CVS log for pvfs2/src/proto/pvfs2-req-proto.h
	Help

Request diff between arbitrary revisions / Display revisions graphically

Merged robust security code into Orange-Branch.

Set the minor protocol version back to 0 (based on discussion)

added new uid management app under src/apps/admin/pvfs2-get-uid
also included are a server and client side sm and the uid mgmt implementation files

merge with Orange-Branch

1. bump up TROVE to version 0.1.6, add a data migration function migrate_collection_0_1_5.
Will create a single dirdata object in distributed directory structure. More dirdata handles will be requested when calling split function.
2. some code cleanup in mgmt-create-root-dir.sm and pvfs2-req-proto.h.

not entering admin mode when creating distributed root directory. will save startup time to <3 seconds on 4 servers.

trove/bmi version change, start working through handle changes from server-side file system creation

remove extents, handle ranges, precreate pools and various other things that aren't needed in their current forms with UUIDs

Committed Windows client code to Orange-Branch.

uuid converstation, PVFS_x_position changes, server/client side first pass done

move PVFS_ds_position to different types when appropriate, pass a flag around instead of using special canary values in the handle. Things will be broken for a while.

changes to PVFS_handle_X

formatting only

Modifications to perf counter system
Minor formatting and documentation edits

merge with Orange-Branch

distributed directory structure for root and lost+found directory. In pvfs2-server.c, it will check if the root directory is set. Need 10~20 seconds for creation when the servers are launched at first time.

redesigned lookup code to work with security

reverted to Orange-Branch version of lookup code. many modifications will
be needed before the code will compile again.

fix to segfault in list-eattr if client provides too large of buffer, found by Mike Marshall. Also includes other changes to enforce attribute name, value and list lengths for eattr operations on the server side and list-eattr on the client side.

Porting client

Merged in changes from Orange-Branch. There were a number of bugs fixed there since this branch was created.

Fixes to lookup.

Next round of distributed directory changes.

incorporate dist-dir-struct to the source tree. It's compilable but not runnable. still have a lot to change and add. Commit as a touch base.

*** empty log message ***

*** empty log message ***

merging Orange Branch changes in

fixes to make the new server state machines compile with robust security

encoding fixes

initial merge with Orange-Branch. much will be broken

lots of little fixes that have been in the works for a while

Preliminary work for distributed directories - making client aware of DIR_ENT_KEY.

*** empty log message ***

*** empty log message ***

removed anything and everything related to certificates

revert cu-security-branch to before the attempted merge with Orange-Branch

merge of B2O-Blue-Sync,HEAD,and latest of Orange.

commiting changes from merge of Orange and Blue into test branch.

Merged failover with tree code.
Modified Files:
  Tag: Orange-Branch
 	prepare src/apps/kernel/linux/pvfs2-client-core.c
 	src/client/sysint/client-state-machine.c
 	src/client/sysint/client-state-machine.h
 	src/client/sysint/finalize.c
	src/client/sysint/sys-getattr.sm
 	src/client/sysint/sys-io.sm
	src/client/sysint/sys-small-io.sm
 	src/common/gossip/gossip.h
	src/common/misc/msgpairarray.sm
 	src/common/misc/pint-event.c
	src/common/misc/pint-event.h
 	src/io/bmi/bmi.c
	src/io/dev/pint-dev.c
	src/io/job/job.c
 	src/kernel/linux-2.6/devpvfs2-req.c
 	src/kernel/linux-2.6/file.c
	src/kernel/linux-2.6/pvfs2-mod.c
 	src/proto/PINT-le-bytefield.c
	src/proto/pvfs2-req-proto.h
 	src/server/create-immutable-copies.sm
	src/server/get-attr.sm
 	src/server/small-io.sm
	src/server/tree-communicate.sm
 	src/server/request-scheduler/request-scheduler.c
 	src/server/request-scheduler/request-scheduler.h

*** empty log message ***

Add retry_msgpair_leaf to request parameters for tree_get_file_size.

*** empty log message ***

Completed coding and debugging of small-io failover.

bug and memory fixes to getvalue operation

checkpoint

checkpoint

Merge branches

checkpoint

merge from HEAD: http://www.pvfs.org/fisheye/changelog/PVFS/?cs=MAIN:pcarns:20090727185444

Timing measurments

storing parent de implementation

Merge of Orange-Branch, Orange-FailOver-Becky, and version 1.164 of sys-io.sm (do not unstuff on a read).
This version of Orange-FailOver-Becky supports failover for large files only.  The next iteration will
support small-io failover.

merged in changes from latest Orange-Branch

remove debugging cruft

Change get-path request

state machine implemenation work

merged in changes from summer at LANL

initial get-path state machine and associated changes

*** empty log message ***

Initial commit of parallel fsck code

Merge range query changes and other fixes

patches from Dries Kimpe:
- Added check before redefining USE_GNU in pvfs2-server.c
- Made use of endecode_fields_xxx macro's consistent and solved warning
in case they were compiled out.  The macro should be used as a function
call, and thus have a semicolumn at the end.

Initial import of branch supporting keyval attribute/value lookup

Completed changes to PVFS_object_attr.  Changes to get-attr.sm are still underday

fixed invalid free() in io.sm.

*** empty log message ***

Added the server to server communication module within pipeline.sm
to pull the small portion of data when the data is not aligned on strip boundaries.

Changed the default file stripe size to 256KB (originally 64KB)
so that pipeline unit is equal to it.

Setting up Orange Branch

*** empty log message ***

reverse merge fixes made to HEAD since 2.8 branch point.

comments

bug fix: let PINT_SERVREQ_MGMT_SETPARAM_FILL handle case of NULL value
argument for operations that don't have a value to set.  This fixes seg
fault in pvfs2-drop-caches utility

- Removed max3 macro and defined max() macro and used max(max(a,b), c)
  for extra_size_PVFS_object_attr definition
- Redefined PVFS_REQ_LIMIT_LISTATTR to 113 which is the largest set that fits in the
  1MB preallocated space.
- Redefined pvfs2-ls.c: MAX_NUM_DIRENTS to 113 to match PVFS_REQ_LIMIT_LISTATTR.
- Trac Ticket #70

new authentication protocol

merging hints/events code into trunk.

tree communication implementation

reverse merge from HEAD.

reverse merge from HEAD to directio branch

added a credential to the setattr request message

committing patch from Bart Taylor to increase the number of directory
entries retrieved per readdir in both gnu ls and pvfs2-ls.

reverse merge of changes to HEAD since branch.  Includes small file changes.

Merging small files branch to head.  Includes server side precreation of
data files and file stuffing.

fixed a race condition between the cond_wait and signal in db-rep-send.sm
fixed some memory leak

dspace, keyval, coll_attr dbs are now created within the environment.
A whole lot of other things try to make replication work

db-replication state machine which processes the incomming replication message

Added flag to mgmt iterate handles function to allow it to request reserved
handle values.  Will be used to get a list of known precreated handle values
so that they can be accounted for in fsck and validate tools.  Untested.

Updated lookup/getattr on server/client.

Pass the fs_id with the replication message

made getattr conform to current conventions. fixed bugs in capability
verification and hash lookup.

set is_rep_master in config to 2 to indicate no replication for mkspace

db-rep-send state machine added

fixed size calculation and cleanup routines for le-bytefield encoding

updated all references in include/, src/client/, src/common/, src/io/,
src/proto/, and src/server/ to use the new PVFS_credential in place of the
old PVFS_credentials. the admin apps in particular need to be updated to use
the new API.

Reverse merged and ported to HEAD.

Comment update, bugfix for encoding size allocation.

Updated encode/decode buffers.  Updated getattr permission checking.

More work on encode/decode.

modified getattr client sm to handle credential and capability. compiles
but doesn't actually work yet. for starters, all the apps that call
PVFS_[i]sys_getattr need to be fixed.

reverse merge of trunk changes since branch.

Updated capability encode/decode.  Fixed capability transmission across the wire...currently asks for a capability with every getattr and gives one back as well.

mo' changes, bug fixes to hints/events.  Fixed encoding bug for hints.

removed extra fields from request/response structures

reverse merge latest trunk changes to small-file-branch and update msgpair
usage where needed

Merge from HEAD.  Fleshed out credential.  Encode/decode added for credential and cleaned up for capability.  Added capability/credential into request/response structures.  Fleshed out comments in security module.

adding setparam to the list of mgmt operations; fixes bug in which
pvfs2-fsck leaves servers stuck in admin mode

protocol size limit for layout information; custom encode functions for
layout in order to enforce limit; remember to free decoding memory for
create request fields

added PVFS_capability to server request/response but without endecode

whoops, revert last commit for now (on wire server layout list has strings, not
BMI_addr_t types)

account for extra size possibly needed in create request to account for
layout information

*** empty log message ***

debugging

debugging updates

*** empty log message ***

*** empty log message ***

removing extent ranges from create request

hints and events.

syncing small-file-branch back up with trunk at small-file-branch-point2 tag
(reverse merge)

Implementation of create-file.

encode stuffed_size in protocol for getattr, make sys-getattr honor the
stuffed_size when possible, increment protocol version number because we are
no longer wire compatible with trunk

fill in the num_dfiles_req field in create requests, was previously being
set to zero

if unstuff is called for a file that has already been unstuffed, then we
need it to behave exactly like a getattr (retrieving dfiles etc. that are
necessary for sys-io.sm).  Tinkering with using getattr nested machine to
serve that purpose.  Also added attr mask to unstuff request.

sorting out some request types

typos

added some comments to guide changes for create-file

set up for create-file mod - stubbed out stuff, etc.
a few small bug fixes to statecomp and statemachine stuff

Big pile-o' changes to make Pete go mad.

The main changes are the implementation of the stuffed create path (now just the create state machine and operation), and the beginnings of the "unstuff" operation.

This also includes the request scheduler and state machine cleanups (required for using prelude with batch_remove) and whatever else I may have happened to change along the way.

This code is able to create files, but writing to them is almost guaranteed to break.

config file control of some precreate parameters.  rather conservative
default parameters for now

merge of config file limit increase and genconfig warning message.

The config file was limited to 16K because of a #define limit in pvfs2-req-proto.h.  I've increased the limit to 64K, which only limits the size of the allocated response structure.  Since the response can have an unlimited size, the config file size shouldn't be limited by the size of eager mode messages in BMI.
Also added the requested warning message to genconfig for config files larger than the limit, which is somewhere around 512 servers now.

Sam's prelude and scheduler updates

The normal path for a server request operation is to jump to the prelude state machine, which tries schedule the request based on the handle+fsid.  This meant that the request scheduler code had a switch/case returning the handle+fsid for each request we defined.  The function was also overloaded with a readonly parameter (whether the request modified any objects on the server), and a scheduling policy parameter (as the return code), which specified whether the request scheduling could be bypassed for this request.

I moved all that to a per-request parameter structure, that gets specified within the server state machine for the request.  Now there is one array of server requests (moved to pvfs2-server-req.c) which the request scheduler and prelude state machine can poke at to get parameters (defined in the state machine for that request), and based decisions on.

The patch includes some other minor changes:

* A fix to the BMI addr ref count increment/decrement.  Previously, the addr ref count was being incremented/decremented in the prelude/final-response state machines.  That requires that those state machines get invoked, which I don't want to have to depend on.  I've moved the INC call to the unexpected_sm, and the DEC call to where the state machine gets finished in the server.  That way we don't have to worry about whether we're using prelude/final-response.

* Cleanup of the request scheduler.  The way we were changing the mode from normal to admin was sort of embedded into the request scheduling code, so we had to hack the mode change from prelude_sm instead of in setparam.  Also, the request scheduler had to be aware of the setparam fields in the request.  I've moved the mode change code to separate request scheduler and job calls.  I think something similar could be done for the timer code (it doesn't really belong in the request scheduler), but I'll leave that for another patch.

* Changed the way we compile the server binary in Makefile.in to allow for multiple object files.

stubbed in state machine to handle refilling precreate pools

inode stuffing operation.  also added a keyval_remove_list.

stubbed in new request type: batch_create.  The plan is for this to react
like create, except that it can create many handles of the same type (on
one server) in one protocol exchange

commit of Murali's unifying config file patch.  Also adding a convert script for converting to the new single fs.conf format.

Zero out pad bytes after distribution.  Fix bug where an aligned distribution would always send an extra 8 bytes.  This requires a protocol version bump.

Zero pad bytes in server req header for valgrind.

Merge HEAD changes to TAS-branch.

walt3 reverse merge from head includes merges of pw's sm changes (no state declarations), cleanup of state machine code, and other general merging/fixes.

Created server requests.  Added them to the PINT_SERVREQ_LOCK_FILL macro.

Changed PVFS_ds_position from int32_t to uint64_t, allowing us to remove DB_RECNUM flag from our databases.

Also changed the dspace comparison function to read the dspace db front to back instead of back to front.

Also changed the dspace_iterate_handles code to use DB_MULTIPLE to improve
server startup times.

Update migration branch to current CVS version

Encoding and decoding.

reverse merge of HEAD to WALT3 branch.

Added necessary cases for PVFS_SERV_LOCK.

- Support for immutable files
- Cleanups in the pvfs2-bufmap interfaces (replace int with size_t)
- Utility for displaying extended attributes and setting them
- Support for disabling atime updates at the object level, file system level (mount time)
  and directories.

backmerging of HEAD to branch...

Merge HEAD into Walt's branch.
Rework new state machines to the new cleanups introduced by Walt.

privatize odd struct for server debug messages

Merge posix-extensions-branch to HEAD
This branch implemented patches to the 2.6.16 kernel for the proposed POSIX I/O extensions and those patches are under
the patches subdirectory.
It also implements the PVFS2 specific hooks for these system calls.

Tools that may be of immediate use to the pvfs2 general audience is the pvfs2-lsplus utility in src/apps/admin
that should be noticeably faster than the pvfs2-ls utility if there are a lot of objects in a single directory.
Other features are left out by configure and are not even built if the kernels do not support those features
and callbacks.

Changed behavior of migration: now the client initiates the source servers I/O. This allows concurrent migrations of a metafile's datafiles.
TODO: add another message exchange to modify a files (metadata) datafiles before the file is deleted. Also added (working) eclipse project properties.

Modification for migration added, some functions still need to be adapted after moving to a new branch...

PVFS2-hint support + minor changes (verbose request scheduler)

Added request statistics to scheduler, reference from datafiles to
parent metafile (needed for migration) and auto migration tool(unfinished).

Reverse merges from HEAD..

Move PVFS_MAX_SERVER_OP into the enum and audit places where it was used in comparisons.  Requiring manual adjustment of that value has bugged me for some time.

Backmerge with HEAD

Migration of datafiles between different servers works. Uses msgpairray.sm,
transfers only one datafile between two servers and rewrites metadata.
Todo: Adaption of request scheduler to ensure that migration and write
I/O do not happen concurrently on source dataserver.
Also allow users to specify command line parameter which supresses
transfer of pvfs2-hints.

Added tool which prints file mapping to server aliases. Added
experimental stubs for migration tools. Added kernel tgid to request ID.
Bugfix of PINT_cached_config_get_one_server_str.

Added PVFS-hint, changed tests for pwrite/odirect/transactions/aio dbpf
 support to allow users to explicitly enable or disable the functionality.
In the aio implementation is a bug since somebody else modified it.

reverse merge from trunk.  working for now.

Fixed a few of the gcc 4.1.x "dereferencing type-punned pointer will break strict-aliasing rules"
warnings

Reverse merge changes from HEAD to posix-extensions-branch.

- Misc. cleanups (replaced a big macro with an inline function)
- Removed parent_*time from the *dirent request protocol since time stamps are now
kept at the server and do not need to be sent over the wire (This is the reason why
major version of the wire protocol has been bumped)
- Adjust macros for the *dirent protocol accordingly.
- Add a few more debugging statements to kmod

Patch fixes the atime, mtime and ctime problems that people have
reported with PVFS2.
Patch changes the notion of who sets these times. Earlier behavior was the
clients providing the times to the server. New behavior is to have the
server set these times except in  the implementation of utimes() system
call whereby clients explicitly set the time (atime and mtime).
NOTE: We dont allow the clients to dictate the ctime setting at all since
that is consistent with all file systems.
So what we have is:
read, readdir -> atime update required
write, truncate -> mtime update required
ctime is updated on every setattr implicitly on the server.

Since there is no need for parent_*time fields in the server request for chdirent,
crdirent and rmdirent, the next step is to remove those fields, fix the endecode
macros and bump up the major version of the protocol.

Patch also implements these times lazily by flushing them only at the time
of a close rather than after every read/write which could be pretty
expensive.

Checkin of previous patches

protocol major version bump:  we changed the small io and geteattr parameters a bit, so this release is incompatible with previous ones.

merge of murali's changes from trunk.

[pcarns]: Patch to allow modifying distribution, distribution parameters and number of
data files to use within a given directory. Such setting will only impact creation of future files
within the same directory. It does not change existing files' parameters. All such hints
are inherited by newly created directories as well..
[pcarns]: Allow trove to continue to read a list of xattrs/keyvals
without stopping if some subset of keys dont exist in the keyval db. Modifies the keyval_read_list() function
to return an array of completely independant error codes for each key that is read.
[pcarns]: adjust encode_string() macro to make it safe to call on NULL strings
[pcarns]: bug in the getattr state machine that would cause it to read dirent_count most of the time
regardless of whether the attribute mask requested it or not.
[slang]: fix a couple of bugs in the get-attr server-side state machine whereby the dirent_count was being read
using the dirent handle instead of the dirdata handle. Fixed a return value from dbpf_keyval_get_handle_info_op_svc
so that get-attr would not needlessly hang..
[vilayann]: fixes for the viewdist utility yet again, since the keys for the xattrs have changed...

Reverse merge from trunk to lockd-branch...

reverse merge of branch

reverse merge of trunk to branch

Critical bugs exposed by Opteron experiments.
includes premature freeing of the state machine message pair array in readdirplus which were somehow not revealed
on x86!?
Added lib64 to openssl.m4 so that we also look for these libraries in the 64 bit path prior to the 32 bit paths.
If openssl was not present or not configured, make fskey_get return a NULL key and 0 keylength instead
of a random pointer that will cause us to crash with a seg fault..
All test module.mk had to be augmented with -lssl -lcrypto to get them to build on opterons.
Maximum number of directory entries (requested at 1 time) cannot be larger than maximum number of handles
that we can handle per server list attr request. Dont know how I can make this a compile-time test...
Fixed a few warnings in the test programs for Opterons.
Upshot is that with this patch, all getdents/getdents64/getdents_plus/getdents64_plus work correctly on Opterons to
the best of my knowledge on the sample workloads ran!!

Avery noticed we were using the worng size variable for sio requests

First cut of the sys_readdirplus implementation where clients can issue a single
system interface call to get the directory entries as well as the attribute information
of the directory entries.
What this means is that we do not need to call a sys_getattr of the individual directory entries
anymore.
(The kernel version is still untested and won't work until the VFS patch is in place! :))

On the server, there is now a new request/state machine called listattr(poor name, i know!)
which essentially does a vectored getattr for a list of handles as opposed to a getattr
which only does for 1. Perhaps, we need to consolidate them into 1 at a later point of time after
others have given this code a look-see!
On the client, there is a new system interface, and a corresponding statemachine implementation
called readdirplus which does the following algorithm
a) Step 1 - Do a regular readdir
b) Step 2 - Collate all meta handles (obtained from the readdir) and send out 1 listattr
to every server that owns atleast 1 handle from the set.
c) Step 3 - Handles that refer to metafiles (determined by the attributes of the response)
need an additional step to retrieve their file sizes. So we collate all data file handles
of all metafiles, and send out 1 listattr to every server that owns atleast 1 handle from
the set.
d) Step 4 - Copy object attributes obtained for all objects into the sys response structure and
also compute file sizes of regular files since we know the individual dfile handles' file sizes
from Step 3.

Clearly, this is a huge win in terms of performance of listing directories which have large number of directory
entries.
Next step is to hammer out the VFS patches and coreutils patch for ls to get the same benefit.
Also added a new utility called pvfs2-lsplus in src/apps/admin that uses this API.

added versioning test.  There's some nasty hacks with set/getenv to be able to test it properly without modifying sysint signatures.  Those will have to be removed at some point.

commit of versioning stuff to branch

up-ing protocol version from small io changes

merging changes from HEAD to dbpf branch

This fixes a bug where invalid sizes were being written on servers that were past the first strip.  It was created from the previous fixes to send the total_bytes to the server for the small_io operation (which are in still needed for writes).  The PINT_REQUEST_STATE_SET_FINAL macro actually needs the entire size of the request, not just the bytes sent to this server.

fix for the read request size being sent in small_io operation.  The displayed error this fixes is:  pint_request.c line 89: PINT_process_request: NULL segmax or bytemax!

fix alignment

fixes to small io request to calculate the size of the actual buffer being sent instead of the size of the memory request.

align geteattr req, small_io resp; small_io req alignment to be fixed by Murali with his patch

merging small I/O changes to trunk.  The majority of this commit includes the
new small I/O state machines on client and server, and the modifications to
the client IO state machine.  It also includes mods to the zero-fill code
to perform zero-fill of holes only if they exist instead of zero-ing the entire
memory region initially.  Also, some debug stuff has been added to the
state machine processing code to print enter/exit of state machines if sm
debug is enabled.  The genconfig now includes ioports/metaports for running
multiple servers on the same host for testing.

zero-fill appears to work now (finally).

fixed a bug in flow caused by outsize being updated correctly in trove now.  Also, s/transfered/transferred/.

*** empty log message ***

small I/O changes in sio branch for now

Based on Pete's observation, remove a superfluous skip4 in the listeattr response structure
since PVFS_ds_position is only 4 bytes long not 8 bytes.
Also removed the superfluous roundup8 macro in the listeattr request structure size calculations.

This patch adds a new system interface (listeattr) to list the extended attributes
of a file.
Also adds support in the VFS kernel module to support the listxattr
system call and a bunch of test programs to debug these.

Updated the FAQ to reflect usage of xattrs through VFS and acls through
VFS.

[pcarns]: add protocol versioning to PVFS2.  Major version changes are
incompatible up or down.  New clients can't talk to old servers, but new
servers can talk to old clients.

updates to my event changes to bring them inline with trunk

merge changes from trunk since branch point into branch.  Effectively moves the branchpoint to the latest trunk.

Murali's patch to add VFS support for extended attributes.

added extended attribute support

removed commented out older code version for final cleanup

cleanup for extended attribute stuff.  removed non-list server ops
in favor of list type and renamed list type.  cleaned up a few items in
request debug programs.

dded delete extended attribute feature

Updates for standardized spelling in pint-request.h and new distribution
function signatures.

updates for adding extended attributes

significant protocol changes to solve 64-bit alignedness issues

change the semantics of decode_skip4 so we can use it in endecode_fields entries
align PVFS_servreq_setattr
align PVFS_symlink_attr
align PVFS_metafile_attr_dfiles
align all strings to 8

begin first request in IO on 8-byte boundary

- fixed test program ls.c and admin program pvfs2-ls.c to not issue
  an extra call to readdir on every run by fixing a slight logic bug
- changed some readdir debugging
- added a directory version that is passed back from the server to
  the client on each successful readdir call (happens to be the
  directory's mtime encoded as an opaque uint64_t)
- updated client and server response structures and encoder to
  accomodate this
- NOTE: the versioning is not perfect since retrieving the directory
  mtime and iterating the handle list isn't an atomic operation with
  respect to the overall readdir op; i.e. there's still a window where
  a difference may not be detected.  then again, since the mtime is in
  seconds granularity, this should not present a major detectable
  problem anyway.  can use a different version number if problems
  arise often in practice

delete unused enum

fix comment

fix glaring alignment failure in PVFS_server_req

- added a PVFS_util method to get the current system and return it as
  a valid PVFS_time type
- added arguments to the crdirent/chdirent/rmdirent operations that
  now can take optional parent atime/mtime/ctime fields
- added support on server *dirent operations that update the parent
  time attributes (if specified) on successful completion of the
  *dirent operation; null/zero time fields are not updated
- modified sys-create/sys-mkdir/sys-symlink/sys-rename/sys-remove to
  work with the new operations properly; in general,
  create/remove/rename operations updates the mtime and ctime of the
  parent directory, which is now properly supported
- moved some common server side attribute handling code into a macro
  to make it re-useable across several server state machines
- updated en{de}coding of modified operations
- misc cleanups

- moved to fully 64 bit unsigned gossip masks
- setparam mgmt method now takes uint64_t values instead of int64_t
- added object type printing to acache debugging
- fixed bug in shared-state-machine getattr that could allow
  attributes to come out of the acache with incorrect attr
  types/fields
- fixed bug is msgpair setup for create/mkdir/symlink
- gossip mask change to uint64_t and many gossip cleanups
- misc cleanups

- added a client/server operation and mgmt method
  (PVFS_mgmt_get_dirdata_handle/PVFS_imgmt_get_dirdata_handle) for
  retrieving the dirdata handle of a specified 'parent' object (if
  any)
- fixed a crash bug (race) in trove that could return an object on
  queue addition that has already been serviced (since we've signaled
  the trove worker thread before returning the id in the first place);
  should not be a problem if we used safe_register for the ids, but
  otherwise not worth it

- fixed typo in the mgmt-remove-object request fill
- added some stubs for mgmt-remove-dirent

- use the new sm and stop relying on the old remove

- fixed bug in pint-dev's test method that didn't clear the outcount
  on error.  this caused the pvfs2-client-core to get short read
  errors, which were handled gracefully (by reposting the dev unexp),
  but it's better if it doesn't see them at all in that case (since
  there really was nothing read).
- started adding some debugging hooks for printing server response
  types (i only did getattr for now since that's the one that appears to
  be causing trouble)
- misc cleanups

- applied Nathan's expandtab patch (expandtab-2.patch.gz)

  first referenced at:
  http://www.beowulf-underground.org/pipermail/pvfs2-developers/2004-July/000745.html

start consolidating pvfs2 server request parameters into a single table to
make it easier to experiment with new requests; so far table only includes
op type, name, and permission requirements; will include more information later

- remove the extra free macros; all responses are freed by the decoder already

NOTE THAT THIS IS A BRANCH COMMIT (tagged as nm-nb-branch).  Feel free
to ignore it completey as it's a snapshot of a work in progress and it
will crash your computer and reformat your hard drive.

This is almost the initial draft of the pvfs2-client-core that
operates in a non-blocking manner.  While it runs, there are still
issues that need to be resolved (that are keeping it out of the main
CVS trunk).  Many other changes were made along the way, so it's more
than just that.

- added compile time option for disabling thread-safety in the client
  library (enabled by default; --disable-thread-safety to disable)
- improved configure summary information emitted at configure time
- added missing non-blocking sysint declarations to sysint header
- re-wrote pvfs2-client-core to use sysint non-blocking operations
  where possible
- made sysint test and testsome() calls more useable from a user point
  of view
- merged dev unexp polling/handling with system interface
- added PINT_sys_dev_unexp call that allows posting unexpected device
  messages so that they can be returned from the sysint testsome
  method in addition to completed sysint operations
- many memory leaks fixed -- many more to go (started adding macros
  for freeing the server response objects)
- added a id_gen_fast_unregister macro that is a no op, to make the
  api more consistent with the id_gen_safe_* calls
- server-config-mgr: report mutex still in use if it is, but also make
  sure not to unlock an already unlocked mutex (valgrind complains)
- many assertions added
- many formatting changes
- many ptr assignments to NULL after freeing in the job interface;
  done while tracking down a problem
- modified the pint-dev device interface to make sure it can handle
  the pvfs2 device in a non-blocking manner
- pint-dev was using buffers larger than it needed across the device;
  fixed them to be the right size
- added a method to free the memory region mapped into the kernel
  through the device
- freed that mapped memory region on pvfs2-client-core shutdown
  (valgrind complained)
- modified device driver to work properly in non-blocking mode from
  userspace by implementing the character device poll method
- modified pint-dev test function to properly handle non-blocking
  responses from the device driver
- modified PINT_flow_reset to not allocate a new mutex unless the old
  one was destroyed
- modified all job uses of the id-generator to use the safe, rather
  than fast, methods (useful for several reasons including safe
  cancellation)
- modified cancellation methods to be able to handle ops that have
  already completed
- modified the job_dev_unexp method to have (and honor) the no
  immediate completion flag if passed (used in the pvfs2-client-core)
- modified the job completion callbacks to make sure to NOT add a
  completed job desc to the completion queue if it's already been
  added (by checking a flag, not scanning).  this is a safety and
  should only be used when a non-thread-safe client library is being
  used in a thread safe env, but we should handle it gracefully anyway
- replaced all kernel allocations of ops through the slab allocator
  handled op_cache to be replaced by a wrapper method (op_alloc(), as
  suggested by Murali) -- the other cache allocations will probably be
  replaced later.  this allows the removal of the extern op_cache
  declaration
- added macros for freeing some of the most heavily used server
  response messages (readdir, lookup, getattr) -- the others are
  coming later
- freed dirents coming out of sysint response object in the
  pvfs2-client

Updates for distribution renamings.

Update for rename of PVFS_Dist to PINT_dist.

- added a chdirent server operation useful during renames of target files
  that already exist.
- still working on cleaning up rename, as it does not properly remove some
  data that is left to be removed

add a mechanism for a server to send a generic response back to the client
when it receives a request with a bad version number or encoding type, so
that the user gets a clear error message at the application level too

make PINT_server_get_config use msgpairs

add a server state machine (driven by a request scheduler timer) to check
for jobs that have timed out once per second; still not doing anything
because timeouts are set to inf. for now

remove spurious yet harmless semicolon

choose a limit on complex datatype (request) sizes and complain if more is demanded

convert request addresses to indices for sending on wire

- clean up msgpair error handling where we assert if the recv completes
  without errors before the send, but error out properly if we can't post
  the recv
- decrease max num dirent constant to avoid gm post failure due to a buffer
  size that is too large
[ thanks, phil for helping me find this stuff ]

- a complete re-write of the client side lookup operation that can properly
  handle '.' and '..' segments, as well as recursive relative/absolute
  symlinks across multiple metadata servers
- addition of a lookup debugging keyword, specific to the lookup operation
- minor changes to test programs to aid development
- added two shell scripts in the test/client/sysint directory to give a
  casual look at some trickier cases that lookup can handle (and to help
  make sure future versions can still handle)

- comment fix
- bump up max dirent count, for later readdir optimizations

This is a full merge of my pvfs2-dbpf-attr-cache-branch, so this is
the cumulative changelog (based on existing cvs messages).

- initial dbpf-attr-cache implementation and cleanups
- added all the hooks to grab configurable cacheable metadata keys
  from a config file.  re-arranged the config file by separating out
  fs specific hints into a StorageHints area.  updated genconfig to
  account for these changes.
- initial working implementation of the unified dbpf-attr and
  keyval/data pair cache.  holy cow this thing is fast... much work to
  be done still.
- initial locking added
- replaced debug level with a new DBPF_ATTRCACHE_DEBUG level
- added safe to/from cache data copies
- added write-through caching for keyval data (still need to do generic attrs)
- updated simple.conf
- allow config file to have multiple AttrCacheKeywords lines
- fix some cache parts that allow symlinks to work (i.e. be cached)
- added a verbose warning for a particular kind of configuration error
  for a user to see if it should happen
- some more robust configure checks
- more configure related changes
  disable vis build if sdl is detected, but sdl-ttf isn't
- fix broken ifdef nesting that caused compilation failure on non
  aio-threaded configurations
- safe copy of attributes (to/from, preparing for write-through impl)
- server cleanups (making way for pre-caching of a number of attrs on server
  startup since it reads them from disk during handle iterate anyway)
- update create test program to report human readable creation error
- put some quick human readable error reporting in pvfs2-client-core
- added some quickhash method we've been wanting but never had
- implemented policy to limit previously unbounded size of attr cache
- adjusted simple.conf to have some good defaults
- adjusted pvfs2-genconfig to match those defaults
- fixed some htable cleanups
- started write-through attr updates
- update changelog and version info, preparing for mainline merge
- implement write-through of dbpf-resize operation
- make sys-truncate return -PVFS_EINVAL if truncate is called on an object
  that can't be truncated
- fix error code return value of sys-truncate
- cosmetic changes to truncate test program output
- added some comments
- updated showconfig with copyright info
- updated sample configs for showconfig
- added a default handle purgatory value if a value of zero is specified
  (e.g. not specified in config file at all)
- for now, remove the cached dbpf-attr object on i/o writes
- slightly robustify the glibc aio callback detection

initial dbpf-attr-cache implementation and cleanups

Changed iod_num and iod_count to server_nr and server_ct.

Changed "int" to "int32_t".

bugfixes; cleanups; more lookup work -- now we support full symlink resolution
by default (link semantics will change later) up to an arbitrary depth of
8 links deep

add missing max-size on mgmt_event_mon encoding rule

remove debugging; better attr handling

first draft of mkdir implementation as a state machine.  it is used by default
in the tree now.  fixed an annoyance in the test create program -- now it uses
getuid/getgid for creating new files.

first draft sysint readdir implementation as a state-machine.  replaces
non-state-machine version by default.  added check for a special case in
the server's readdir that can skip a lot of work if no entries were requested
in a readdir operation.  fixed a bug in the readdir sysint test program that
called readdir 1 too many times in the general case.

add encoding rules for most request and response messages

added untested mgmt API hooks for retrieving server event logs

added stubs for server mgmt operation to retrieve event log

fix asymmetry on the one odd request type so that struct name, operation enum name, and item name match

filled in request protocol structures for new mgmt function to retrieve
low level dspace information, including encode/decode routines.
- untested
- still need true scheduler support

Fixed bug in macro that fills in create request so that dataspace types are now set correctly for metafiles and datafiles.

seperated/renamed variables/#defines that controls how many jobs we test for
at a time- number of initial unexpected jobs is now independent of the
number of jobs that we will test for at any given time...

gave setparam the ability to report back the original value of the parameter
for each server, not utilizing anywhere yet

filled in client side hooks for iterate_handles operation, untested

stubbed in server operation for iterate handles

added flags to request protocol to allow requests that modify the file
system to proceed when servers are in admin mode

fix up the encoding of perf monitor messages a little, sanity check sizes

more incremental progress on performance monitoring

stubbing in more components of the perf monitoring framework

started implementing server request that will retrieve performance
statistics using a mgmt operation

added a state machine that runs continuously in the back ground to
manage performance statistics (doesn't actually do anything yet), along
the way added support for starting state machines that are not
triggered by protocol requests

started on an admin tool that shows statfs-ish type stuff, not finished
yet

implemented mgmt function to retrieve statfs style information from
each independent server

added a new PVFS_statfs structure (incomplete- just a few fields to
start with), implemented trove_collection_getinfo() to retrieve this
information from file system, hooked into server statfs implementation.
completely untested, and no client side support yet

. landing a new and improved client-side truncate.
. completing the landing of a fully-functioning server-side truncate.

added statfs request type, and a trivial stub for server state machine

pushed the decimal values for major,minor, and sub version numbers
into pvfs2-config.h and used as a basis for protocl version number

finally added sys-symlink.sm (thought I had before).  more changes to get
it closer to working.

Fill in the create implementation as a state machine.  Proper pcache hooks
should now be in place as well, where applicable.  Areas that need work are
rolling back in error cases and usage of the dcache.

scattered server cleanups; most notable change is the dynamic allocation
of the job structures (if less than a hard max)

Put remaining code in place to handle memory datatypes, not ready yet
(pending a fix to some corner case in req processing).  Everything should
still work as it was prior to this commit.  When we do enable memory
datatypes, I'll need to go back and uncomment some code and adjust all
PVFS_sys_{read/write} calls to pass in the right datatype (they are just
taking NULL for now).  PVFS_sys_{read/write} API is in its final form
(I think...) for now, at least.  For now this code is dependent upon
the fact that the file datatype isn't implicitly tiled yet, need to get
memory datatypes going before relaxing that constraint...

changed reference to datatypes called "io_req..." to "file_req...", to make
the distinction clearer between file datatypes and memory datatypes once the
latter is added

started work on a mgmt function that can send a noop to an arbitrary
server for diagnostic purposes

started adding hooks for a NOOP request

Removed all references to PVFS_SYS_IO_{READ,WRITE}, instead using PVFS_IO_xxx versions exclusively.

put a few pieces in place for mgmt_setparam operation, hopefully located
strategically so as not to conflict on merge with the flush...

this is going to cause a lot of warnings until the rest of my flush changes are landed, but start teaching more parts of pvfs2 about flush

header shuffling- you probably have to do a distclean to build after
updating

some tweaks to the server getattr state machine to make the attribute mask more accurate (i hope).  also more of the client io
state machine.

initial support for "flush"

client-side remove state machine fully implemented.  needs major cleanup next.

client-side remove state machine can remove datafiles now too.

removed unused header file.
more progress on the client-side state machine front.

Pulled out some common functionality into functions and macros to help with reuse.

mkdir is now mostly functional, while crdirent is broken.  Some misc cleanups
as well.

Made necessary changes for sending handle ranges over to the server.  This
required changes to the trove_dspace_create call, so a lot of stuff had to
change to make this compile.  mkdir is broken and may be for a while.

Server receives handle ranges now for creates, but doesn't do anything with
them yet.  That's to come.

started on a new protocol encoder (little endian bytefield encoding).  still
very rought and untested, but need to check in to test some other stuff.
also added a #define indicating the maximum request number in protocol.

started setting up generic encoding header to prepend all messages; not
actually used yet

went ahead and pushed through the addition of a file datatype offset at
sysint, server, and request protocol level.  This stuff needed to happen
regardless of how we actually implement the feature at the lowest level.

Untested!

more sanity checking of request protocol boundaries, mostly on server side.
Also elected to remove config buffer size field from getconfig request, and
instead have that boundary dictated by protocol headers

more #define / limits cleanup; biggest change is the addition of
PVFS_SEGMENT_MAX to limit the size of each path segment

started defining limits on some the variable sized fields within the
request protocol; not enforcing yet

cleaned up some header file inclusions

Piles of changes:
- removed rsize field from request protocol (useless in memory representation,
  duplicate of BMI reported size at wire level, used inconsistently
  anyway)
- fixed some contig encoder bugs (failed to add 4 byte trailer to some
  encodings when it wasn't convenient)
- fixed some unrelated bugs in the job handling path of client side I/O
  operations
- added some memory leaks in contig encoder, will fix later...

more header file cleanup, this time io req/resp and comments

minor cleanup, and removal of a pile of unused/undefined structures

cleaned up the getconfig req/resp a little

readdir req/resp cleanup

got rid of the "generic" response structure, cleaned up several response
structure definitions

cleaned up lookup_path response structure- in particular made the size of
the handle and attribute arrays explicit

cleaned up setattr request definition a bit

cleaned up getattr request a bit, removed old geteattr structures

cleaned up the remove request a bit, removed old batch structures

cleaned up create request definition a bit

got rid of some extra typedefs, cleaned up some naming conventions

removed some unused statfs structures.  We'll figure out what we need here
later.

minor change, explicit enumeration

re-ordered the request type numbers, changed the way the state machine table is intialized on the server side, and removed another MIA request type

removed some server request types that aren't actually implemented

very minor cleanup of req proto header

discarded some unused extended attribute structures

next phase of attribute cleanup:
- added mask field to obj attribute structure
- remove seperate mask value from pinode struct and several functions,
  including sysint functions
- made server store mask value properly and report it back- even though it
  still ignores which mask fields are set in getattr and setattr requests
- for now, made it so that number of datafiles _cannot_ be specified at
  sys_create time by setting fields in the attributes structure directly.
  Until this issue is resolved, there is currently no way to pick the
  number of datafiles at create time- it is always chosen by system.
- misc. other related cleanup

found another field (this time in the create request) called
object_type, changed it to type PVFS_ds_type instead of int and
updated code to use object type values for it instead of attribute
mask values.

got rid of the PVFS_token type and replaced with PVFS_ds_position where
appropriate

got rid of PVFS_count32 type; replaced with int32_t or uint32_t or int
where appropriate

got rid of the PVFS_count64 type

got rid of the PVFS_bitfield type, and replaced with something more
generic everywhere it was used

renamed enum flowproto_type to enum PVFS_flowproto_type and moved to
pvfs2-types.h

got rid of PVFS_string type and replaced with char* everywhere

step one of header cleanup: ran pvfs2-mangle-in-place-80col.sh on these
files to get the code format closer to standard

Replaced many instances of the bucket/mask stuff with just handle stuff.

put in hooks to manually specify which flowprotocol to use, tested for
a few trial cases

Some more rough getconfig cleanups.  Removed the fs_name from the request
since we no longer do a per-fs getconfig query.

Another big set of getconfig changes.  A bit more polished than the last
revision, but not yet perfect.

Server configuration structure now caches the config files in memory to
avoid having to check the disk each time a getconfig is done.

This is a *terribly* intrusive patch that may break everything.  I'm checking
it in now as I'm hesitant to wait on this until tomorrow, but rest assured I
will test and fix loose ends tomorrow.  It works for me currently, with the
only caveat being that when you start the server, you MUST specify the
full path name to the fs.conf file (i.e. simple.conf)

Report bugs loudly anyway, so I'll know what to attack tomorrow.

'rmdir' no longer a supported operation.  instead, 'remove' will remove
directory objects in addition to datafile and metafile objects.

Big patch.  The I/O path is broken in this- I'm in the middle of
changing the protocol for I/O operations.  A new response has been
added to announce when writes are complete on the server side, new
functions have been added for receiving arrays of acks on the
client side, other client side req processing functions have been
cleaned up, buffer verification has been turned on in the I/O test
(with some modifications), several other changes.

completed io state machine and tested some

added hooks in protocol and in io test program to send io request and
distribution information; still need support in request encoder/decoder,
though

added some fields to the I/O request structure that I think I might
need

filled in a little more of the steps needed for an I/O operation (up
to the point of checking the size of the bstream to operate on).
Realized that I was missing some job functionality at this point...

minor cleanup, added data file size to io ack

cleaned a lot of old junk out; mostly outdated i/o description
structures

added in job code; added in request processor code where empty
files were accidentally checked in last time

added request scheduler and request protocol headers into new tree