Thank you very much. Best Regards, Hongsheng Lu
On Tue, Apr 20, 2010 at 11:40 AM, <[email protected]> wrote: > Send m5-users mailing list submissions to > [email protected] > > To subscribe or unsubscribe via the World Wide Web, visit > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > or, via email, send a message with subject or body 'help' to > [email protected] > > You can reach the person managing the list at > [email protected] > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of m5-users digest..." > > > Today's Topics: > > 1. Re: SSH on m5 (nathan binkert) > 2. Re: Accessing m5.debug functions in gdb (nathan binkert) > 3. Re: reset & dump stats (nathan binkert) > 4. Re: Banked Cache (nathan binkert) > 5. Re: m5threads (nathan binkert) > 6. Re: m5threads (Eberle) > 7. Re: m5threads (Eberle) > 8. Re: reset & dump stats (Adam Beece) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Mon, 19 Apr 2010 16:11:16 -0700 > From: nathan binkert <[email protected]> > Subject: Re: [m5-users] SSH on m5 > To: M5 users mailing list <[email protected]> > Message-ID: > <[email protected]> > Content-Type: text/plain; charset=ISO-8859-1 > > > mounting swap...modprobe: FATAL: Could not load > > /lib/modules/2.6.13/modules.dep: No such file or directory > > > > modprobe: FATAL: Could not load /lib/modules/2.6.13/modules.dep: No > > such file or directory > > Those errors can be ignored. I can't remember exactly what they were, > but apache is basically trying to see if the kernel supports something > (like ipv6) and modprobe for the module fails. > > > But when I use NetPerf benchmark, there is no such error. > > > > I figured that NetPerf is probably statically linked in. > This isn't the issue. > > > But I got the following error, not sure what it means here with > > respect to the simulator. > > I simply invoked sshd as /opt/sbin/sshd in the rcS script. > > > > "starting sshd.../tmp/script: line 43: /opt/sbin/sshd: Input/output > error" > Not sure what exactly the problem is, but my guess is that it's a > filesystem problem or some problem with the sshd binary. Where did > you get the binary from? > > > Could you point me to the right direction of what is the good and > > easier way to port ssh to m5? > You want to use crosstools to build a crosscompiler toolchain and then > use that to build sshd. Alternatively, you can try to use a gentoo > stage3 tarball to populate a disk image. > > Nate > > > ------------------------------ > > Message: 2 > Date: Mon, 19 Apr 2010 16:12:41 -0700 > From: nathan binkert <[email protected]> > Subject: Re: [m5-users] Accessing m5.debug functions in gdb > To: M5 users mailing list <[email protected]> > Message-ID: > <[email protected]> > Content-Type: text/plain; charset=ISO-8859-1 > > > I have been struggling with a performance bottleneck in the o3cpu model > for > > a few weeks (IPC capped at 3, regardless of the width of the core), and > I've > > been using gdb on m5.debug to try to find the bottleneck. From the > debugging > > information on the wiki ( > http://www.m5sim.org/wiki/index.php/Debugging_M5) I > > should be able to call some of the dump functions as I step through the > > simulation. However, I get the following error: > > (gdb) call eventq_dump() > > No symbol "eventq_dump" in current context. > > This usually happens only if the simulator hasn't actually started > yet. You need to start the simulator by doing something like: > (gdb) break SimStartup > (gdb) run > > Then at the first breakpoint, M5 is loaded and running, then you can > call various functions. > > Nate > > > ------------------------------ > > Message: 3 > Date: Mon, 19 Apr 2010 16:17:46 -0700 > From: nathan binkert <[email protected]> > Subject: Re: [m5-users] reset & dump stats > To: M5 users mailing list <[email protected]> > Message-ID: > <[email protected]> > Content-Type: text/plain; charset=ISO-8859-1 > > > I was wondering if there was any more complete documentation on > > resetstats and dumpstats than can be found at > > http://www.m5sim.org/wiki/index.php/M5ops ? > > > > resetstats, in particular does not appear to reset the total instruction > count. > > sim_insts, sim_second, sim_freq, host_inst_rate, host_mem_usage, > host_seconds, and host_tick_rate are all special stats in that they > reflect parameters about the simulator itself and the host, not the > simulated program. None get reset. That could be changed by adding a > bit of code. See src/sim/stat_control.cc > > Nate > > > ------------------------------ > > Message: 4 > Date: Mon, 19 Apr 2010 21:07:07 -0700 > From: nathan binkert <[email protected]> > Subject: Re: [m5-users] Banked Cache > To: M5 users mailing list <[email protected]>, Jiayuan Meng > <[email protected]>, Richard Strong <[email protected]> > Message-ID: > <[email protected]> > Content-Type: text/plain; charset=ISO-8859-1 > > Unfortunately, Jiayuan's patches have not yet made it back into the > main M5 tree, so if you want to use them, you'll probably need to get > help from Jiayuan himself or perhaps Rick Strong. > > Nate > > > I am a beginner on M5 and doing a project on cache performance. However, > I > > found M5 seems not supporting banked cache although I found the webs > > > > http://article.gmane.org/gmane.comp.emulators.m5.users/5110/match=bankand > > http://article.gmane.org/gmane.comp.emulators.m5.users/2911/match=bank > > > > > > (under the help of Siming Chen, who posted a similar email with the name > > "About Bank and Crossbar Model" on 04-11-2010 in the email archive). It > > seems there is a patch for this > > > > functionality. So, could some one please help me out? Thank you very > much. > > > > > > Additionally, attached is a pdf published by jiayuan Meng online. I found > it > > via Google and attach it for convenience. > > > > > > Best Regards, > > Hongsheng Lu > > > > > > > > > > _______________________________________________ > > m5-users mailing list > > [email protected] > > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > > > > > ------------------------------ > > Message: 5 > Date: Mon, 19 Apr 2010 21:13:07 -0700 > From: nathan binkert <[email protected]> > Subject: Re: [m5-users] m5threads > To: M5 users mailing list <[email protected]> > Message-ID: > <[email protected]> > Content-Type: text/plain; charset=ISO-8859-1 > > > I have created a m5threads static linkable library for Sparc, with the > files > > available in the m5 repository. I used the Sparc compiler available in > the > > m5 website. > > The thing is: when I link my multithreaded helloworld app with this lib, > I > > get the following error: > > > > eberl...@bellatrix:~/UFSC/Wannabe/helloworld$ > sparc64-unknown-linux-gnu-g++ > > -static -L. -lm5pthread hello.cpp -o hello > > /tmp/ccfh1vud.o(.text+0x2b0): In function `main': > > : undefined reference to `pthread_join' > > collect2: ld returned 1 exit status > > I think your link order is bad. generally, -l options need to come > after object files, otherwise, the linker doesn't know to pull out the > relevant functions. (The linker processes files in order.) > > > ------------------------------ > > Message: 6 > Date: Tue, 20 Apr 2010 08:51:59 -0300 > From: Eberle <[email protected]> > Subject: Re: [m5-users] m5threads > To: M5 users mailing list <[email protected]> > Message-ID: > <[email protected]> > Content-Type: text/plain; charset="utf-8" > > Thanks for the reply. > I don't think that is the problem because the first pthread function called > in the hello.cpp was correctly found. > I was looking inside the pthreads library generated and I noticed that this > pthread_join function wasn't listed there. I really can't imagine a reason. > > To create the library I executed the following two lines: > > sparc64-unknown-linux-gnu-gcc -c pthread.c -o m5pthread.o > sparc64-unknown-linux-gnu-ar csr libm5pthread.a m5pthread.o > > > Funny thing... > > > On Tue, Apr 20, 2010 at 1:13 AM, nathan binkert <[email protected]> wrote: > > > > I have created a m5threads static linkable library for Sparc, with the > > files > > > available in the m5 repository. I used the Sparc compiler available in > > the > > > m5 website. > > > The thing is: when I link my multithreaded helloworld app with this > lib, > > I > > > get the following error: > > > > > > eberl...@bellatrix:~/UFSC/Wannabe/helloworld$ > > sparc64-unknown-linux-gnu-g++ > > > -static -L. -lm5pthread hello.cpp -o hello > > > /tmp/ccfh1vud.o(.text+0x2b0): In function `main': > > > : undefined reference to `pthread_join' > > > collect2: ld returned 1 exit status > > > > I think your link order is bad. generally, -l options need to come > > after object files, otherwise, the linker doesn't know to pull out the > > relevant functions. (The linker processes files in order.) > > _______________________________________________ > > m5-users mailing list > > [email protected] > > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: > http://m5sim.org/cgi-bin/mailman/private/m5-users/attachments/20100420/105d5561/attachment-0001.htm > > ------------------------------ > > Message: 7 > Date: Tue, 20 Apr 2010 11:13:18 -0300 > From: Eberle <[email protected]> > Subject: Re: [m5-users] m5threads > To: M5 users mailing list <[email protected]> > Message-ID: > <[email protected]> > Content-Type: text/plain; charset="utf-8" > > I was able to successfuly compile my test program without calling > pthread_join(). > But, when I run the executable on the platform with 2 SPARCs, the following > happened: > > command line: ./build/SPARC_SE/m5.fast configs/splash2/teste.py -n 2 -b > Helloworld > Global frequency set at 1000000000000 ticks per second > 0: system.remote_gdb.listener: listening for remote gdb on port 7000 > 0: system.remote_gdb.listener: listening for remote gdb on port 7001 > info: Entering event queue @ 0. Starting simulation... > Initiating > experiment... > <- This is the first of my helloworld printed outputs. > panic: Tried to execute unmapped address 0. > @ cycle 76104000 > [invoke:build/SPARC_SE/arch/sparc/faults.cc, line 629] > Memory Usage: 200844 KBytes > For more information see: http://www.m5sim.org/panic/95542d88 > Program aborted at cycle 76104000 > Aborted > > > I suppose it happens when I try to create a second thread. > The teste.py script is the splash2/run.py modified to use my executable > file > instead. > > Has someone successfully compiled and ran a multithreaded Sparc software? > > -- > Eberle A. Rambo. > > > On Tue, Apr 20, 2010 at 8:51 AM, Eberle <[email protected]> wrote: > > > Thanks for the reply. > > I don't think that is the problem because the first pthread function > called > > in the hello.cpp was correctly found. > > I was looking inside the pthreads library generated and I noticed that > this > > pthread_join function wasn't listed there. I really can't imagine a > reason. > > > > To create the library I executed the following two lines: > > > > sparc64-unknown-linux-gnu-gcc -c pthread.c -o m5pthread.o > > sparc64-unknown-linux-gnu-ar csr libm5pthread.a m5pthread.o > > > > > > Funny thing... > > > > > > > > On Tue, Apr 20, 2010 at 1:13 AM, nathan binkert <[email protected]> > wrote: > > > >> > I have created a m5threads static linkable library for Sparc, with the > >> files > >> > available in the m5 repository. I used the Sparc compiler available in > >> the > >> > m5 website. > >> > The thing is: when I link my multithreaded helloworld app with this > lib, > >> I > >> > get the following error: > >> > > >> > eberl...@bellatrix:~/UFSC/Wannabe/helloworld$ > >> sparc64-unknown-linux-gnu-g++ > >> > -static -L. -lm5pthread hello.cpp -o hello > >> > /tmp/ccfh1vud.o(.text+0x2b0): In function `main': > >> > : undefined reference to `pthread_join' > >> > collect2: ld returned 1 exit status > >> > >> I think your link order is bad. generally, -l options need to come > >> after object files, otherwise, the linker doesn't know to pull out the > >> relevant functions. (The linker processes files in order.) > >> _______________________________________________ > >> m5-users mailing list > >> [email protected] > >> http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > >> > > > > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: > http://m5sim.org/cgi-bin/mailman/private/m5-users/attachments/20100420/2b053758/attachment-0001.htm > > ------------------------------ > > Message: 8 > Date: Tue, 20 Apr 2010 11:35:14 -0400 > From: Adam Beece <[email protected]> > Subject: Re: [m5-users] reset & dump stats > To: M5 users mailing list <[email protected]> > Message-ID: > <[email protected]> > Content-Type: text/plain; charset=ISO-8859-1 > > Thank you. > > Oddly enough, my experiments definitely show that sim_seconds is > reset, at least with the m5-stable repository. > > Adm > > On Mon, Apr 19, 2010 at 7:17 PM, nathan binkert <[email protected]> wrote: > >> I was wondering if there was any more complete documentation on > >> resetstats and dumpstats than can be found at > >> http://www.m5sim.org/wiki/index.php/M5ops ? > >> > >> resetstats, in particular does not appear to reset the total instruction > count. > > > > sim_insts, sim_second, sim_freq, host_inst_rate, host_mem_usage, > > host_seconds, and host_tick_rate are all special stats in that they > > reflect parameters about the simulator itself and the host, not the > > simulated program. ?None get reset. ?That could be changed by adding a > > bit of code. ?See src/sim/stat_control.cc > > > > ?Nate > > _______________________________________________ > > m5-users mailing list > > [email protected] > > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > > > > > ------------------------------ > > _______________________________________________ > m5-users mailing list > [email protected] > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > > End of m5-users Digest, Vol 45, Issue 24 > **************************************** >
_______________________________________________ m5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
