[gem5-users] --script does not work
Hi all, I currently face a problem with --script option, that is , the full disk system never executes my script. I have checked that from the checkpoint I restored, the system user is root, and my script has been chmoded with 755. My first command is to do another checkpoint, but it never works. Any reason this could be ? Many thanks, My command is set as below: /home/mao/gem5Research/build/X86/gem5.opt /home/mao/gem5Research/configs/example/fs.py --disk-image=/home/mao/diskimage/disks/ubuntu-16.04_BENCH.img --kernel=/home/mao/diskimage/binaries/VM-4.14-newcopy --checkpoint-dir=/home/mao/gem5Research/scriptMY/newbatch/SPEC_CHECKPOINTS/pmeM1 --mem-size=16GB --cpu-type=AtomicSimpleCPU --mem-type=DDR4_2400_4x16 --l3cache --caches -n 8 -r 2 --script /home/mao/gem5Research/scriptMY/newbatch/parabatch/pmeM1.rCS > /home/mao/gem5Research/scriptMY/newbatch/SPEC_CHECKPOINTS/pmeM1/T5 -- Mao Ye, ___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
Re: [gem5-users] How to attach multiple disk images a simulation with gem5 fs.py?
With fs.py, I had it working at some point with a small ugly hack to FSConfig.py I think. It is on another machine so I cannot check right now. I will probably make a cleaner way of doing this in the near future. However, the same method did not apply to a custom full-system script I am working on, so I am still investigating. On Thu, Jun 14, 2018 at 7:23 PM Ciro Santilli wrote: > At https://www.mail-archive.com/gem5-users@gem5.org/msg15233.html > Jason mentioned the technique of attaching two disk images to restore > a checkpoint and run a different benchmark. > > Is it possible to specify multiple disk images when using fs.py? From > the source code I don't think it is currently supported, but just > double checking before I decided to patch it or not. > > It seems that multiple --disk-image= options just overwrite one another. > > gem5 60600f09c25255b3c8f72da7fb49100e2682093a > ___ > 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
Re: [gem5-users] How to attach multiple disk images a simulation with gem5 fs.py?
Thanks for confirming, Feel free to CC me directly if you end up submitting a patch. On Fri, Jun 15, 2018 at 1:37 PM, Timon Evenblij wrote: > With fs.py, I had it working at some point with a small ugly hack to > FSConfig.py I think. It is on another machine so I cannot check right now. > I will probably make a cleaner way of doing this in the near future. > > However, the same method did not apply to a custom full-system script I am > working on, so I am still investigating. > > On Thu, Jun 14, 2018 at 7:23 PM Ciro Santilli > wrote: > >> At https://www.mail-archive.com/gem5-users@gem5.org/msg15233.html >> Jason mentioned the technique of attaching two disk images to restore >> a checkpoint and run a different benchmark. >> >> Is it possible to specify multiple disk images when using fs.py? From >> the source code I don't think it is currently supported, but just >> double checking before I decided to patch it or not. >> >> It seems that multiple --disk-image= options just overwrite one another. >> >> gem5 60600f09c25255b3c8f72da7fb49100e2682093a >> ___ >> 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
Re: [gem5-users] How to change the number of data message flits size + garnet
Yes you can change these values. The flit size is essentially the link width, and its default is 16B (128b), defines jn src/mem/ruby/network/garnet2.0/GarnetNetwork.py src/mem/ruby/network/Network.py defines the default control message size which is 8B (64b). The default data message size is in src/mem/ruby/network/network.cc and is control msg size + cache block size , so (8+64)B = 72B Accordingly the number of control flits and data flits get computed. You can change any of these to model the number of flits per packet that you are interested in. Note — if you increase the number of control / data flits, you may need to increase the control and data VC depths accordingly in src/mem/ruby/network/garnet2.0/GarnetNetwork.py Cheers, Tushar On Jun 15, 2018, 2:36 AM -0400, F. A. Faisal , wrote: Dear All, In the default garnet configuration, the default value of ni_flit_size = 16 results in control messages fitting within 1 flit, and data messages fitting within 5 flits. However, I would like to increase the number of data message flit size to 10 flits or the control message size to 5 flits. [In addition, I am also unsure about changing ni_flit_size to any value, can change the actual number of flit size in data packets or the control packets] Please let me know the possible configuration to perform such changes. Thanks in advance for your kindness. Thanks and best regards, F. A. Faisal - School of Information Science, JAIST Doctoral Student ___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
Re: [gem5-users] How to attach multiple disk images a simulation with gem5 fs.py?
Hi Ciro, I'm not sure if you can do it with fs.py (I bet not) :). When you set the ide controller up (e.g., pc.south_bridge.ide.disks) it takes a python list of length 2 of disk images. Cheers, Jason On Fri, Jun 15, 2018 at 6:43 AM Ciro Santilli wrote: > Thanks for confirming, > > Feel free to CC me directly if you end up submitting a patch. > > On Fri, Jun 15, 2018 at 1:37 PM, Timon Evenblij > wrote: > >> With fs.py, I had it working at some point with a small ugly hack to >> FSConfig.py I think. It is on another machine so I cannot check right now. >> I will probably make a cleaner way of doing this in the near future. >> >> However, the same method did not apply to a custom full-system script I >> am working on, so I am still investigating. >> >> On Thu, Jun 14, 2018 at 7:23 PM Ciro Santilli >> wrote: >> >>> At https://www.mail-archive.com/gem5-users@gem5.org/msg15233.html >>> Jason mentioned the technique of attaching two disk images to restore >>> a checkpoint and run a different benchmark. >>> >>> Is it possible to specify multiple disk images when using fs.py? From >>> the source code I don't think it is currently supported, but just >>> double checking before I decided to patch it or not. >>> >>> It seems that multiple --disk-image= options just overwrite one another. >>> >>> gem5 60600f09c25255b3c8f72da7fb49100e2682093a >>> ___ >>> 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
Re: [gem5-users] Gem5 - McPAT, fs_big.little.py Stats.txt to XML
Hi George, in order to make this work you need to start with a proper template that can handle the bigLITTLE architecture. Then from stats.txt you will get the proper numerical values that are associated with you simulation (i.e. utilization factors) and from the gem5 config json file you'll get several microarchitectural info that have to be passed also to the template xml. So your work starts primarily with McPAT in trying to compile a proper template that most resembles the system that you are trying to simulate in gem5. Then the script is not that difficult. You can also try our conversion scripts that I think handle some cases (e.g. atomic cpus) that the conversion you mentioned does not handle. Check it here: https://github.com/H2020-COSSIM/cMcPAT I repeat though, that your work starts with the proper template file. Another option if you want to avoid McPAT is a tool developed by Chalmers University in the same project (COSSIM) that models the power consumption of ARM bigLITTLE according to the instruction mix. I think the most relevant paper is this one: https://ieeexplore.ieee.org/abstract/document/8106950/ and you can contact the authors for more as for all I know they have not released the code in a public repository. Best, Andreas On Thu, Jun 14, 2018 at 5:06 PM, Georgios S. Bousdras < georgios.bousd...@ulb.ac.be> wrote: > Hi, > > I want to measure the power of a fs_bigLITTLE system using McPAT. > Is there any method to convert the stats.txt to the proper xml file? > I tried this script https://github.com/Ayymoose/gem5-mcpat-parser but It > gives me error at the template.xml. > Is it possible to convert it manually ? > If you know a different program expect McPAT let me know. > > Best Regards, > George > ___ > 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] How to change the number of data message flits size + garnet
Dear All, In the default garnet configuration, the default value of *ni_flit_size = *16 results in control messages fitting within 1 flit, and data messages fitting within 5 flits. However, I would like to increase the number of data message flit size to 10 flits or the control message size to 5 flits. [In addition, I am also unsure about changing ni_flit_size to any value, can change the actual number of flit size in data packets or the control packets] Please let me know the possible configuration to perform such changes. Thanks in advance for your kindness. Thanks and best regards, F. A. Faisal - School of Information Science, JAIST Doctoral Student ___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users