CVS log for emstar/fusd/BUILD
	Help

Request diff between arbitrary revisions / Display revisions graphically / Jump to LXR

fixing 26 kernel build for nims-stargate etc

added support for autobuilding of 2.6 kernel

minor fixes to make shlibs compile

* moved fusdnet out of fusdd and into a separate process
* added --enable-remote option into misc_init that waits for fusdnet
* added fusdnet() macro for running fusdnet when desired
* moved logging in fusdd_server over to elog
* fixed bug in link.run

lots of changes to support the 802.11 ceiling array
* emcee deprecated in favor of emsim --ceiling
* ceiling configuration now done through .sim file
* proxy-host directive enables connection via fusdnet to remote host
* hostmote runs either locally as part of the node, or if proxy-host
  set, connects via fusdnet to another server
* mote devices now are part of the /dev/sim/group... hierarchy
* new --no-ceiling directive added to channel model to prevent it
  from being run when in ceiling mode

Big commit -- the bulk of the Cygwin compatibility is being checked in here.
The Cygwin port doesn't really run yet since FUSD hasn't been ported, but,
with this checkin, the rest of the repository compiles.

oops -- took out too many libraries

use BUILD_KMODULES variable

fixed other copy of dse... wish there were not two copies

various uninit vars fixed and changes to BUILD files in prep for shared
library support

take out openrace until it works with fusdd

accidentally checked in commented out kfusd..

took openrace back out

added openrace back in.. doesn't hang fusdd anymore

Changed build file to take advantage of the new simple_targets statement

took out old target

* some fixes to the previously-added fusdnet support in libfusd.c
* reorg and first working version of FUSDnet
* FUSDnet is now officially part of the fusdd build

more tweaking of regression tests

added FUSD regression tests

build new fudsnet stuff..

Bug fix to fusdd, in some cases it would delete and then immediately
re-create devices when it got a new status, causing races.

Used advanced new C-comment technology in the BUILD file

client sends connection message to server now.

added small utility to read binary version of the fusd status dev

* major checkin of fusdnet work.

* fusd API changes
  * fusd_context_t instead of g_event_t
  * added option to disable fusdnet functionality
  * propagated changes throughout libdev etc
  * changes to fusd_msg to support new fusdnet message type

* audio server
  * got rid of stupid rflink junk
  * fixed fusd api change

Added fusd/fusdd/tcp_server - a very simple server that decodes fusd_msg_t
messages and provides a listen socket to test the fusdd_client.c code.

Changed fusdd_client so that it creates a fusd-device instead of a
packet-device.

please don't check in build files that build experimental code

Took out segment in command_dev_open that kills a command-device if
a usage call-back is instantiated, but returns a NULL pointer.

Added code to fusdd_client.c - in order to create a :
dev/fusd/net device which accepts commands to connect to remote hosts.
Once it creates a socket and connection, it creates a new packet-device
(the name is provided in the connection call to the cmd-device)
which accepts commands to the host and will automatically send whatever
is written to the packet-device to the remote host.

Added code to emtos to translate dbg commands in TinyOS to elog commands.

Minor fixes to error messages, added todo list..

* Modified FUSD to support unlimited numbers of devices in non-devfs mode
* Modified FUSD status protocol to support diffs with a 10-second refresh

FUSD no longer depends on DevFS!!!

TO use this feature, compile on a system with a non-devfs kernel,
insert the module, and run the fusdd user daemon.  Otherwise it
works identically to devfs fusd.

This also involves a change to the fusd status output, so old fusd's
are not compatible with the new fusd.

Added an additional case where forging a close might be necessary.
Was sometimes causing processes not to see balanced open/close pairs.
Mainfested by emrun thinking that a process had not yet died when it had,
but it simply hadn't gotten the close() notification for the log or
control device.

** FINALLY FIXED BUG CAUSING PROCESSES TO GET STUCK IN D STATE -- in
fusd_client_poll, there was a call to LOCK_FUSD_FILE followed by
LOCK_FUSD_DEV.  The LOCK_FUSD_DEV call might jump out if the dev was
a zombie, leaving the file locked.  WHOO HOO!!!!

** Changed (slightly) semantics of callback returns: changes to
device_info made in fusd_file_info are no longer honored.  device_info
can ONLY be set when the device is registered.

** Various other cleanups in handling of zombies -- zombie checks are
now never performed unless the device is locked.  A lot of "zombie_dev"
goto targets are, as a result, gone.  But I think this is a good thing,
as they can be confusing (see first entry in changelog!)

Checked in the epic bugfix to finally remove what I think are the very last
of the SMP-unsafe problems: the infamous race between clients opening and
unregistration.  Passes a whole slew of regression tests, including one
that was able to consistently deadlock the old code.

final conversion away from old Makefile

build the kernel modules

more fixes...... does this actually WORK???????

many more changes; hopefully got a handle on the recursive dep problem
between .d files and the directories that they sit in.

started the process of converting the code itself to conform to the new
build system (mainly fully-qualifying include filenames)

minor bugfix and fusd now compiles

more checkpointing -- relative dirs work, the basic structure of bootstrapping
everything from the root directory works, more black magic with gcc -MM
stuff, objdir tokens integrated, concept of 'native build' on which the
easybuild system itself runs.... lots and lots of details.  but then
again, is a build system really anything BUT lots and lots of details?

new build system actually builds FUSD now, with a little help

change comment