[gem5-users] Do gem5 CPU support float computation natively?
Hi all, I have run a matrix test code on my server. When I changed all double type to float type in this code, float really get less time. Theoretically, if your hardware support float computation natively, you will speedup of your program with float type. But when I run them on gem5, float doesn't speedup program. I get the following result:【gem5 command:./build/X86/gem5.fast configs/example/se.py -c /home/zzh/code/matrix_double -n 2 --mem-size=2048MB --caches --l2cache --l2_size 256kB --cpu-clock 1.46GHz --sys-clock 1.46GHz --cpu-type=DerivO3CPU】 So my question is, does gem5's cpu model support float computation natively? If it support, why I get above rsult? ___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
[gem5-users] why gem5 show the opposite result with my server ?
Hi all, My server's specifications are as follows: Intel E5310 @1.6GHz, 8 cores, each core has 32K L1 data cache, 32K L1 instruction cache and 4096K L2 cache. Running two conparison program(double version and float version) on my server with the same input, running time of double version is 2m25s, and float version is 2m7s. float version < double version ( Due to the deviation of float version's result within the acceptable range, so we don't consider its impact ) But when I put two conparison program to run on gem5, I got the opposite result. We used SE mode to simulate Atom Embedded environment, with the command line of " ./build/X86/gem5.fast configs/example/se.py -c /home/zzh/code/double_scale -o 16 -n 2 --mem-size=2048MB --caches --l2cache --l2_size 256kB --cpu-clock 1.46GHz --sys-clock 1.46GHz --cpu-type=AtomicSimpleCPU " Finally I got two stats.txt as follows: Double version’s stats.txt: sim_seconds 337.821068 # Number of seconds simulated sim_ticks sim_ticks337821067887820 #Number of ticks simulated host_inst_rate2164548 # Simulator instruction rate (inst/s) host_op_rate 3369417 # Simulator op (including micro ops) rate (op/s) host_tick_rate2820528361# Simulator tick rate (ticks/s) host_mem_usage 2323216 # Number of bytes of host memory used host_seconds 119772.26 # Real time elapsed on the host sim_insts 259252813524#Number of instructions simulated system.mem_ctrls_0.totalEnergy129723290069280#Total energy per rank( pJ ) system.mem_ctrls_0.averagePower 384 #Core power per rank Float version’s stats.txt: sim_seconds 349.758410 # Number of seconds simulated sim_ticks sim_ticks 349758409947185 #Number of ticks simulated host_inst_rate 2423477# Simulator instruction rate (inst/s) host_op_rate3935604 # Simulator op (including micro ops) rate (op/s) host_tick_rate 3269516689 # Simulator tick rate (ticks/s) host_mem_usage 2324240 # Number of bytes of host memory used host_seconds 106975.57 # Real time elapsed on the host sim_insts259252795117 #Number of instructions simulated system.mem_ctrls_0.totalEnergy 134307229419840#Total energy per rank( pJ ) system.mem_ctrls_0.averagePower384#Core power per rank From the above data, why run time (sim_seconds) show the opposite result with the server? double version of the time is less than the float version. And in terms of the total energy per rank, double version is also smaller than the float version. Though the number of instructions simulated( sim_insts ) of double version is greater than float version. ___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
[gem5-users] How to do power analysis with AtomicSimpleCPU
Hi all, I run my program on gem5 with se mode. “./build/X86/gem5.opt configs/example/se.py -c /home/zzh/code/scale_pre_00 -o 5 -n 2 --mem-size=2048MB --caches --l2cache --l2_size 1MB --cpu-clock 1.46GHz --cpu-type=detailed” But, it cost much long time with detailed CPU. In order to speed up the operation, I change the cpu type from detailed CPU to AtomicSimpleCPU, the fastest cpu type. Now I meet a problem. When using detailedCPU, I get the stats.txt for power analysis. I use the MCPAT tool and use gem5tomcpat to change stats.txt to appropriate xml input file. gem5tomcpat's link is https://bitbucket.org/dskhudia/gem5tomcpat I have already modify its template-xeon.xml to fit my own situation, and obtain correct power file. template-xeon.xml need to take a lot of entries from the stats.txt. But I use the AtomicSimpleCPU instead, stats.txt generated now has no corresponding value, such as system.cpu0.branchPred.condPredicted、system.cpu0.branchPred.condIncorrect and so on. So what should I do? How to do power analysis with AtomicSimpleCPU ? ___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
[gem5-users] Gem5 running too slow, how to speed up
Hello, I run an image processing program on gem5 with the SE mode. Using the command: ./build/X86/gem5.opt configs/example/se.py -c /home/zzh/code/test -o 7 -n 2 --mem-size=2048MB --caches --l2cache --l2_size 256kB --cpu-clock 1.46GHz --cpu-typ=DerivO3CPU But it has been running for 6 days,not yet ended,and is still running. How can I accelerate this operation? It costs too much time. ___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
Re: [gem5-users] McPAT show the same power log file
Hi, Thanks, it really works for me. What's the meaning of this variable “Runtime Dynamic”? Does it mean calculating the average power of the component in the simulation process? Another question is, Can I use this "runtime dynamic" to measure different programs' performance ? -原始邮件- 发件人:"Midhun P" 发送时间:2017-07-06 16:00:25 (星期四) 收件人: "gem5 users mailing list" 抄送: 主题: Re: [gem5-users] McPAT show the same power log file Hi, The input which you are feeding to McPAT is not correct, that is why you are getting inf as runtime dynamic. Compare your template with sample inputs came with McPAT. Most probably some values specified in your template-xeon will be stats.system.cpu.xxx. that you have to change to stats.system.cpu0.xxx. if you are running multicore simulation cpu tag wont be there. There will be multiple statistics corresponds to different cores in the format cpu0, cpu1, cpu2...etc.. change those in template. eg : stats.system.cpu.numCycles need to change to stats.system.cpu0.numCycles Read the warnings coming during gem5 output to McPAT input conversion. You can see warnings related with these tags.. Regards, Midhun P On 6 July 2017 at 12:42, 赵智慧 wrote: Hi,Midhun P Thanks for your answer. But I want to know why the value of runtime dynamic is inf, rather than a specific value? And my goal is to measure the energy consumption of different programs with the same configuration in gem5,, how can I do? -原始邮件- 发件人:"Midhun P" 发送时间:2017-07-06 12:35:27 (星期四) 收件人: "gem5 users mailing list" 抄送: 主题: Re: [gem5-users] McPAT show the same power log file Hi, Only runtime dynamic will be different for different programs. All other metrics (area, leakage etc) are depends on system config, so on different programs with same system configuration, you will get same value. Regards, Midhun P On 06-Jul-2017 7:42 AM, "赵智慧" wrote: Hi, I ran different programs on gem5's SE mode. “./build/X86/gem5.opt configs/example/se.py -c /home/zzh/code/scale_pre_00 -o 5 -n 2 --mem-size=2048MB --caches --l2cache --l2_size 1MB --cpu-clock 1.46GHz --cpu-type=detailed” Then I used McPAT to analyze their power.I follow the URL of the method, with the inside of the files——GEM5ToMcPAT.py、template-xeon.xml https://bitbucket.org/dskhudia/gem5tomcpat/src But I get the contents of the two power files actually exactly the same. What's the problem with this? Or,power is not related to the running program? Power is only related to the hardware structure of the simulation? ___ 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 -- Regards, Midhun P mail to : midhunpmad...@gmail.com midhunpmad...@live.com Mobile : +91-9946001223___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
Re: [gem5-users] McPAT show the same power log file
Hi,Midhun P Thanks for your answer. But I want to know why the value of runtime dynamic is inf, rather than a specific value? And my goal is to measure the energy consumption of different programs with the same configuration in gem5,, how can I do? -原始邮件- 发件人:"Midhun P" 发送时间:2017-07-06 12:35:27 (星期四) 收件人: "gem5 users mailing list" 抄送: 主题: Re: [gem5-users] McPAT show the same power log file Hi, Only runtime dynamic will be different for different programs. All other metrics (area, leakage etc) are depends on system config, so on different programs with same system configuration, you will get same value. Regards, Midhun P On 06-Jul-2017 7:42 AM, "赵智慧" wrote: Hi, I ran different programs on gem5's SE mode. “./build/X86/gem5.opt configs/example/se.py -c /home/zzh/code/scale_pre_00 -o 5 -n 2 --mem-size=2048MB --caches --l2cache --l2_size 1MB --cpu-clock 1.46GHz --cpu-type=detailed” Then I used McPAT to analyze their power.I follow the URL of the method, with the inside of the files——GEM5ToMcPAT.py、template-xeon.xml https://bitbucket.org/dskhudia/gem5tomcpat/src But I get the contents of the two power files actually exactly the same. What's the problem with this? Or,power is not related to the running program? Power is only related to the hardware structure of the simulation? ___ 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] McPAT show the same power log file
Hi, I ran different programs on gem5's SE mode. “./build/X86/gem5.opt configs/example/se.py -c /home/zzh/code/scale_pre_00 -o 5 -n 2 --mem-size=2048MB --caches --l2cache --l2_size 1MB --cpu-clock 1.46GHz --cpu-type=detailed” Then I used McPAT to analyze their power.I follow the URL of the method, with the inside of the files——GEM5ToMcPAT.py、template-xeon.xml https://bitbucket.org/dskhudia/gem5tomcpat/src But I get the contents of the two power files actually exactly the same. What's the problem with this? Or,power is not related to the running program? Power is only related to the hardware structure of the simulation?___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
Re: [gem5-users] gem5 SE with Win10's 14.04 Ubuntu
It really works for me. Thank you. -原始邮件- 发件人:"Ferran Olid" 发送时间:2017-07-05 15:11:49 (星期三) 收件人: gem5-users@gem5.org 抄送: 主题: Re: [gem5-users] gem5 SE with Win10's 14.04 Ubuntu It is a gem5 option, not a se.py script option. Try "build/X86/gem5.opt --remote-gdb-port=0 configs/example/se.py -c tests/test-progs/hello/bin/x86/linux/hello" Ferran O. On 05/07/17 02:54, 赵智慧 wrote: But when I run "build/X86/gem5.opt configs/example/se.py -c tests/test-progs/hello/bin/x86/linux/hello --remote-gdb-port=0" it shows "se.py:error:no such option:--remote-gdb-port". So how to use this command ? -原始邮件- 发件人:"Jason Lowe-Power" 发送时间:2017-07-04 22:11:32 (星期二) 收件人: "gem5 users mailing list" 抄送: 主题: Re: [gem5-users] gem5 SE with Win10's 14.04 Ubuntu A quick fix is to run with --remote-gdb-port=0. Jason On Mon, Jul 3, 2017 at 10:45 PM 赵智慧 wrote: Hello,all I activated Win10's built-in Ubuntu, and then install gem5 on this ubuntu. After compiling the X86 architecture with scons, I run the command “build/X86/gem5.opt configs/example/se.py -c tests/test-progs/hello/bin/x86/linux/hello”.But it show the errors below: panic: Could not set up async IO so, what should I do to solve the problem? ___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users ___ gem5-users mailing list gem5-users@gem5.orghttp://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 SE with Win10's 14.04 Ubuntu
But when I run "build/X86/gem5.opt configs/example/se.py -c tests/test-progs/hello/bin/x86/linux/hello --remote-gdb-port=0" it shows "se.py:error:no such option:--remote-gdb-port". So how to use this command ? -原始邮件- 发件人:"Jason Lowe-Power" 发送时间:2017-07-04 22:11:32 (星期二) 收件人: "gem5 users mailing list" 抄送: 主题: Re: [gem5-users] gem5 SE with Win10's 14.04 Ubuntu A quick fix is to run with --remote-gdb-port=0. Jason On Mon, Jul 3, 2017 at 10:45 PM 赵智慧 wrote: Hello,all I activated Win10's built-in Ubuntu, and then install gem5 on this ubuntu. After compiling the X86 architecture with scons, I run the command “build/X86/gem5.opt configs/example/se.py -c tests/test-progs/hello/bin/x86/linux/hello”.But it show the errors below: panic: Could not set up async IO so, what should I do to solve the problem? ___ 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] gem5 SE with Win10's 14.04 Ubuntu
Hello,all I activated Win10's built-in Ubuntu, and then install gem5 on this ubuntu. After compiling the X86 architecture with scons, I run the command “build/X86/gem5.opt configs/example/se.py -c tests/test-progs/hello/bin/x86/linux/hello”.But it show the errors below: panic: Could not set up async IO so, what should I do to solve the problem?___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
Re: [gem5-users] can't run new disk image(Ubuntu 14.04.img) in gem5 X86 FS mode
Hello, I have enabled devtmpfs in "Maintain a devtmpfs filesystem to mount at /dev" with the step of make menuconfig. But it didn't work. FS mode still have the problem. -原始邮件- 发件人:"Jason Lowe-Power" 发送时间:2017-06-22 22:15:20 (星期四) 收件人: "gem5 users mailing list" 抄送: 主题: Re: [gem5-users] can't run new disk image(Ubuntu 14.04.img) in gem5 X86 FS mode Hello, From here: http://www.lowepower.com/jason/setting-up-gem5-full-system.html Problem: Can't mount /dev > Failed to mount devtmpfs at /dev: No such device > Freezing execution. Something like the above (this was taken from arch linux boot). The problem is that that the right devfs is not compiled into the kernel. You need to make sure that devtmpfs is enabled in your kernel. You may have to recompile the kernel with devtmpfs enabled. Jason On Thu, Jun 22, 2017 at 2:45 AM 赵智慧 wrote: hello, everyone The default disk(linux-x86.img) can't support new version GLIBC(version 2.14) that my program needed. So I create new disk image(ubuntu 14.04) to replace it, and I use the new kernel(x86-64_vmlinux_2.6.28.4). But when I run gem5, the m5term output like this: Mount failed for selinuxfs on /sys/fs/selinux: No such file or directory init: plymouth-upstart-bridge main process(948) terminated with status 1 init: plymouth-upstart-bridge main process ended, respawning init: plymouth-upstart-bridge main process(948) terminated with status 1 init: plymouth-upstart-bridge main process ended, respawning --- I see someone saying that this is not an error and will not affect the simulation. But like this,I can not get into the simulation system. When I change the ubuntu-14.04.img to ubuntu-12.04.5.img,the m5term output like this: mknod:extra operand '64' Try 'mknod --help' for more information mknod:extra operand '64' Try 'mknod --help' for more information mknod:extra operand '64' Try 'mknod --help' for more information Can anybody help me to slove the problem? It confused me for a long time. Thank you. ___ 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] can't run new disk image(Ubuntu 14.04.img) in gem5 X86 FS mode
Thanks for your answer. But which step can make sure the devtmpfs is enabled when compiling a kernel? 发件人:"Jason Lowe-Power" 发送日期:2017-06-22 22:15:20 收件人:"gem5 users mailing list" 抄送人: 主 题:Re: [gem5-users] can't run new disk image(Ubuntu 14.04.img) in gem5 X86 FS mode Hello, From here: http://www.lowepower.com/jason/setting-up-gem5-full-system.html Problem: Can't mount /dev > Failed to mount devtmpfs at /dev: No such device > Freezing execution. Something like the above (this was taken from arch linux boot). The problem is that that the right devfs is not compiled into the kernel. You need to make sure that devtmpfs is enabled in your kernel. You may have to recompile the kernel with devtmpfs enabled. Jason On Thu, Jun 22, 2017 at 2:45 AM 赵智慧 wrote: hello, everyone The default disk(linux-x86.img) can't support new version GLIBC(version 2.14) that my program needed. So I create new disk image(ubuntu 14.04) to replace it, and I use the new kernel(x86-64_vmlinux_2.6.28.4). But when I run gem5, the m5term output like this: Mount failed for selinuxfs on /sys/fs/selinux: No such file or directory init: plymouth-upstart-bridge main process(948) terminated with status 1 init: plymouth-upstart-bridge main process ended, respawning init: plymouth-upstart-bridge main process(948) terminated with status 1 init: plymouth-upstart-bridge main process ended, respawning --- I see someone saying that this is not an error and will not affect the simulation. But like this,I can not get into the simulation system. When I change the ubuntu-14.04.img to ubuntu-12.04.5.img,the m5term output like this: mknod:extra operand '64' Try 'mknod --help' for more information mknod:extra operand '64' Try 'mknod --help' for more information mknod:extra operand '64' Try 'mknod --help' for more information Can anybody help me to slove the problem? It confused me for a long time. Thank you. ___ 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] can't run new disk image(Ubuntu 14.04.img) in gem5 X86 FS mode
hello, everyone The default disk(linux-x86.img) can't support new version GLIBC(version 2.14) that my program needed. So I create new disk image(ubuntu 14.04) to replace it, and I use the new kernel(x86-64_vmlinux_2.6.28.4). But when I run gem5, the m5term output like this: Mount failed for selinuxfs on /sys/fs/selinux: No such file or directory init: plymouth-upstart-bridge main process(948) terminated with status 1 init: plymouth-upstart-bridge main process ended, respawning init: plymouth-upstart-bridge main process(948) terminated with status 1 init: plymouth-upstart-bridge main process ended, respawning --- I see someone saying that this is not an error and will not affect the simulation. But like this,I can not get into the simulation system. When I change the ubuntu-14.04.img to ubuntu-12.04.5.img,the m5term output like this: mknod:extra operand '64' Try 'mknod --help' for more information mknod:extra operand '64' Try 'mknod --help' for more information mknod:extra operand '64' Try 'mknod --help' for more information Can anybody help me to slove the problem? It confused me for a long time. Thank you. ___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
Re: [gem5-users] FS-gem5-X86_stats.txt show nothing
:01 ../ -rwxr-xr-x 1 root root 10490348 Mar 28 2017 TestGauss2* -rw-r--r-- 1 root root 608454 Jan 1 00:01 a.bmp drwxr-xr-x 2 root root 4096 Aug 6 2014 bin/ drwxr-xr-x 2 root root 4096 Apr 19 2012 boot/ drwxr-xr-x 3 root root 4096 Jan 1 00:00 dev/ drwxr-xr-x 41 root root 4096 Jan 1 00:00 etc/ drwxr-xr-x 2 root root 4096 Apr 19 2012 home/ drwxr-xr-x 11 root root 4096 Mar 25 2017 lib/ drwxr-xr-x 2 root root 4096 Aug 6 2014 lib64/ drwx-- 2 root root16384 Mar 24 2017 lost+found/ drwxr-xr-x 2 root root 4096 Aug 6 2014 media/ drwxr-xr-x 2 root root 4096 Apr 19 2012 mnt/ drwxr-xr-x 2 root root 4096 Aug 6 2014 opt/ dr-xr-xr-x 39 root root0 Jan 1 00:00 proc/ drwx-- 2 root root 4096 Aug 6 2014 root/ drwxr-xr-x 5 root root 4096 Aug 6 2014 run/ drwxr-xr-x 2 root root 4096 Mar 24 2017 sbin/ drwxr-xr-x 2 root root 4096 Mar 5 2012 selinux/ drwxr-xr-x 2 root root 4096 Aug 6 2014 srv/ drwxr-xr-x 12 root root0 Jan 1 00:00 sys/ -rw-r--r-- 1 root root 608454 Mar 25 2017 test.bmp drwxrwxrwt 2 root root 4096 Aug 6 2014 tmp/ drwxr-xr-x 10 root root 4096 Aug 6 2014 usr/ drwxr-xr-x 11 root root 4096 Aug 6 2014 var/ root@localhost:/# sz a.bmp bash: sz: command not found root@localhost:/# exit init: mountall-shell main process (968) terminated with status 127 -原始邮件- 发件人:"Jason Lowe-Power" 发送时间:2017-03-31 21:44:20 (星期五) 收件人: "gem5 users mailing list" 抄送: 主题: Re: [gem5-users] FS-gem5-X86_stats.txt show nothing Hi, What was the output to stdout/stderr? What is the output in system.pc.com_1.terminal? This looks like gem5 ran into an error before dumping the stats, but it's hard to know without more information. Jason On Thu, Mar 30, 2017 at 8:33 PM 赵智慧 wrote: Hi, After running a image gaussian blur processing program on Gem5 in FS mode, I enter the directory of gem5/m5out/. But the stats.txt was 0 byte, showing nothing. I use the following commands to run: zzh@ubuntu14:~/gem5$ sudo build/X86/gem5.opt configs/example/fs.py --disk-image=ubuntu-12.04.img --kernel=/home/zzh/gem5/x86Dist/binaries/x86_64-vmlinux-2.6.28.4-smp ___ 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] FS-gem5-X86_stats.txt show nothing
Hi, After running a image gaussian blur processing program on Gem5 in FS mode, I enter the directory of gem5/m5out/. But the stats.txt was 0 byte, showing nothing. I use the following commands to run: zzh@ubuntu14:~/gem5$ sudo build/X86/gem5.opt configs/example/fs.py --disk-image=ubuntu-12.04.img --kernel=/home/zzh/gem5/x86Dist/binaries/x86_64-vmlinux-2.6.28.4-smp___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
Re: [gem5-users] run gem5 in FS mode with this kernel, though, I get this error: Kernel panic - not syncing: Attempted to kill the idle task!
Hi, But the provided .config only include 2.6.22.9 , 2.6.25.1, 2.6.28.4 in the file config-x86.tar.bz2. So where can I download the other .config for gem5's kernel ? and how can I enable smp feature and specify the maximum amount of cores in the file .config ? -原始邮件- 发件人:"Javier Cano Cano" 发送时间:2017-03-23 23:31:03 (星期四) 收件人: "gem5 users mailing list" 抄送: 主题: Re: [gem5-users] run gem5 in FS mode with this kernel, though, I get this error: Kernel panic - not syncing: Attempted to kill the idle task! Hi, If you want to upgrade your kernel, use the 3.4.112 version and the provided .config. Just make sure to enable smp feature and specify the maximum amount of cores if you want to run multi-core simulations. Have a nice day, Best, Cano. 2017-03-23 13:25 GMT+01:00 赵智慧 : Hi, When I running X86 full system on gem5 with the kernel of x86_64-vmlinux-2.6.28.4.smp,I have met the errors:Kernel panic - not syncing: Attempted to kill the idle task! Details as follows: kernel BUG at arch/x86/kernel/xsave.c:323! invalid opcode: [#1] last sysfs file: CPU 0 Modules linked in: Pid: 0, comm: swapper Tainted: GW 2.6.28.4-hg5ea85692a53e-dirty #2 RIP: 0010:[] [] xsave_cntxt_init+0x35/0x130 RSP: 0018:807ddf48 EFLAGS: 00b8 RAX: 002d RBX: RCX: 80843460 RDX: RSI: 0d74 RDI: 80773180 RBP: R08: R09: 03fd R10: R11: R12: R13: R14: R15: FS: () GS:80835020() knlGS: CS: 0010 DS: 0018 ES: 0018 CR0: 80050033 CR2: CR3: 00201000 CR4: 06a0 DR0: DR1: DR2: DR3: DR6: 0ff0 DR7: Process swapper (pid: 0, threadinfo 807dc000, task 80766380) Stack: 80804fb3 8080aec0 80805cee 8000897669002087 8022ee104b10 80809ec0 8080aec0 807de981 8080aec0 Call Trace: [] fpu_init+0x3e/0x8b [] cpu_init+0x222/0x240 [] start_kernel+0x16f/0x2d9 [] x86_64_start_kernel+0xd9/0xdd Code: 48 c1 e2 20 89 c0 48 8d 34 02 48 89 f0 48 89 35 12 1b 27 00 83 e0 03 48 83 f8 03 74 12 48 c7 c7 60 82 6d 80 31 c0 e8 7f 69 01 00 <0f> 0b eb fe f6 05 57 1c 21 00 04 48 c7 05 e5 1a 27 00 03 00 00 RIP [] xsave_cntxt_init+0x35/0x130 RSP ---[ end trace 4eaa2a86a8e2da22 ]--- Kernel panic - not syncing: Attempted to kill the idle task! I'm sure my other startup steps are no problem. Now, only combinations of Linux-x86.img and x86_64-vmlinux-2.6.22.9 can work well.The kernel of x86_64-vmlinux-2.6.28.4, I have tried the version of download from http://www.cs.utexas.edu/~parsec_m5/x86_64-vmlinux-2.6.28.4-smp , and I also tried compile the kernel from sourcecode as the steps in http://gem5.org/Linux_kernel. I also tried build a new disk image,but other combinations all have the question above. ___ 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] run gem5 in FS mode with this kernel, though, I get this error: Kernel panic - not syncing: Attempted to kill the idle task!
Hi, When I running X86 full system on gem5 with the kernel of x86_64-vmlinux-2.6.28.4.smp,I have met the errors:Kernel panic - not syncing: Attempted to kill the idle task! Details as follows: kernel BUG at arch/x86/kernel/xsave.c:323! invalid opcode: [#1] last sysfs file: CPU 0 Modules linked in: Pid: 0, comm: swapper Tainted: GW 2.6.28.4-hg5ea85692a53e-dirty #2 RIP: 0010:[] [] xsave_cntxt_init+0x35/0x130 RSP: 0018:807ddf48 EFLAGS: 00b8 RAX: 002d RBX: RCX: 80843460 RDX: RSI: 0d74 RDI: 80773180 RBP: R08: R09: 03fd R10: R11: R12: R13: R14: R15: FS: () GS:80835020() knlGS: CS: 0010 DS: 0018 ES: 0018 CR0: 80050033 CR2: CR3: 00201000 CR4: 06a0 DR0: DR1: DR2: DR3: DR6: 0ff0 DR7: Process swapper (pid: 0, threadinfo 807dc000, task 80766380) Stack: 80804fb3 8080aec0 80805cee 8000897669002087 8022ee104b10 80809ec0 8080aec0 807de981 8080aec0 Call Trace: [] fpu_init+0x3e/0x8b [] cpu_init+0x222/0x240 [] start_kernel+0x16f/0x2d9 [] x86_64_start_kernel+0xd9/0xdd Code: 48 c1 e2 20 89 c0 48 8d 34 02 48 89 f0 48 89 35 12 1b 27 00 83 e0 03 48 83 f8 03 74 12 48 c7 c7 60 82 6d 80 31 c0 e8 7f 69 01 00 <0f> 0b eb fe f6 05 57 1c 21 00 04 48 c7 05 e5 1a 27 00 03 00 00 RIP [] xsave_cntxt_init+0x35/0x130 RSP ---[ end trace 4eaa2a86a8e2da22 ]--- Kernel panic - not syncing: Attempted to kill the idle task! I'm sure my other startup steps are no problem. Now, only combinations of Linux-x86.img and x86_64-vmlinux-2.6.22.9 can work well.The kernel of x86_64-vmlinux-2.6.28.4, I have tried the version of download from http://www.cs.utexas.edu/~parsec_m5/x86_64-vmlinux-2.6.28.4-smp , and I also tried compile the kernel from sourcecode as the steps in http://gem5.org/Linux_kernel. I also tried build a new disk image,but other combinations all have the question above.___ gem5-users mailing list gem5-users@gem5.org http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users