replacing umlaute and fixing address in headers Haldeneggsteig -> Universitaetsstrasse

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

libbarrelfish: heap: explicitly store and propagate heap pagesize instead of relying on heap alignment to infer pagesize

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

Fix VA layout so that thread slabs are <4GB

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

Always use frame_identify() instead of raw invocation

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

T191: x86_64: frame_identify returns size in bytes

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

T153: avoid code duplication when setting slot allocator

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

32bit arm: fix large page support. Close T148.

This implements a bit field to track the mapping of individual L2 tables for
armv5 and armv7 user space page table management. Having this tracking
bitfield we can use 1MB sections without having weird interference between
unused L2 tables and creating new section mappings.

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

morecore: fixing leaking slot allocator.(closes T153

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

32bit arm: fix large page support. Close T148.

This implements a bit field to track the mapping of individual L2 tables for
armv5 and armv7 user space page table management. Having this tracking
bitfield we can use 1MB sections without having weird interference between
unused L2 tables and creating new section mappings.

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

libbarrelfish: vspace_mmu_aware: fix frame_alloc() retry condition.

Fix retry condition when connection to memory server not established yet.

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

Fix large page code so it compiles on 32bit / ARM.

Note: the ARM version does not support large pages properly at this point.

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

libbarrelfish: fix morecore heap with 1GB pages.

This fixes the vspace_mmu_aware vregion to actually allocate 1GB pages if
requested through morecore=0x40000000 on the command line.

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

mmu aware vspace: added sane fallback for basecn allocations.

Signed-off-by: Simon Gerber <simon.gerber@hp.com>

vspace_mmu_aware: fixed edge case in assertions.

Assertions didn't allow freeing the region starting at offset zero in the
vregion.

Signed-off-by: Simon Gerber <simon.gerber@hp.com>

Added remapping functionality for malloc.

Signed-off-by: Simon Gerber <simon.gerber@hp.com>

make mmu_aware vspace code support multiple page sizes.

Signed-off-by: Simon Gerber <simon.gerber@hp.com>

fix heap memory reclamation bug which usually manifested as an apparent out-of-memory condition in spawnd

remove support for Beehive

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

Reverted the changeset:88 which was breaking the networked application startup.

Fixed memory reclamation leak. Need to update consumed field.

Fixed two memory reclamation problems:

1. Need to update the consumed field, otherwise vspace library leaks memory.

2. Have every domain keep up to 1MB of free memory around as fast reserve,
before returning to memory server. This circumvents a problem with
asynchronous memory reclamation: If memory is allocated and freed in a
tight loop, allocating memory happens faster than it will be freed
(freeing is an asynchronous operation). This commit works on the
assumption that allocating and then freeing more than 1MB in a tight loop
is not something one does regularly. Also, keeping a little bit of memory
as local reserve should speed up fine-grain memory management.

Merge from ASPLOS tree.

Initial version of public stable barrelfish repository.