replacing umlaute and fixing address in headers Haldeneggsteig -> Universitaetsstrasse

Signed-off-by: Reto Achermann <reto.achermann@inf.ethz.ch>

flounder: implementing support to pass caps over UMP connection for endpoints

Signed-off-by: Reto Achermann <reto.achermann@inf.ethz.ch>

Refactor process manager to enable spawnd discovery.

The process manager now allocates a special LMP endpoint for monitor.0 and
gives said endpoint to monitor.0 as part of the export_cb. After monitor.0
spawns spawnd, it uses that endpoint to send the new spawnd's iref to the
process manager. The latter then binds with spawnd, to keep track of its state.

For app cores, a similar workflow applies, except inter-monitor communication
is now performed. Namely, the new app core's monitor sends the new app core's
spawnd to monitor.0, which forwards it to the process manager. The latter then
binds with the app core's spawnd just the same.

Diagrams for the discovery protocol can be found at https://goo.gl/eJE37u.

Signed-off-by: Razvan Damachi <razvan.damachi@gmail.com>

T276: Fix monitor revoke check to check for overlapping descendants instead of any descendants

Signed-off-by: Simon Gerber <simon.gerber@inf.ethz.ch>

Adding maximum size to variable arrays in flounder definitions. Converting
one TArray to normal arguments.

Signed-off-by: Adam Turowski <adam.turowski@inf.ethz.ch>

monitor: remove unnecessary log2ceil and 2^ for span_domain_request()

T273: monitor: cleanup variable and argument names for cspace level

Signed-off-by: Simon Gerber <simon.gerber@inf.ethz.ch>

Interface definitions: Changing calls/responses into messages.

Signed-off-by: Adam Turowski <adam.turowski@inf.ethz.ch>

T191: update monitor capops retype

Signed-off-by: Simon Gerber <simon.gerber@inf.ethz.ch>

T24: remove remaining SCC code

Signed-off-by: Moritz Hoffmann <moritz.hoffmann@inf.ethz.ch>

Made errval_t a flounder builtin.

The errval type is used internally by flounder, but has since been duplicated
to be passed around through interfaces by user code. This led to a
proliferation of ifname_errval_t types, which were used interchangably in
user-level code. This breaks once strict typedef checking is enforced, as in
recent GCC versions. This is a bandaid fix, and there's a larger underlying
issue still to be resolved around how to deal with globally-defined types in
flounder.

Signed-off-by: David Cock <david.cock@inf.ethz.ch>

Made errval_t a flounder builtin.

The errval type is used internally by flounder, but has since been duplicated
to be passed around through interfaces by user code. This led to a
proliferation of ifname_errval_t types, which were used interchangably in
user-level code. This breaks once strict typedef checking is enforced, as in
recent GCC versions. This is a bandaid fix, and there's a larger underlying
issue still to be resolved around how to deal with globally-defined types in
flounder.

Signed-off-by: David Cock <david.cock@inf.ethz.ch>

multihop: fix cap ownership tranfer over multihop channel.

Signed-off-by: Simon Gerber <simon.gerber@inf.ethz.ch>

Remove no longer needed message handlers.

Added moving kcb off shared cpu driver, currently crashes in newly started kernel on target core.

x86boot can give kcb from a core to core 0.

invoke by calling: x86boot auto give 1 0
Give kcb to other cores than 0 currently untested, and probably does not yet
work.

Working on a multihop power-down method for remote cores.

Working on power-down over monitor.

Remove Beehive intermon messages.

Beehvive support was removed a long time ago.

Remove Beehive intermon messages.

Beehvive support was removed a long time ago.

Conflicts:
if/intermon.if

Make x86_32 compile; some of the monitor invocations are disabled right now.

Documentation for core boot-up.

Added and clarified some comments for the intermonitor communication that
happens when a new core is booted in the system.

Formatting intermon.if interface file.

Time seems to have ravaged it.

Rewrote distributed revoke and delete handling.

Added "retrieve" operation for revoke.

Implemented support for "give_away" transfers in capops_copy.

Added layer to encapsulate transferring caps across intermon.

Added individual remote relations bits for copies/ancestors/descendants.

Do not engage monitors in distributed cap ops until mem alloc is setup.

First stage of integrating "copy" into ump cap transfer.

Changed copy result to cnode & slot.

Added valid bits param to cap copy result.

Made capops state type more opaque in intermon.if.

Added capop messages to intermon.if.

Reindented intermon.if.

New tracing infrastructure

This is a Distributed Systems Lab project by
Alexander Grest <agrest@student.ethz.ch> and David Stolz <stolzda@ethz.ch>

Added replication for ramfs iref among monitors.

--HG--
branch : octopus.chips-remove

simplify out-of-band cap transfer to avoid asynchronously reporting errors back to the sender

As discussed on the mailing list (10.01.2012 "Proposed change to
out-of-band cap transfer"), this change changes the semantics so that
an error in delivery of the cap is reported to the receiver (rather
than asynchronously to the sender) along with a NULL capref. This
simplifies the implementation within the monitor, and results in saner
error handling on both sides.

remove support for Beehive

see the mailing list discussion at the end of December 2011 for motivation

improvements/clean-up to multihop routing table management

* remove pointless ack messages from monitor back to route_setup
* general cleanup of rts.c internal logic
* added functionality to grow the routing table through a given
forwarding core through a new set of cores (this is an initial
kludge for very simple heterogeneous setups like MIC)
* removed assumption of num_cores - 1 == max_coreid from monitor
(it's still present in route_setup)

(this is a short-term improvement to the existing code; it doesn't
fix any of the real issues we identified on the mailing list)

general cleanup of monitor internals

* start to abtract management of bindings to other monitors
(in preparation for multihop support)

* move code out of header files -- if the inlining is such a big
win, we can put it back later, but it seemed excessive

* fix a lot of irritating misnamed variables and
fields: most "closure" objects, and a lot of "st" pointers,
were in fact pointers to bindings

* deprecated num_cores_request message

* removed pingpong code from intermon interface

* removed some dead code

there should be no changes to runtime behaviour or APIs/interfaces
outside the monitor

--HG--
rename : usr/monitor/iref.h => usr/monitor/iref.c

use typedefs in multihop code

rename MAX_CPUS to MAX_COREID, and increase it to 254 (to match an 8-bit coreid_t)

* Kludged around the limitations of coremask_t enough to keep rcap_db_null compiling.
* Removed the config option that allowed users to lower it. this
doesn't make much sense any more, unless we are going in for really
silly compile-time optimisations. This change requires updating
Config.hs

Remove old routing/group communication library.

As discussed briefly on the list (22.{02,12}.2011), the library was
unused since early this year, and has been deprecated because it:
* is unable to interface to flounder/THC
* assumes that a set of core IDs can be expressed in a single-word bitmap
* doesn't work on machines with large (> 32?) numbers of cores (but
this is probably just a bug?)

Since we have the new multihop code in the tree, and seem to have
decided that this is an architectural dead-end, I'm removing this code
in the interest of least confusion as we start building functionality
on the new multihop code.

The one major side-effect of this change is that it is no longer
possible to compile the non-NULL remote cap management database
implementations (rcap_db_central, rcap_db_twopc). The code remains in
the tree for future resurrection, however.

Apply the multihop patch by alexander grest based on his bachelor thesis

Initial version of public stable barrelfish repository.