replacing umlaute and fixing address in headers Haldeneggsteig -> Universitaetsstrasse

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

T24: remove remaining SCC code

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

Removed RCK_EMU and updated SCC tech note.

Added local loopback device

Fixed a bug in webserver which was causing problem in NFS loading of files.
There still another bug which surfaces in handle_tx_done due to spp_index
off by one with read_reg. Also the pending_tx buffers tries to become
less than zero. The way to test this bug is to mirror the website once
the webserver starts. When mirroring the website second time, it should
show the related print statements

Fix in bulk transfer mechanism concerning a full queue (from producer view)
Without this fix sp_set_read_index wont succeed if the queue is full.
---
lib/procon/procon.c | 8 ++++++--
1 files changed, 6 insertions(+), 2 deletions(-)

Fixed a bug which was causing httperf to crash.
There was a problem of using index without wrapping it with size of
circular buffer.

Minor changes to get the merge working

Fixed some indentation issues in ether_control.if
and modified procon.c to fix the race condition that Antonie had seen

Removed the additional ring from packet send path. The driver code for e1000n
and e10k is modified but code for other drivers (eMAC, and rtl) is still to
be modified.

slimmed down the usage of shared_queue in receive and send path
Also, modified analyze_traces.py to accommodate antoine's benchmarking

Added profiling code for latency in connection with e10k.

Removed the use of disp_name() from debugging in procon.c to avoid compilation
issues

Fixed the issue with loading large files over NFS.
The problem was mainly due to the way notifications from the driver were
handled by the application. Now, every notification from driver triggers
both handle_tx_done and handle_incoming_packets.

The NFS is still slow, but most probably due to the use of inefficient pbuf
type PBUF_RAM

Fixed some compilation problems and modified the use of functions
inside asserts as asserts are not compiled in production level deployments
(harness deployments)

Removed most of the debugging code which was breaking the compilation for arm

Cleaned up the debug print statement to reduce the noise in boot process.

Minor improvements to speedup the RX benchmark, added more code
for measurements.

Changes to get RX benchmark working.
There are lot of small changes in this commit and also extra code for debugging
and performance evaluation which will be cleaned in future versions.

Fixed the problem of random retransmits happening in send path.
The problem was in the way tx_notify was handled and the assumption that
hardware might return notification out-of-order.

Partially working version with two bugs
1. Assertion failure about "Slot not clear" due to the unclear logic in
sp_process_tx_done
2. Same packet getting delivered twice and hence generating Unknown XID warning
in rpc.c code

Mostly working version of new bulk transport

Cleaned up most of the debugging code in procon and mem_barrelfish
Still there is a debugging code in ethersrv and also, mem_free is still wedging

Fixed the memory overwriting problem. Now the netd is getting the IP
but pbuf free is going crazy. Also I need to cleanup all the debugging code
added to find the memory overwriting problem.

Found temporary fix for the assertion failure in pbuf initialization code
Somehow, adding slots into shared_pool is somehow leading to the changes
in the variable buffer_list and corrupting it.

Fixed some issues with benchmark to make it more suitable as TX benchmark

Fixed some subtle bugs in procon library

reverted the changes to get the working state of network stack

Added code to deal with tx_done communication with spp
but this code is broken. Need more debugging to get it working

Minor changes in procon library to fix few minor issues

Sending packets to driver from applications is working with procon library
The library has become little complicated and needs some cleanup, but
good part is that, it is working.

Added code to map the shared_pool memory into driver

shared_pools are getting registered with buffers
Now, need to load them properly into driver address space.

Added code in procon library to create shared frames

Added a library procon for producer/consumer based shared buffer management.
Also, added the library in the Hakefiles of all applications which needs it.