8723b08dbf
Transaction Level Modeling (TLM2.0) is widely used in industry for creating virtual platforms (IEEE 1666 SystemC). This patch contains a standard compliant implementation of an external gem5 port, that enables the usage of gem5 as a TLM initiator component in SystemC based virtual platforms. Both TLM coding paradigms loosely timed (b_transport) and aproximately timed (nb_transport) are supported. Compared to the original patch a TLM memory manager was added. Furthermore, the transaction object was removed and for each TLM payload a PacketPointer that points to the original gem5 packet is added as an TLM extension. For event handling single events are now created. Committed by: Nilay Vaish <nilay@cs.wisc.edu>
97 lines
3.1 KiB
Plaintext
97 lines
3.1 KiB
Plaintext
This directory contains a demo of a coupling between gem5 and SystemC-TLM.
|
|
It is based on the gem5-systemc implementation in utils/systemc.
|
|
First a simple example with gem5's traffic generator is shown, later an full
|
|
system example.
|
|
|
|
Files:
|
|
|
|
main.cc -- demonstration top level
|
|
sc_port.{cc,hh} -- transactor that translates beween gem5 and tlm
|
|
sc_mm.{cc,hh} -- implementation of a tlm memory manager
|
|
sc_ext.{cc,hh} -- a TLM extension that carries the gem5 packet
|
|
sc_target.{cc,hh} -- an example TLM LT/AT memory module
|
|
tlm.py -- simple gem5 configuration
|
|
tgen.cfg -- configuration file for the traceplayer
|
|
|
|
Other Files will be used from utils/systemc example:
|
|
|
|
sc_logger.{cc,hh},
|
|
sc_module.{cc,hh},
|
|
sc_gem5_control.{cc,hh},
|
|
stats.{cc,hh}
|
|
|
|
|
|
I. Traffic Generator Setup
|
|
==========================
|
|
|
|
To build:
|
|
|
|
First build a normal gem5 (cxx-config not needed, Python needed).
|
|
Second build gem5 as a library with cxx-config support and (optionally)
|
|
without python.
|
|
|
|
> cd ../..
|
|
> scons build/ARM/gem5.opt
|
|
> scons --with-cxx-config --without-python build/ARM/libgem5_opt.so
|
|
> cd util/systemc_tlm
|
|
|
|
Set a proper LD_LIBRARY_PATH e.g. for bash:
|
|
> export LD_LIBRARY_PATH="$LD_LIBRARY_PATH:/path/to/gem5/build/ARM/"
|
|
|
|
Then edit the Makefile to set the paths for SystemC and run make
|
|
|
|
> make
|
|
|
|
Make a config file for the C++-configured gem5 using normal gem5
|
|
|
|
> ../../build/ARM/gem5.opt ./tlm.py
|
|
|
|
The message "fatal: Can't find port handler type 'tlm'" is okay.
|
|
The configuration will be stored in the m5out/ directory
|
|
|
|
The binary 'gem5.opt.sc', that has been created in the make step,
|
|
can now be used to load in the generated config file from the previous
|
|
normal gem5 run.
|
|
|
|
Try:
|
|
|
|
> ./gem5.opt.sc m5out/config.ini -e 1000000
|
|
|
|
It should run a simulation for 1us.
|
|
|
|
To see more information what happens inside the TLM module use the -D flag:
|
|
|
|
> ./gem5.opt.sc m5out/config.ini -e 1000000 -D
|
|
|
|
To see more information about the port coupling use:
|
|
|
|
> ./gem5.opt.sc m5out/config.ini -e 1000000 -d ExternalPort
|
|
|
|
II. Full System Setup
|
|
=====================
|
|
|
|
Build gem5 as discribed in Section I. Then, make a config file for the
|
|
C++-configured gem5 using normal gem5
|
|
|
|
> ../../build/ARM/gem5.opt ../../configs/example/fs.py --tlm-memory=memory \
|
|
--cpu-type=timing --num-cpu=1 --mem-type=SimpleMemory --mem-size=512MB \
|
|
--mem-channels=1 --caches --l2cache --machine-type=VExpress_EMM \
|
|
--dtb-filename=vexpress.aarch32.ll_20131205.0-gem5.1cpu.dtb \
|
|
--kernel=vmlinux.aarch32.ll_20131205.0-gem5 \
|
|
--disk-image=linux-aarch32-ael.img
|
|
|
|
The message "fatal: Can't find port handler type 'tlm'" is okay.
|
|
The configuration will be stored in the m5out/ directory
|
|
|
|
The binary 'gem5.opt.sc' can now be used to load in the generated config
|
|
file from the previous normal gem5 run.
|
|
|
|
Try:
|
|
|
|
> ./gem5.opt.sc m5out/config.ini -o 2147483648
|
|
|
|
The parameter -o specifies the begining of the memory region (0x80000000).
|
|
The system should boot now.
|
|
|
|
For conveniance a run_gem5.sh file holds all those commands
|