[gem5-users] Re: Dumping stats into individual files

Hi, I don't know whether gem5 has an option for that but you can use the standard library's "rename" function each time you call m5_dump_stats as a quick hack. Best, Muhammet Subhankar Pal via gem5-users , 1 Mar 2021 Pzt, 05:59 tarihinde şunu yazdı: > Hi, > > I am using gem5 se.py to run some

[gem5-users] Re: Out of Memory while running GPU Benchmark

memory usage show 30305260 KBytes ~ 29 GB ? > What does this value indicate ? > > Thank you, > Sampad > > On Mon, Sep 14, 2020 at 3:30 PM Muhammet Abdullah Soytürk < > muhammetabdullahsoyt...@gmail.com> wrote: > >> Hi, >> >> Is there any chance that th

[gem5-users] Re: Out of Memory while running GPU Benchmark

Hi, Is there any chance that the input you provide is bigger than 2GB? If the input size is bigger than the memory size, you cannot simulate it in SE mode (since there is no paging support for SE mode). As you can understand from the error message, you need to increase the size of the memory.

[gem5-users] Re: Problem with gem5 memory latency

Hi Ahmed, You could either follow Abhishek's advice and use default configuration scripts (se.py or fs.py) or you can create your own script to manipulate those options. Writing your own configuration script is recommended (You can check this book to learn how to

[gem5-users] Re: Stat dump after each N instructions - CPU and/or AMD GPU

I don't know how legal is this and whether it has side effects but you might try schedStatEvent to schedule the dumps. Muhammet Matt Sinclair via gem5-users , 27 May 2020 Çar, 00:13 tarihinde şunu yazdı: > I'm not sure if

[gem5-users] Simulation OpenMP Applications with se.py

Hi all, I am trying to simulate an openmp application with 4 cores in SE mode. It seems to be working fine when I run it with this command: build/RISCV/gem5.opt configs/example/se.py -c --caches --l2cache --cpu-type=DerivO3CPU -n 4 but while I was searching I have seen that there was

[gem5-users] Correct Way to Fast-Forward

Hi all, I want to fast forward my benchmark during its input graph initialization phase and then switch to O3 cpu. I have a couple questions about that. - My system is a multicore system (let's say 4 cores). Should I fast forward with 4 atomic cpus and restore with 4 O3 cpus or one cpu for

Re: [gem5-users] How to get the high-level architectural view of the system I am simulating?

Hi Rubel, The command you execute does not simulate the system that you put in the image. There is no l2 bus and l2 cache in the system that you simulate. If you want to have l2 cache you need to add --l2cache as an argument to the script. In order to see what's on the system you execute, you

Re: [gem5-users] Problem passing a SimObject as Parameter

Hi Francisco, Apparently gem5 cannot register your object to object list. With the information you gave, I can only guess. - Did you register your object in the SConscript file? - Did you recompile gem5.opt after you created your SimObject? While you compiling the gem5 binary with scons

[gem5-users] Page Table Fault When Accessing Virtual Address

Hi all, I am trying to run a simulation with se.py. It works perfectly fine on my machine but when I try to run it on a workstation I am getting this error: panic: Page table fault when accessing virtual address 0x7f8ff900 I am using RISCV isa and the latest version of gem5. This is

[gem5-users] Sysmalloc assertion

Hi all, I wrote a script that executes several gem5 simulations. It works perfectly on my laptop but when I clone it to the workstation I am getting this error: pr(This is the applications name): malloc.c:2389: sysmalloc: Assertion `(old_top == initial_top (av) && old_size == 0) || ((unsigned

Re: [gem5-users] 回复： Installing Shared Libraries to Image in Full Simulation

e I am doing ssh into the machine and I ssh >>>>>>>>>>>>>>> using "ssh -Y" >>>>>>>>>>>>>>> >>>> for >>>>>>>>>>>>>>> >>>>

Re: [gem5-users] Installing Shared Libraries to Image in Full Simulation

; https://github.com/darchr/gem5art/blob/master/docs/disks/boot-exit/exit.sh) >>>> which are written to /root/.bashrc are responsible for exiting the >>>> simulation as soon as linux is booted. >>>> >>>> One thing you can do to avoid exiting t

Re: [gem5-users] Kernel for x64

Hi Niranjan, You can check this tutorial to see which kernels are booting successfully. Best, Muhammet niranjan soundararajan , 26 Şub 2020 Çar, 09:31 tarihinde şunu yazdı: > Hello > > Has anyone got a recent Linux kernel

Re: [gem5-users] Paging in System Emulation Mode

Hi Muhammet, > > As far as I know, this isn't possible. However, you can configure your > memory size to be as large as you need. > > Cheers, > Jason > > On Mon, Feb 24, 2020 at 11:51 PM Muhammet Abdullah Soytürk < > muhammetabdullahsoyt...@gmail.com> wrote: &

[gem5-users] Paging in System Emulation Mode

Hi all, Is paging possible in system emulation mode? I want to simulate a system with RISCV isa and provide an input bigger than RAM size. Best, Muhammet ___ gem5-users mailing list gem5-users@gem5.org

[gem5-users] Read/Write to Misc Registers

Hi all, Since m5ops are not available for RISCV, I am trying to define a region of interest by using misc registers. I added a custom register to https://github.com/gem5/gem5/blob/master/src/arch/riscv/registers.hh How can I read the register that I added or write to it so that I can start

[gem5-users] RISC-V Region of Interest

Hi all, Is there any way to define region of interest in an application compiled for RISC-V isa? Best, Muhammet ___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users

[gem5-users] Annotating Region of Interest for RISCV ISA

Hi all, As far as I see there is no m5op_riscv.S file in util/m5 which is needed to annotate the region of interest. Is there any other way that I can annotate an application for riscv isa? Best, Muhammet ___ gem5-users mailing list gem5-users@gem5.org

[gem5-users] RISC-V Multicore Support

Hi all, I have seen that multicore support for RISC-V ISA was implemented by some researchers from Cornell University. I ran into these slides. In one of the slides, they try to pull some changes (

Re: [gem5-users] Incorrect Result When Switching to Detailed CPU

-tutorial.html Abhishek Singh , 19 Şub 2020 Çar, 20:51 tarihinde şunu yazdı: > Hey, > Can you post details on commit number and ISA, so that we can reproduce > the output and see > > On Wed, Feb 19, 2020 at 12:35 PM Muhammet Abdullah Soytürk < > muhammetabdullahsoyt...@gmail.co

[gem5-users] Incorrect Result When Switching to Detailed CPU

Hi all, I am trying to simulate PageRank in full system simulation. The application generates a kronecker graph at the beginning (e.g. ./pr -g 10 generates a kronecker graph witih 1024 nodes). The problem is that when I try to simulate the whole application with KVM, I get the correct result but

[gem5-users] About kvm-x86

Hi all, I am using kvm to boot linux and switch to o3 cpu in full system simulation and it seems to be working but I am getting a bunch of warnings which I am not sure they can be ignored. Here are the warnings that I am getting: warn: ldt: Illegal segment type (0) warn: ldt: P flag not set

[gem5-users] Marking Region of Interest

Hi all, I want to simulate an application in full system simulation with region of interest. I want to mark the region of interest (ROI) in the application, simulate the system with kvm until the beginning of ROI, switch to detailed cpu for ROI, and switch back to kvm for the rest of the

Re: [gem5-users] PerfKvmCounter Error

By the way I am using vmlinux-4.19.83. Muhammet Abdullah Soytürk , 10 Şub 2020 Pzt, 19:45 tarihinde şunu yazdı: > Yes I did but still I am getting the same error message. I have tried > > echo "12345" | echo -1 > /proc/sys/kernel/perf_event_paranoid > > and >

Re: [gem5-users] PerfKvmCounter Error

Yes I did but still I am getting the same error message. I have tried echo "12345" | echo -1 > /proc/sys/kernel/perf_event_paranoid and echo "12345" | sudo sysctl -w kernel.perf_event_paranoid=1 in the equivalent file to npb-install.sh. It does not change anything.

[gem5-users] PerfKvmCounter Error

Hi all, I am trying to run PageRank application in GAP Benchmark with full system simulation. I followed the following npb tutorial: https://gem5art.readthedocs.io/en/latest/tutorials/npb-tutorial.html. and changed the necessary files to make my application work but I am getting a KVM error that

[gem5-users] Packer Template

Hi all, Fourth method In creating disk images documentation ( http://www.gem5.org/documentation/general_docs/fullsystem/disks) is "Using Packer to create a disk image" and the documentation is mentioning about a default packer script called "template.json". I could not find that default script

Re: [gem5-users] Installing Shared Libraries to Image in Full Simulation

e following post on > the mailing list: > > https://www.mail-archive.com/gem5-users@gem5.org/msg17274.html > > You can also refer to gem5art's documentation on building your own disk > images: > > https://gem5art.readthedocs.io/en/latest/main-doc/disks.html > > -Ayaz > >

[gem5-users] Installing Shared Libraries to Image in Full Simulation

Hi all, What is the best way to install shared libraries to linux-x86.img? Since apt or dpkg is not installed, I don't know how to add the shared libraries to the images that I need to execute PageRank in GAP benchmark suite. Best, Muhammet ___

[gem5-users] Thread Creation Failed Error

Hi all, I am trying to simulate the PageRank application available in GAP Benchmark Suite. I get the following error when I try to run it with a simple configuration. Command: build/X86/gem5.opt configs/learning_gem5/part1/simple.py. (I have modified simple.py to run PageRank.) "Thread creation

[gem5-users] About gem5art

Hi all, I just checked gem5art. It seems like a convenient tool to me but I am not sure if it is.mature enough to use without much headache since I saw that it's a relatively new project. Should I start to use it or should I wait a bit longer until they publish a more stable version? Regards,

Re: [gem5-users] gem5 stable release proposal [PLEASE VOTE!]

Hi Jason, For the first one, I think master should be stable. For the second one, I think gem5 should be released three times per year. Cheers, Muhammet Jason Lowe-Power , 16 Ara 2019 Pzt, 22:50 tarihinde şunu yazdı: > Hi all, > > As many of you have seen on gem5-dev, we are going to be

Re: [gem5-users] Understanding the CacheRepl debug flag output

Hi Charitha, Tag of invalid blocks is always 0x which is MaxAddr. You can check allocateBlock function of base.cc under cache to get a better understanding. If you also want to know how it invalidates a block, you can check invalidate function of cache_blk.hh Best, Muhammet

[gem5-users] Where is the "actual data" that is referenced by the block?

Hi all, In cache_blk.hh, there is a field "uint8_t* data" and in the comment it states that this field contains a copy of the data in this block for easy access and it should be kept consistant with the "actual data" referenced by this block. Where is this "actual data" referenced by this block?

[gem5-users] Lowest and Highest Block Address in A Cache

Hi all, I am trying to find the lowest and highest possible block address so that I can generate a random block address. How can I find those addresses? I can only reach to the block addresses that were targeted by the application that is simulated when I use Cache debug flag. Regards, Muhammet

[gem5-users] Unmapped address after fault injection

Hi all, I have been trying to inject some permanent faults into the dcache and when I inject faults to data of some block addresses there is no problem but for other addresses I get "Tried to read unmapped address x" error. x part changes from address to address. It is sometimes 0