Go to file
Andreas Hansson d1f3a3b91a gcc: Enable Link-Time Optimization for gcc >= 4.6
This patch adds Link-Time Optimization when building the fast target
using gcc >= 4.6, and adds a scons flag to disable it (-no-lto). No
check is performed to guarantee that the linker supports LTO and use
of the linker plugin, so the user has to ensure that binutils GNU ld
>= 2.21 or the gold linker is available. Typically, if gcc >= 4.6 is
available, the latter should not be a problem. Currently the LTO
option is only useful for gcc >= 4.6, due to the limited support on
clang and earlier versions of gcc. The intention is to also add
support for clang once the LTO integration matures.

The same number of jobs is used for the parallel phase of LTO as the
jobs specified on the scons command line, using the -flto=n flag that
was introduced with gcc 4.6. The gold linker also supports concurrent
and incremental linking, but this is not used at this point.

The compilation and linking time is increased by almost 50% on
average, although ARM seems to be particularly demanding with an
increase of almost 100%. Also beware when using this as gcc uses a
tremendous amount of memory and temp space in the process. You have
been warned.

After some careful consideration, and plenty discussions, the flag is
only added to the fast target, and the warning that was issued in an
earlier version of this patch is now removed. Similarly, the flag used
to enable LTO, now the default is to use it, and the flag has been
modified to disable LTO. The rationale behind this decision is that
opt is used for development, whereas fast is only used for long runs,
e.g. regressions or more elaborate experiments where the additional
compile and link time is amortized by a much larger run time.

When it comes to the return on investment, the regression seems to be
roughly 15% faster with LTO. For a bit more detail, I ran twolf on
ARM.fast, with three repeated runs, and they all finish within 42
minutes (+- 25 seconds) without LTO and 31 minutes (+- 25 seconds)
with LTO, i.e. LTO gives an impressive >25% speed-up for this case.

Without LTO (ARM.fast twolf)

real	42m37.632s
user	42m34.448s
sys	0m0.390s

real	41m51.793s
user	41m50.384s
sys	0m0.131s

real	41m45.491s
user	41m39.791s
sys	0m0.139s

With LTO (ARM.fast twolf)

real	30m33.588s
user	30m5.701s
sys	0m0.141s

real	31m27.791s
user	31m24.674s
sys	0m0.111s

real	31m25.500s
user	31m16.731s
sys	0m0.106s
2012-09-14 12:13:22 -04:00
build_opts Regression: Add a test for x86 timing full system ruby simulation 2012-04-25 22:43:36 -05:00
configs se.py Ruby: Connect TLB walker ports 2012-09-12 21:42:57 -05:00
ext clang/gcc: Fix compilation issues with clang 3.0 and gcc 4.6 2012-04-14 05:43:31 -04:00
src gcc: Enable Link-Time Optimization for gcc >= 4.6 2012-09-14 12:13:22 -04:00
system ARM: Fix issue with with way MPIDR is read to include affinity levels. 2012-09-07 14:20:53 -05:00
tests Stats: Remove the reference stats that are no longer present 2012-09-13 08:02:55 -04:00
util ARM: Fix the compiler and platform identification for building on ARM. 2012-09-07 14:20:53 -05:00
.hgignore .hgignore: added src/doxygen 2010-07-27 20:00:38 -07:00
.hgtags Added tag stable_2012_06_28 for changeset f75ee4849c40 2012-07-19 16:53:02 -07:00
COPYING copyright: Add code for finding all copyright blocks and create a COPYING file 2011-06-02 17:36:07 -07:00
LICENSE copyright: Add code for finding all copyright blocks and create a COPYING file 2011-06-02 17:36:07 -07:00
README Info: Clean up some info files. 2011-02-14 21:36:37 -08:00
SConstruct gcc: Enable Link-Time Optimization for gcc >= 4.6 2012-09-14 12:13:22 -04:00

This is the M5 simulator.

For detailed information about building the simulator and getting
started please refer to http://www.m5sim.org.

Specific pages of interest are:
http://www.m5sim.org/wiki/index.php/Compiling_M5
http://www.m5sim.org/wiki/index.php/Running_M5

Short version:

1. If you don't have SCons version 0.98.1 or newer, get it from
http://wwww.scons.org.

2. If you don't have SWIG version 1.3.31 or newer, get it from
http://wwww.swig.org.

3. Make sure you also have gcc version 3.4.6 or newer, Python 2.4 or newer
(the dev version with header files), zlib, and the m4 preprocessor.

4. In this directory, type 'scons build/ALPHA_SE/tests/debug/quick'.  This
will build the debug version of the m5 binary (m5.debug) for the Alpha
syscall emulation target, and run the quick regression tests on it.

If you have questions, please send mail to m5-users@m5sim.org

WHAT'S INCLUDED (AND NOT)
-------------------------

The basic source release includes these subdirectories:
 - m5:
   - configs: simulation configuration scripts
   - ext: less-common external packages needed to build m5
   - src: source code of the m5 simulator
   - system: source for some optional system software for simulated systems
   - tests: regression tests
   - util: useful utility programs and files

To run full-system simulations, you will need compiled system firmware
(console and PALcode for Alpha), kernel binaries and one or more disk images. 
These files for Alpha are collected in a separate archive, m5_system.tar.bz2.
This file can he downloaded separately.

Depending on the ISA used, M5 may support Linux 2.4/2.6, FreeBSD, and the
proprietary Compaq/HP Tru64 version of Unix. We are able to distribute Linux
and FreeBSD bootdisks, but we are unable to distribute bootable disk images of
Tru64 Unix. If you have a Tru64 license and are interested in
obtaining disk images, contact us at m5-users@m5sim.org