until
we receive our first non-duplicate acknowledgement. In TCP that means an
acknowledgement that moves the window. In RX, it is an acknowledgment
that ACKs a new packet.
Change-Id: If8e461dd91315be845397dd1bf42771c9223d156
Reviewed-on: http://gerrit.openafs.org/4869
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
The RX transmit engine limits the number of packets sent whilst in
loss recovery to one per invocation of the transmit engine. As the
engine cannot be called by the application thread whilst in recovery,
this means that we end up being limited to one packet per ACK received,
which means that despite a growing congestion window we'll only send
one packet per RTT (in effect, a congenstion window of 1).
This will remain the case until we exit recovery, and all of a sudden
can send a large number of packets. If this is larger than the current
capacity of the network, we'll probably end straight back in recovery
again.
Let the congestion window do its job, by removing this arbitrary limit.
Change-Id: I9f21138662b0918d6d2b885f0fb5ada7fa4c79ec
Reviewed-on: http://gerrit.openafs.org/4868
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Two different threads can cause a call to enter recovery. The event
thread will move a call into recovery as a result of a timeout, or
the listener thread will move it there following a fast retransmit.
In both of these cases, recovery looks different. In the case of
a timeout, we enter slow start, starting as if we were begininning
transmission for the first time. Following fast retransmit, we enter
fast recovery, with different starting parameters than those coming
from slow start.
As a reslt, the current behaviour, where either call sitting in
FAST_RECOVERY_WAIT causes the other to simply return is inappropriate.
Further investigation indiciates that FAST_RECOVER_WAIT is actually
uncessary. There is no harm caused to a thread which is currently
blocked on the network in the middle of a transmit, in adjusting the
window size underneath it. As both of these states collapse the window,
that thread will simply cease sending earlier.
So, simplify the code, and remove the potential race between event and
listener by removing the FAST_RECOVER_WAIT state.
Change-Id: Ic2e7606136ca04c869685345b63101c346ce702b
Reviewed-on: http://gerrit.openafs.org/4867
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Since I mistakenly wrote commit 36e2d13b, RX hasn't entered congestion
avoidance when a loss event occurs. This is bad, because on todays
networks the majority of packet losses are due to some form of
congestion.
Now that the timeout code has been restructured, the chances of entering
the retransmit routine in error are much much smaller, so this code
needs to be restored.
This change reverts 36e2d13b55. However,
the original RX code has the problem that it assumes that all forms of
fast recovery are the same - in particular, that the call settings that
result from entering fast recovery due to a fast retransmit are
identical to those resulting from a timeout. This is not the case, and
this will be fixed in a later change.
Change-Id: Iedb34437db9fcfbc90307b01e566a8d089eef4bb
Reviewed-on: http://gerrit.openafs.org/4866
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
When we retransmit a packet, we may be doing so because the RTT of the
connection has grown dramatically larger than earlier within the call.
However, RX doesn't permit all ACKs to retransmitted packets to be
counted within the RTT calculation.
So, adopt the same approach as Karn developed for TCP, and as described
in detail in RFC2988. When a retransmit event occurs, backoff the
connection RTT by doubling its value, and hold at this doubled value
until either another retransmit occurs (in which case we back off again,
up to a predetermined ceiling), or we receive an ACK packet which we
can use within the RTT calculation, in which case we drop back down to
the newly measured value.
This change replaces the per-packet backoff strategy originally
implemented in RX (which, whilst allowing resent packets more chance of
arriving, doesn't help with computing a correct RTT).
Change-Id: I715dc673a0b379733e3be26fa594ea13f492a58f
Reviewed-on: http://gerrit.openafs.org/4865
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
With the existing clock_Add code, the following:
struct clock a = {2, 800000};
clock_Add(&a, &a);
gives a clock value of {6, 600000}, rather than the expected {5, 60000}.
This is because the ordering of instructions leads it to double count
the carry on the seconds field. Reorder the instructions so that the
carry is correctly applied.
Change-Id: Ia71b387ce521a11e4caf9ec200907efe1d2be8ff
Reviewed-on: http://gerrit.openafs.org/4864
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Create a new function, rxi_Resend, which is the entry point to running
the transmit queue as a result of a resend event. This concentrates all
of the resend logic into one place, removes the need for
rxi_StartUnlocked, and means that rxi_Start's arguments don't need to
match those of an event handler.
Change-Id: I550ebbbae63b7d659bb980eea709a14882038914
Reviewed-on: http://gerrit.openafs.org/4863
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
RX maintains a retryTime for every packet that it has transmitted,
which is held as the time that that packet was sent, plus the smoothed
RTT of the connection. If a packet is in the queue with a retryTime
older than the current time, then it is resent at the first opportunity.
In some circumstances, this first opportunity will be as a result of
the resend event timer expiring, in others it will happen as part of
a normal queue run.
There are a number of problems with this approach on congested networks.
Firstly, on a network with a large window size, which is in "normal"
flow, it means that we will never actually perform fast retransmit as
the timeout for this packet will have expired before we have received
any further ACKs. This is because, on a network with a relatively stable
RTT the ACK for packet n+1, n+2, or n+3 cannot arrive before the
expected time of arrival of the ACK for packet n. As we retry
immediately this expected time of arrival has passed, we never have the
opportunity of using these later ACKs to learn that packet n is lost.
Secondly, the fact that we may resend packets from a "normal" queue run,
rather than as a result of a resend event, means that there is no clear
entry point for resends. As resends should be assumed to be a result of
network congestion, and result in both the call throttling back, and the
RTT being increased, this lack of a clean entry point makes things
tricky.
As a solution, this patch changes the way in which retransmit times are
applied to use the algorithm described in RFC2988.
*) Whenever we send a new packet, we start a timer for the current call
rto value if one isn't already running.
*) Whenever we receive an ACK that acknowledges new data, and we have
packets that are sent but not yet acknowledged, we restart the
retransmit timer using the current rto value.
This alogrithm solves the first problem, as it means that if the
connection is still flowing, we will continue to receive ACKs, and we
can enter fast retransmit.
In implementation terms, we longer track a retryTime per packet, and
instead simply record if a packet has been sent or not. Packets which
have been sent may only be resent as a result of a resend timer
expiring, or of entering fast retransmit, so solving the second issue.
Change-Id: Ic1e2b32876197f8a1fb620b7d70c0c3ec3f6bb88
Reviewed-on: http://gerrit.openafs.org/4862
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
RX uses the TCP RTT smoothing algorithm as described in RFC2988.
However, the TCP algorithm is designed to accept samples from a
single connection, accepting a new sample once per RTT.
RFC2988 suggests that "when multiple samples are taken
per RTT the [ alogrithm ] may keep an inadequate RTT history."
In RX's implementation, we use a single instance of this alogrithm
per peer, and input all of the samples from all of the active calls
and connections into this same instance. This leads to us taking
a significantly (potentially many magnitudes) larger number of samples
per RTT, and rapidly losing the RTT history. With RX's implementation,
short lived network events may easily bias the RTT, and cause large
numbers of packets to time out.
This change fixes this by moving the RTT calculation onto a per call
basis. We still update the peer with our caclulated value, so that new
calls may be created with an RTT corresponding to the current value for
the connection, rather than having to start high and converge downwards.
Change-Id: I2ed6bce63adf160c03518686ec25cbecc5084f5f
Reviewed-on: http://gerrit.openafs.org/4861
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: Derrick Brashear <shadow@dementia.org>
The "testclient" utility is built as part of the build on Windows.
Fix it so that it actually builds on Unix, so we can test changes to
testclient there.
Change-Id: I751e905f6d9676598e3b7be9ccaf321250804f11
Reviewed-on: http://gerrit.openafs.org/4872
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Currently, the rx server thread idle queue is used in an LRU manner.
This means that we round robin requests between all of the threads
configured on a given system, which means that we end up thrashing
CPU caches on machines whose workload doesn't require that all of
the configured threads be used.
Change this so that we always use the most recently idle thread. This
isn't as "fair" to all of our waiting threads, but should mean that we
scale better on SMP machines, as a thread that is recently idle is
likely to have been recently scheduled.
Performance numbers to follow ...
Change-Id: I6072183b5ca72754755c3ac854e30a064c4187e5
Reviewed-on: http://gerrit.openafs.org/4871
Reviewed-by: Derrick Brashear <shadow@dementia.org>
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Jeffrey Altman <jaltman@openafs.org>
Tested-by: Jeffrey Altman <jaltman@openafs.org>
include our shared libraries in the package when we have them
Change-Id: I18a38c69410a7fded361b05479c93698e3d618e2
Reviewed-on: http://gerrit.openafs.org/4870
Tested-by: BuildBot <buildbot@rampaginggeek.com>
Reviewed-by: Derrick Brashear <shadow@dementia.org>
