[gem5-users] Reg v21.2 release
Hi, When is gem5 v21.2 scheduled for release? Looking forward to use some of the enhancements that go in it. Thanks Sindhuja ___ gem5-users mailing list -- gem5-users@gem5.org To unsubscribe send an email to gem5-users-le...@gem5.org %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
[gem5-users] Error: Can't find a working Python installation
Hi Community, I get Error:Can't find a working Python installation with option "-with-asan" whereas no error when the option is not included. Why is this so? I use GEM5 v21.1.0.1. Error : [root@sindhuja-5520 gem5]# scons -j8 --with-asan scons: Reading SConscript files ... Checking for linker -Wl,--as-needed support... yes Warning: To suppress false positive leaks, set the LSAN_OPTIONS environment variable to "suppressions=/root/dse/gem5/util/lsan-suppressions" when running gem5 Warning: LSAN_OPTIONS=suppressions=suppressions=/root/dse/gem5/util/lsan-suppressions Checking for compiler -gz support... yes Checking for linker -gz support... no Warning: Can't enable executable debug section compression Info: Using Python config: python3-config Checking for C header file Python.h... yes Checking Python version... no Error: Can't find a working Python installation No Error: [root@sindhuja-5520 gem5]# scons -j8 scons: Reading SConscript files ... Checking for linker -Wl,--as-needed support... yes Checking for compiler -gz support... yes Checking for linker -gz support... yes Info: Using Python config: python3-config Checking for C header file Python.h... yes Checking Python version... 3.6.8 Checking for accept(0,0,0) in C++ library None... yes Checking for zlibVersion() in C++ library z... yes Checking for C library tcmalloc... yes Thanks & Regards Sindhuja ___ gem5-users mailing list -- gem5-users@gem5.org To unsubscribe send an email to gem5-users-le...@gem5.org %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
[gem5-users] Re: Query: Valgrind speed in FS
Hi Jason, Thanks for the suggestions. I tried --without-tcmalloc and used option --suppressions. Still this is so very slow. If I am able to get this working, I will update here. @Gabe: Thanks for your comments Thanks & Regards Sindhuja From: Jason Lowe-Power Sent: Tuesday, September 7, 2021 3:28 PM To: gem5 users mailing list Cc: Sindhuja Gopalakrishnan Elango Subject: Re: [gem5-users] Query: Valgrind speed in FS Hi Sindhuja, Yes, there is an expectation that valgrind causes a slowdown. Let me give you a couple of suggestions. 1. Make sure you compile without tcmalloc (e.g., scons build//gem5.opt --without-tcmalloc). Using tcmalloc will make valgrind miss all allocations. 2. Use the suppressions file in util/valgrind-suppressions. This will hide most of the python "errors". It was recently updated to suppress many more errors that aren't really errors. That said, I've had some recent problems myself with valgrind and the new suppressions file being *very* slow. I think this is something we need to look into. If you have any ideas on how to improve the performance, we would appreciate hearing them! Otherwise, I guess we'll all just have to keep waiting overnight ;) Cheers, Jason On Tue, Sep 7, 2021 at 2:20 PM Sindhuja Gopalakrishnan Elango via gem5-users mailto:gem5-users@gem5.org>> wrote: Hi Community, I ran into bad_alloc issues in GEM5 with the full system simulation of SPEC 2006 benchmarks. Suspecting a memory leak, I wanted to use valgrind to understand better. Without valgrind option, it takes less than 10 minutes for kernel to boot and also run the 400.perlbench/attrs benchmark. But with valgrind, it has crossed 2 hrs and the kernel hasn’t yet booted. Usage: valgrind --log-file=attrs.val.txt --error-limit=no $GEM5_CMD I would like to know if this slowdown is reasonable with valgrind? And do you have any suggestions for a memory leak detection tool that is faster than valgrind and works well with gem5? Appreciate your time and effort. Thanks much. Best Regards Sindhuja ___ gem5-users mailing list -- gem5-users@gem5.org<mailto:gem5-users@gem5.org> To unsubscribe send an email to gem5-users-le...@gem5.org<mailto:gem5-users-le...@gem5.org> %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s ___ gem5-users mailing list -- gem5-users@gem5.org To unsubscribe send an email to gem5-users-le...@gem5.org %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
[gem5-users] Query: Valgrind speed in FS
Hi Community, I ran into bad_alloc issues in GEM5 with the full system simulation of SPEC 2006 benchmarks. Suspecting a memory leak, I wanted to use valgrind to understand better. Without valgrind option, it takes less than 10 minutes for kernel to boot and also run the 400.perlbench/attrs benchmark. But with valgrind, it has crossed 2 hrs and the kernel hasn't yet booted. Usage: valgrind --log-file=attrs.val.txt --error-limit=no $GEM5_CMD I would like to know if this slowdown is reasonable with valgrind? And do you have any suggestions for a memory leak detection tool that is faster than valgrind and works well with gem5? Appreciate your time and effort. Thanks much. Best Regards Sindhuja ___ gem5-users mailing list -- gem5-users@gem5.org To unsubscribe send an email to gem5-users-le...@gem5.org %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
[gem5-users] Re: Issue: Checkpoint in FS mode
Hi Pedro, Thanks for taking the time to reply. You're right. The implementation was missing in my own python script. After including that, I am able to generate the checkpoint. Thanks & Regards Sindhuja -Original Message- From: Pedro Becker via gem5-users Sent: Thursday, September 2, 2021 3:57 AM To: gem5-users@gem5.org Cc: Pedro Becker Subject: [gem5-users] Re: Issue: Checkpoint in FS mode Hi Sindhuja, Are you using your own python script to configure the simulation? If I'm not mistaken calling the /sbin/m5 checkpoint exits the m5.simulate() (from the script), returning the exit_event (a string "checkpoint"). Then, your python script should capture that exit_event and create the checkpoint. This is done in the default scripts (e.g., fs.py/Simulation.py) with something like this: exit_event = m5.simulate() exit_cause = exit_event.getCause() while exit_event == "checkpoint": m5.checkpoint(joinpath(cptdir, "cpt.%d")) exit_event = m5.simulate() exit_cause = exit_event.getCause() But if you have your own python script, then you have to implement this behavior yourself. Best, Pedro. ___ gem5-users mailing list -- gem5-users@gem5.org To unsubscribe send an email to gem5-users-le...@gem5.org %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s ___ gem5-users mailing list -- gem5-users@gem5.org To unsubscribe send an email to gem5-users-le...@gem5.org %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
[gem5-users] Issue: Checkpoint in FS mode
Hi, I would like to create a checkpoint in Full-System simulation before running a benchmark. For that, in the start of the boot script (/etc/gem5init) I have included the below lines. #!/bin/sh +x # Checkpoint the first execution echo "Checkpointing simulation..." /sbin/m5 checkpoint As soon as the "/sbin/m5 checkpoint" line is executed, the simulation is immediately terminated and there is no checkpoint generated in m5out directory. I have used AtomicSimpleCPU and gem5.opt executable. /sbin/m5 readfile, /sbin/m5 writefile options in the same boot script seem to work fine. Why is /sbin/m5 checkpoint generation not working in this set up? Appreciate your help. Thanks. Best Regards Sindhuja ___ gem5-users mailing list -- gem5-users@gem5.org To unsubscribe send an email to gem5-users-le...@gem5.org %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
[gem5-users] Re: SPEC INT 2006 memory requirements
Hi Chandrahas, Thank you for the suggestion. I tried with the guest system of 3 GB and host system of 12 GB. Still the same issue. Will it help if the gem5 executable is compiled with flag –without-tcmalloc? Thanks & Regards Sindhuja From: Chandrahas Tirumalasetty Sent: Friday, August 27, 2021 3:10 PM Cc: Sindhuja Gopalakrishnan Elango Subject: Re: [gem5-users] SPEC INT 2006 memory requirements Hello Sindhuja, The memory allocation will depend on the system you are simulating. If you are simulating a system with 8GB of memory, the gem5 process in your system will try to malloc() 8GB from the host. Depending on the memory availability, your host kernel might outright reject the memory allocation, which is happening in your case. Maybe if you have 12GB in your system, try simulating a system with 6GB memory, it should work. On Fri, Aug 27, 2021 at 4:46 PM Sindhuja Gopalakrishnan Elango via gem5-users mailto:gem5-users@gem5.org>> wrote: Hi, What is the guest system memory requirement for running SPEC INT 2006 benchmarks in full-system simulation mode? The kernel that I use has a 3 GB RAM and 4GB swap space. The host system has 12 GB RAM and 4GB swap space. I tried with both atomic and timing CPU’s and used gem5.fast as well as gem5.opt binaries. But, most of the benchmarks from the reference set gets killed with the above setting. src/central_freelist.cc:333] tcmalloc: allocation failed 8192 gem5 has encountered a segmentation fault! However for the same SPEC binaries, system call emulation seems to be just fine. Can anyone help here? Thanks & Regards Sindhuja ___ gem5-users mailing list -- gem5-users@gem5.org<mailto:gem5-users@gem5.org> To unsubscribe send an email to gem5-users-le...@gem5.org<mailto:gem5-users-le...@gem5.org> %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s -- Chandrahas Tirumalasetty PhD student at Dept. of ECE, TAMU mail: chandrahas...@tamu.edu<mailto:chandrahas...@tamu.edu> ___ gem5-users mailing list -- gem5-users@gem5.org To unsubscribe send an email to gem5-users-le...@gem5.org %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s
[gem5-users] SPEC INT 2006 memory requirements
Hi, What is the guest system memory requirement for running SPEC INT 2006 benchmarks in full-system simulation mode? The kernel that I use has a 3 GB RAM and 4GB swap space. The host system has 12 GB RAM and 4GB swap space. I tried with both atomic and timing CPU's and used gem5.fast as well as gem5.opt binaries. But, most of the benchmarks from the reference set gets killed with the above setting. src/central_freelist.cc:333] tcmalloc: allocation failed 8192 gem5 has encountered a segmentation fault! However for the same SPEC binaries, system call emulation seems to be just fine. Can anyone help here? Thanks & Regards Sindhuja ___ gem5-users mailing list -- gem5-users@gem5.org To unsubscribe send an email to gem5-users-le...@gem5.org %(web_page_url)slistinfo%(cgiext)s/%(_internal_name)s