One of the merits of my fork is that I clearly documented what works
and what does not unlike other repos I've seen so far.

Therefore, you can immediately see on the readme that Raytrace and
Vips didn't work and how they failed.

My setup might work for native QEMU aarch64 compile, but I haven't
tested. But if someone does, send a pull request, I also want that to
work. I did test native x86 compilation and got almost everything
working there (not even that was working upstream).

The bottom line is that Princeton is not responding anymore, so the
only way parrsec can ever work is if the gem5 community forks it and
maintains it, which is what I'm doing.

I tried to use Bamos' repo first, and I would rather contribute to it
rather than have my own fork, but he is not responding either
apparently: https://github.com/bamos/parsec-benchmark/issues/3 and
last commit is very old.

The other option would be ARM's RSK
https://github.com/arm-university/arm-gem5-rsk but I doubt that they
have made everything work either based on their patches, and they
haven't documented what works and what does not, and it is not nice to
contribute there since they don't have an issue tracker and you'd have
to patch their patches.

At least on my side, I guarantee that if anyone sends me an issue or
pull request on my fork, that I will try to reproduce it or review it,
and merge, be it for native or cross compilation.


On Sat, May 12, 2018 at 8:14 PM, Tung Hoang <hoangthanht...@gmail.com> wrote:
> You can compile parsec by using either cross compiler or qemu and add binary
> benchmark into linaro image.
>
> I am not sure if Ciro’s parsec can compile all benchmark for aarch64 but I
> could not. At least Raytrace and Vips are not compiled with patches.
>
> /T
>
> On Sat, May 12, 2018 at 9:00 AM commerce _com <commercecom...@gmail.com>
> wrote:
>>
>> hi Ciro;
>>
>> so i downloaded parsec which is in the link you sent me but did i copy it
>> directly into the linaro arch64 image or i add the patch for arch64 i did
>> not understand well ??
>> because I need to copy parsec 3.0 in the linaro image to run with gem5.
>>
>> thanks;
>>
>>
>> 2018-05-12 17:28 GMT+02:00 Ciro Santilli <ciro.santi...@gmail.com>:
>>>
>>> I don't think those patches are enough to run all benchmarks, and I
>>> have never seen a set of patches that just worked for all benchmarks so
>>> far.
>>>
>>> I have part of parsec benchmarks working at:
>>>
>>> https://github.com/cirosantilli/linux-kernel-module-cheat/tree/d1f676a680aca90e525f14d90605393a9742d09b#parsec-benchmark
>>> and I'm clearly documenting what works and does not at:
>>> https://github.com/cirosantilli/parsec-benchmark
>>>
>>> On Sat, May 12, 2018 at 3:17 PM, commerce _com <commercecom...@gmail.com>
>>> wrote:
>>> > hi all;
>>> >
>>> > I need to compile parsec 3.0 with Cross-Compile:
>>> >
>>> > here are the steps that I followed:
>>> >
>>> > I downloaded gcc-linaro-5.5.0-2017.10-x86_64_aarch64-linux-gnu and I
>>> > modified the xcompile-patch.diff file as follows:
>>> >
>>> >    # CC_HOME is root installation of the C compiler
>>> > - export CC_HOME = "/ usr"
>>> > + export CC_HOME = "/ media / ali / ali / parsec-3.0 /
>>> > gcc-linaro-5.5.0-2017.10-x86_64_aarch64-linux-gnu"
>>> >    # BINUTIL_HOME is root installation of the GNU binutils
>>> > - export BINUTIL_HOME = "/ usr"
>>> > + export BINUTIL_HOME = "/ media / ali / ali / parsec-3.0 /
>>> > gcc-linaro-5.5.0-2017.10-x86_64_aarch64-linux-gnu / aarch64-linux-gnu"
>>> >    # GNUTOOL_HOME is root installation of the GNU tools
>>> >    export GNUTOOL_HOME = "/ usr"
>>> >    # BINARY_PREFIX is the string which is used as prefix for the GNU
>>> > binaries
>>> >    export BINARY_PREFIX = ""
>>> > + # GCC_PREFIX is the string which is used as prefix for the compilers
>>> > and
>>> > preprocessors
>>> > + export GCC_PREFIX = "aarch64-linux-gnu-"
>>> >
>>> > I typed the command:
>>> > patch -p1 <xcompile-patch.diff
>>> >
>>> > I found the following result:
>>> > patching file config / gcc.bldconf
>>> > patching file pkgs / apps / blackscholes / parsec /
>>> > gcc-pthreads.bldconf
>>> > patching file pkgs / apps / bodytrack / parsec / gcc-pthreads.bldconf
>>> > patching file pkgs / apps / facesim / parsec / gcc-pthreads.bldconf
>>> > patching file pkgs / apps / ferret / parsec / gcc-pthreads.bldconf
>>> > patching file pkgs / apps / fluidanimate / parsec /
>>> > gcc-pthreads.bldconf
>>> > patching file pkgs / apps / raytrace / parsec / gcc-pthreads.bldconf
>>> > patching file pkgs / apps / swaptions / parsec / gcc-pthreads.bldconf
>>> > patching pkgs file / apps / vips / parsec / gcc-pthreads.bldconf
>>> > patching pkgs file / apps / x264 / parsec / gcc-pthreads.bldconf
>>> > patching file pkgs / kernel / canneal / parsec / gcc-pthreads.bldconf
>>> > patching file pkgs / kernel / dedup / parsec / gcc-pthreads.bldconf
>>> > patching file pkgs / kernels / streamcluster / parsec /
>>> > gcc-pthreads.bldconf
>>> > patching file pkgs /libs/glib/parsec/gcc.bldconf
>>> > patching file pkgs /libs/gsl/parsec/gcc.bldconf
>>> > patching file pkgs / libs / hooks / parsec / gcc.bldconf
>>> > patching file pkgs / libs / hooks / src / hooks.c
>>> > patching pkgs file / libs / libjpeg / parsec / gcc.bldconf
>>> > patching file pkgs /libs/libxml2/parsec/gcc.bldconf
>>> > patching file pkgs / libs / mesa / parsec / gcc.bldconf
>>> > patching file pkgs / libs / parmacs / parsec / gcc.bldconf
>>> > patching file pkgs /libs/ssl/parsec/gcc.bldconf
>>> > patching file pkgs /libs/tbblib/parsec/gcc.bldconf
>>> > patching file pkgs / libs / uptcpip / parsec / gcc.bldconf
>>> > patching file pkgs /libs/zlib/parsec/gcc.bldconf
>>> > patching file pkgs / tools / cmake / parsec / gcc.bldconf
>>> > patching file pkgs / tools / libtool / parsec / gcc.bldconf
>>> > patching file pkgs / tools / yasm / parsec / gcc.bldconf
>>> >
>>> > after I typed the following commands:
>>> >
>>> > export PARSECPLAT = "aarch64-linux"
>>> > source ./env.sh
>>> > parsecmgmt -a build -c gcc-hooks -p blackscholes
>>> >
>>> > and I found this error is in the last line:
>>> >
>>> > [PARSEC] Packages to build:  parsec.blackscholes
>>> >
>>> > [PARSEC] [========== Building package parsec.blackscholes [1]
>>> > ==========]
>>> > [PARSEC] [---------- Analyzing package parsec.blackscholes ----------]
>>> > [PARSEC] parsec.blackscholes depends on: hooks
>>> > [PARSEC] [---------- Analyzing package parsec.hooks ----------]
>>> > [PARSEC] parsec.hooks depends on: libtool
>>> > [PARSEC] [---------- Analyzing package parsec.libtool ----------]
>>> > [PARSEC] parsec.libtool does not depend on any other packages.
>>> > [PARSEC] [---------- Building package parsec.libtool ----------]
>>> > [PARSEC] Running 'env LIBS=
>>> > /media/ali/ali/parsec-3.0/pkgs/tools/libtool/src/configure
>>> > --build=x86_64-linux-gnu --host=aarch64-linux-gnu
>>> >
>>> > --prefix=/media/ali/ali/parsec-3.0/pkgs/tools/libtool/inst/aarch64-linux.gcc-hooks':
>>> > ## ------------------------- ##
>>> > ## Configuring libtool 2.2.6 ##
>>> > ## ------------------------- ##
>>> >
>>> > checking for a BSD-compatible install...
>>> > /media/ali/ali/parsec-3.0/bin/install -c
>>> > checking whether build environment is sane... yes
>>> > checking for a thread-safe mkdir -p... /bin/mkdir -p
>>> > checking for gawk... no
>>> > checking for mawk... mawk
>>> > checking whether /usr/bin/make sets $(MAKE)... yes
>>> > checking for aarch64-linux-gnu-strip...
>>> >
>>> > /media/ali/ali/parsec-3.0/gcc-linaro-5.5.0-2017.10-x86_64_aarch64-linux-gnu/aarch64-linux-gnu/bin/strip
>>> > checking whether subdir libobjs are useable... yes
>>> > checking for aarch64-linux-gnu-gcc...
>>> >
>>> > /media/ali/ali/parsec-3.0/gcc-linaro-5.5.0-2017.10-x86_64_aarch64-linux-gnu/bin/aarch64-linux-gnu-gcc
>>> > checking for C compiler default output file name... a.out
>>> > checking whether the C compiler works... yes
>>> > checking whether we are cross compiling... yes
>>> > checking for suffix of executables...
>>> > checking for suffix of object files... o
>>> > checking whether we are using the GNU C compiler... yes
>>> > checking whether
>>> >
>>> > /media/ali/ali/parsec-3.0/gcc-linaro-5.5.0-2017.10-x86_64_aarch64-linux-gnu/bin/aarch64-linux-gnu-gcc
>>> > accepts -g... yes
>>> > checking for
>>> >
>>> > /media/ali/ali/parsec-3.0/gcc-linaro-5.5.0-2017.10-x86_64_aarch64-linux-gnu/bin/aarch64-linux-gnu-gcc
>>> > option to accept ISO C89... none needed
>>> > checking for style of include used by /usr/bin/make... GNU
>>> > checking dependency style of
>>> >
>>> > /media/ali/ali/parsec-3.0/gcc-linaro-5.5.0-2017.10-x86_64_aarch64-linux-gnu/bin/aarch64-linux-gnu-gcc...
>>> > gcc3
>>> > checking whether
>>> >
>>> > /media/ali/ali/parsec-3.0/gcc-linaro-5.5.0-2017.10-x86_64_aarch64-linux-gnu/bin/aarch64-linux-gnu-gcc
>>> > and cc understand -c and -o together... yes
>>> > checking how to run the C preprocessor...
>>> >
>>> > /media/ali/ali/parsec-3.0/gcc-linaro-5.5.0-2017.10-x86_64_aarch64-linux-gnu/bin/aarch64-linux-gnu-cpp
>>> > checking build system type... x86_64-pc-linux-gnu
>>> > checking host system type... Invalid configuration `aarch64-linux-gnu':
>>> > machine `aarch64' not recognized
>>> > configure: error: /bin/bash
>>> >
>>> > /media/ali/ali/parsec-3.0/pkgs/tools/libtool/src/libltdl/config/config.sub
>>> > aarch64-linux-gnu failed
>>> > [PARSEC] Error: 'env LIBS=
>>> > /media/ali/ali/parsec-3.0/pkgs/tools/libtool/src/configure
>>> > --build=x86_64-linux-gnu --host=aarch64-linux-gnu
>>> >
>>> > --prefix=/media/ali/ali/parsec-3.0/pkgs/tools/libtool/inst/aarch64-linux.gcc-hooks'
>>> > failed.
>>> >
>>> >
>>> > THANKS;
>>> >
>>> >
>>> > _______________________________________________
>>> > gem5-users mailing list
>>> > gem5-users@gem5.org
>>> > http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
>>> _______________________________________________
>>> gem5-users mailing list
>>> gem5-users@gem5.org
>>> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
>>
>>
>> _______________________________________________
>> gem5-users mailing list
>> gem5-users@gem5.org
>> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
_______________________________________________
gem5-users mailing list
gem5-users@gem5.org
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users

Reply via email to