I am using gem5 to simulate performance power variation running different application on different core sizes. I tried to adjust the cache size, issue width, and some related parameters. The problem is that if I need to run SPEC benchmarks on that emulated system, I have to use a console window. And the job is always kicked out by the server (I set up gem5 simulator on a server). Is there a way to use a script to run benchmarks so that I can just submit the script to the server without using the head node?
Thanks, Yanqi ________________________________________ From: [email protected] [[email protected]] on behalf of [email protected] [[email protected]] Sent: Thursday, April 26, 2012 12:00 PM To: [email protected] Subject: gem5-dev Digest, Vol 60, Issue 97 Send gem5-dev mailing list submissions to [email protected] To subscribe or unsubscribe via the World Wide Web, visit http://m5sim.org/mailman/listinfo/gem5-dev 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 gem5-dev digest..." Today's Topics: 1. Re: Cron <m5test@zizzer> /z/m5/regression/do-regression quick (Steve Reinhardt) 2. Re: Cron <m5test@zizzer> /z/m5/regression/do-regression quick (Steve Reinhardt) 3. Re: Cron <m5test@zizzer> /z/m5/regression/do-regression quick (Nilay Vaish) 4. Re: Cron <m5test@zizzer> /z/m5/regression/do-regression quick (Steve Reinhardt) 5. Re: Review Request: MEM: Add the communication monitor (Andreas Hansson) ---------------------------------------------------------------------- Message: 1 Date: Thu, 26 Apr 2012 06:52:35 -0700 From: Steve Reinhardt <[email protected]> To: gem5 Developer List <[email protected]> Subject: Re: [gem5-dev] Cron <m5test@zizzer> /z/m5/regression/do-regression quick Message-ID: <CAHgMoh-y9Hjgu0d8Mm4QHo02ONwg9KWzScAtdshKc=_wfip...@mail.gmail.com> Content-Type: text/plain; charset=ISO-8859-1 Yes, your change to util/regress is bad, you have a comma inside a string instead of outside. You should actually use util/regress to run regressions before you commit, since that's what the nightly regression script does. Then you would have caught this beforehand. Steve On Thu, Apr 26, 2012 at 6:23 AM, Nilay Vaish <[email protected]> wrote: > Can some one check what went wrong? It is likely that I committed some > mistake. > > -- > Nilay > > > On Thu, 26 Apr 2012, Cron Daemon wrote: > > >> See /z/m5/regression/regress-2012-04-26-03:00:**01 for details. >> >> ______________________________**_________________ >> gem5-dev mailing list >> [email protected] >> http://m5sim.org/mailman/**listinfo/gem5-dev<http://m5sim.org/mailman/listinfo/gem5-dev> >> >> ______________________________**_________________ > gem5-dev mailing list > [email protected] > http://m5sim.org/mailman/**listinfo/gem5-dev<http://m5sim.org/mailman/listinfo/gem5-dev> > ------------------------------ Message: 2 Date: Thu, 26 Apr 2012 06:58:57 -0700 From: Steve Reinhardt <[email protected]> To: gem5 Developer List <[email protected]> Subject: Re: [gem5-dev] Cron <m5test@zizzer> /z/m5/regression/do-regression quick Message-ID: <CAHgMoh-Lp=FyeBhq6H1v52aVw=gg7hnf3a0dy7vjkwaa3e4...@mail.gmail.com> Content-Type: text/plain; charset=ISO-8859-1 The bug is in util/regress (see below), but my quick diagnosis was wrong in the details... I just glanced at the diff and didn't notice that the thing you changed is really a long string split across several lines and not a list. I think the actual issue is that you're missing a comma. Here's the regression output; note the problems at the end of the scons args generated by util/regress. pulling from http://repo.gem5.org/gem5 searching for changes adding changesets adding manifests adding file changes added 5 changesets with 21 changes to 19 files 19 files updated, 0 files merged, 0 files removed, 0 files unresolved pulling from http://repo.gem5.org/encumbered searching for changes no changes found Error: no non-leaf 'build' dir found on target path /z/m5/regression/zizzer/gem5/build Child returned 1 When attemping to execute: scons --ignore-style -k USE_MYSQL=no RUBY=True -j 7 -Q build/ALPHA/gem5.debug build/ALPHA_MOESI_hammer/gem5.debug build/ALPHA_MESI_CMP_directory/gem5.debug build/ALPHA_MOESI_CMP_directory/gem5.debug build/ALPHA_MOESI_CMP_token/gem5.debug build/MIPS/gem5.debug build/POWER/gem5.debug build/SPARC/gem5.debug build/X86/gem5.debug build/ X86_MESI_CMP_directoryARM/gem5.debug build/ALPHA/gem5.fast build/ALPHA_MOESI_hammer/gem5.fast build/ALPHA_MESI_CMP_directory/gem5.fast build/ALPHA_MOESI_CMP_directory/gem5.fast build/ALPHA_MOESI_CMP_token/gem5.fast build/MIPS/gem5.fast build/POWER/gem5.fast build/SPARC/gem5.fast build/X86/gem5.fast build/ X86_MESI_CMP_directoryARM/gem5.fast build/ALPHA/tests/opt/quick/se build/ALPHA/tests/opt/quick/fs build/ALPHA_MOESI_hammer/tests/opt/quick/se build/ALPHA_MOESI_hammer/tests/opt/quick/fs build/ALPHA_MESI_CMP_directory/tests/opt/quick/se build/ALPHA_MESI_CMP_directory/tests/opt/quick/fs build/ALPHA_MOESI_CMP_directory/tests/opt/quick/se build/ALPHA_MOESI_CMP_directory/tests/opt/quick/fs build/ALPHA_MOESI_CMP_token/tests/opt/quick/se build/ALPHA_MOESI_CMP_token/tests/opt/quick/fs build/MIPS/tests/opt/quick/se build/MIPS/tests/opt/quick/fs build/POWER/tests/opt/quick/se build/POWER/tests/opt/quick/fs build/SPARC/tests/opt/quick/se build/X86/tests/opt/quick/se build/X86/tests/opt/quick/fs build/ X86_MESI_CMP_directoryARM/tests/opt/quick/se build/ X86_MESI_CMP_directoryARM/tests/opt/quick/fs Child returned 1 When attemping to execute: util/regress '--scons-opts' '-k USE_MYSQL=no RUBY=True -j 7 -Q' 'quick' On Thu, Apr 26, 2012 at 6:52 AM, Steve Reinhardt <[email protected]> wrote: > Yes, your change to util/regress is bad, you have a comma inside a string > instead of outside. > > You should actually use util/regress to run regressions before you commit, > since that's what the nightly regression script does. Then you would have > caught this beforehand. > > Steve > > On Thu, Apr 26, 2012 at 6:23 AM, Nilay Vaish <[email protected]> wrote: > >> Can some one check what went wrong? It is likely that I committed some >> mistake. >> >> -- >> Nilay >> >> >> On Thu, 26 Apr 2012, Cron Daemon wrote: >> >> >>> See /z/m5/regression/regress-2012-04-26-03:00:**01 for details. >>> >>> ______________________________**_________________ >>> gem5-dev mailing list >>> [email protected] >>> http://m5sim.org/mailman/**listinfo/gem5-dev<http://m5sim.org/mailman/listinfo/gem5-dev> >>> >>> ______________________________**_________________ >> gem5-dev mailing list >> [email protected] >> http://m5sim.org/mailman/**listinfo/gem5-dev<http://m5sim.org/mailman/listinfo/gem5-dev> >> > > ------------------------------ Message: 3 Date: Thu, 26 Apr 2012 09:24:25 -0500 (CDT) From: Nilay Vaish <[email protected]> To: gem5 Developer List <[email protected]> Subject: Re: [gem5-dev] Cron <m5test@zizzer> /z/m5/regression/do-regression quick Message-ID: <[email protected]> Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed I was using util/regress but I was specifying the 'builds' option. Hence I failed to capture this problem. It seems there are two problems. One is the missing is comma, and surprisingly there cannot be a space between ',' and X86_MESI_CMP_directory. -- Nilay On Thu, 26 Apr 2012, Steve Reinhardt wrote: > The bug is in util/regress (see below), but my quick diagnosis was wrong in > the details... I just glanced at the diff and didn't notice that the thing > you changed is really a long string split across several lines and not a > list. I think the actual issue is that you're missing a comma. > > Here's the regression output; note the problems at the end of the scons > args generated by util/regress. > > pulling from http://repo.gem5.org/gem5 > searching for changes > adding changesets > adding manifests > adding file changes > added 5 changesets with 21 changes to 19 files > 19 files updated, 0 files merged, 0 files removed, 0 files unresolved > pulling from http://repo.gem5.org/encumbered > searching for changes > no changes found > Error: no non-leaf 'build' dir found on target path > /z/m5/regression/zizzer/gem5/build > Child returned 1 > When attemping to execute: scons --ignore-style -k USE_MYSQL=no RUBY=True > -j 7 -Q build/ALPHA/gem5.debug build/ALPHA_MOESI_hammer/gem5.debug > build/ALPHA_MESI_CMP_directory/gem5.debug > build/ALPHA_MOESI_CMP_directory/gem5.debug > build/ALPHA_MOESI_CMP_token/gem5.debug build/MIPS/gem5.debug > build/POWER/gem5.debug build/SPARC/gem5.debug build/X86/gem5.debug build/ > X86_MESI_CMP_directoryARM/gem5.debug build/ALPHA/gem5.fast > build/ALPHA_MOESI_hammer/gem5.fast build/ALPHA_MESI_CMP_directory/gem5.fast > build/ALPHA_MOESI_CMP_directory/gem5.fast > build/ALPHA_MOESI_CMP_token/gem5.fast build/MIPS/gem5.fast > build/POWER/gem5.fast build/SPARC/gem5.fast build/X86/gem5.fast build/ > X86_MESI_CMP_directoryARM/gem5.fast build/ALPHA/tests/opt/quick/se > build/ALPHA/tests/opt/quick/fs build/ALPHA_MOESI_hammer/tests/opt/quick/se > build/ALPHA_MOESI_hammer/tests/opt/quick/fs > build/ALPHA_MESI_CMP_directory/tests/opt/quick/se > build/ALPHA_MESI_CMP_directory/tests/opt/quick/fs > build/ALPHA_MOESI_CMP_directory/tests/opt/quick/se > build/ALPHA_MOESI_CMP_directory/tests/opt/quick/fs > build/ALPHA_MOESI_CMP_token/tests/opt/quick/se > build/ALPHA_MOESI_CMP_token/tests/opt/quick/fs > build/MIPS/tests/opt/quick/se build/MIPS/tests/opt/quick/fs > build/POWER/tests/opt/quick/se build/POWER/tests/opt/quick/fs > build/SPARC/tests/opt/quick/se build/X86/tests/opt/quick/se > build/X86/tests/opt/quick/fs build/ > X86_MESI_CMP_directoryARM/tests/opt/quick/se build/ > X86_MESI_CMP_directoryARM/tests/opt/quick/fs > Child returned 1 > When attemping to execute: util/regress '--scons-opts' '-k USE_MYSQL=no > RUBY=True -j 7 -Q' 'quick' > > > On Thu, Apr 26, 2012 at 6:52 AM, Steve Reinhardt <[email protected]> wrote: > >> Yes, your change to util/regress is bad, you have a comma inside a string >> instead of outside. >> >> You should actually use util/regress to run regressions before you commit, >> since that's what the nightly regression script does. Then you would have >> caught this beforehand. >> >> Steve >> >> On Thu, Apr 26, 2012 at 6:23 AM, Nilay Vaish <[email protected]> wrote: >> >>> Can some one check what went wrong? It is likely that I committed some >>> mistake. >>> >>> -- >>> Nilay >>> >>> >>> On Thu, 26 Apr 2012, Cron Daemon wrote: >>> >>> >>>> See /z/m5/regression/regress-2012-04-26-03:00:**01 for details. >>>> >>>> ______________________________**_________________ >>>> gem5-dev mailing list >>>> [email protected] >>>> http://m5sim.org/mailman/**listinfo/gem5-dev<http://m5sim.org/mailman/listinfo/gem5-dev> >>>> >>>> ______________________________**_________________ >>> gem5-dev mailing list >>> [email protected] >>> http://m5sim.org/mailman/**listinfo/gem5-dev<http://m5sim.org/mailman/listinfo/gem5-dev> >>> >> >> > _______________________________________________ > gem5-dev mailing list > [email protected] > http://m5sim.org/mailman/listinfo/gem5-dev > ------------------------------ Message: 4 Date: Thu, 26 Apr 2012 07:38:23 -0700 From: Steve Reinhardt <[email protected]> To: gem5 Developer List <[email protected]> Subject: Re: [gem5-dev] Cron <m5test@zizzer> /z/m5/regression/do-regression quick Message-ID: <cahgmoh99aeutk93prv7fmmkvty+hhtyp0ojwuwsr04hsprw...@mail.gmail.com> Content-Type: text/plain; charset=ISO-8859-1 I see. The lack of tolerance for a space is because you're providing a string value as the default for a command-line parameter (where spaces normally wouldn't be present anyway), not a python list... similar to my original confusion. That said, the split_if_nonempty() method could be tweaked to strip leading spaces if we cared. Steve On Thu, Apr 26, 2012 at 7:24 AM, Nilay Vaish <[email protected]> wrote: > I was using util/regress but I was specifying the 'builds' option. Hence I > failed to capture this problem. It seems there are two problems. One is the > missing is comma, and surprisingly there cannot be a space between ',' and > X86_MESI_CMP_directory. > > -- > Nilay > > > On Thu, 26 Apr 2012, Steve Reinhardt wrote: > > The bug is in util/regress (see below), but my quick diagnosis was wrong >> in >> the details... I just glanced at the diff and didn't notice that the thing >> you changed is really a long string split across several lines and not a >> list. I think the actual issue is that you're missing a comma. >> >> Here's the regression output; note the problems at the end of the scons >> args generated by util/regress. >> >> pulling from http://repo.gem5.org/gem5 >> searching for changes >> adding changesets >> adding manifests >> adding file changes >> added 5 changesets with 21 changes to 19 files >> 19 files updated, 0 files merged, 0 files removed, 0 files unresolved >> pulling from >> http://repo.gem5.org/**encumbered<http://repo.gem5.org/encumbered> >> searching for changes >> no changes found >> Error: no non-leaf 'build' dir found on target path >> /z/m5/regression/zizzer/gem5/**build >> Child returned 1 >> When attemping to execute: scons --ignore-style -k USE_MYSQL=no RUBY=True >> -j 7 -Q build/ALPHA/gem5.debug build/ALPHA_MOESI_hammer/gem5.**debug >> build/ALPHA_MESI_CMP_**directory/gem5.debug >> build/ALPHA_MOESI_CMP_**directory/gem5.debug >> build/ALPHA_MOESI_CMP_token/**gem5.debug build/MIPS/gem5.debug >> build/POWER/gem5.debug build/SPARC/gem5.debug build/X86/gem5.debug build/ >> X86_MESI_CMP_directoryARM/**gem5.debug build/ALPHA/gem5.fast >> build/ALPHA_MOESI_hammer/gem5.**fast build/ALPHA_MESI_CMP_** >> directory/gem5.fast >> build/ALPHA_MOESI_CMP_**directory/gem5.fast >> build/ALPHA_MOESI_CMP_token/**gem5.fast build/MIPS/gem5.fast >> build/POWER/gem5.fast build/SPARC/gem5.fast build/X86/gem5.fast build/ >> X86_MESI_CMP_directoryARM/**gem5.fast build/ALPHA/tests/opt/quick/se >> build/ALPHA/tests/opt/quick/fs build/ALPHA_MOESI_hammer/** >> tests/opt/quick/se >> build/ALPHA_MOESI_hammer/**tests/opt/quick/fs >> build/ALPHA_MESI_CMP_**directory/tests/opt/quick/se >> build/ALPHA_MESI_CMP_**directory/tests/opt/quick/fs >> build/ALPHA_MOESI_CMP_**directory/tests/opt/quick/se >> build/ALPHA_MOESI_CMP_**directory/tests/opt/quick/fs >> build/ALPHA_MOESI_CMP_token/**tests/opt/quick/se >> build/ALPHA_MOESI_CMP_token/**tests/opt/quick/fs >> build/MIPS/tests/opt/quick/se build/MIPS/tests/opt/quick/fs >> build/POWER/tests/opt/quick/se build/POWER/tests/opt/quick/fs >> build/SPARC/tests/opt/quick/se build/X86/tests/opt/quick/se >> build/X86/tests/opt/quick/fs build/ >> X86_MESI_CMP_directoryARM/**tests/opt/quick/se build/ >> X86_MESI_CMP_directoryARM/**tests/opt/quick/fs >> Child returned 1 >> When attemping to execute: util/regress '--scons-opts' '-k USE_MYSQL=no >> RUBY=True -j 7 -Q' 'quick' >> >> >> On Thu, Apr 26, 2012 at 6:52 AM, Steve Reinhardt <[email protected]> >> wrote: >> >> Yes, your change to util/regress is bad, you have a comma inside a string >>> instead of outside. >>> >>> You should actually use util/regress to run regressions before you >>> commit, >>> since that's what the nightly regression script does. Then you would >>> have >>> caught this beforehand. >>> >>> Steve >>> >>> On Thu, Apr 26, 2012 at 6:23 AM, Nilay Vaish <[email protected]> wrote: >>> >>> Can some one check what went wrong? It is likely that I committed some >>>> mistake. >>>> >>>> -- >>>> Nilay >>>> >>>> >>>> On Thu, 26 Apr 2012, Cron Daemon wrote: >>>> >>>> >>>> See /z/m5/regression/regress-2012-04-26-03:00:****01 for details. >>>>> >>>>> ______________________________****_________________ >>>>> gem5-dev mailing list >>>>> [email protected] >>>>> http://m5sim.org/mailman/****listinfo/gem5-dev<http://m5sim.org/mailman/**listinfo/gem5-dev> >>>>> <http://**m5sim.org/mailman/listinfo/**gem5-dev<http://m5sim.org/mailman/listinfo/gem5-dev> >>>>> > >>>>> >>>>> ______________________________****_________________ >>>>> >>>> gem5-dev mailing list >>>> [email protected] >>>> http://m5sim.org/mailman/****listinfo/gem5-dev<http://m5sim.org/mailman/**listinfo/gem5-dev> >>>> <http://**m5sim.org/mailman/listinfo/**gem5-dev<http://m5sim.org/mailman/listinfo/gem5-dev> >>>> > >>>> >>>> >>> >>> ______________________________**_________________ >> gem5-dev mailing list >> [email protected] >> http://m5sim.org/mailman/**listinfo/gem5-dev<http://m5sim.org/mailman/listinfo/gem5-dev> >> >> ______________________________**_________________ > gem5-dev mailing list > [email protected] > http://m5sim.org/mailman/**listinfo/gem5-dev<http://m5sim.org/mailman/listinfo/gem5-dev> > ------------------------------ Message: 5 Date: Thu, 26 Apr 2012 15:04:28 -0000 From: "Andreas Hansson" <[email protected]> To: "Andreas Hansson" <[email protected]>, "Default" <[email protected]> Subject: Re: [gem5-dev] Review Request: MEM: Add the communication monitor Message-ID: <[email protected]> Content-Type: text/plain; charset="utf-8" ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://reviews.gem5.org/r/1159/ ----------------------------------------------------------- (Updated April 26, 2012, 8:04 a.m.) Review request for Default. Description ------- MEM: Add the communication monitor This patch adds a communication monitor MemObject that can be inserted between a master and slave port to provide a range of statistics about the communication passing through it. The communication monitor is non-invasive and does not change any properties or timing of the packets, with the exception of adding a sender state to be able to track latency. The statistics are only collected in timing mode (not atomic) to avoid slowing down any fast forwarding. An example of the statistics captured by the monitor are: read/write burst lengths, bandwidth, request-response latency, outstanding transactions, inter transaction time, transaction count, and address distribution. The monitor can be used in combination with periodic resetting and dumping of stats (through schedStatEvent) to study the behaviour over time. In future patches, a selection of convenience scripts will be added to aid in visualising the statistics collected by the monitor. Diffs (updated) ----- src/mem/CommMonitor.py PRE-CREATION src/mem/SConscript 6d11b01e2c53 src/mem/comm_monitor.hh PRE-CREATION src/mem/comm_monitor.cc PRE-CREATION Diff: http://reviews.gem5.org/r/1159/diff/ Testing ------- Used in combination with various full system simulations Thanks, Andreas Hansson ------------------------------ _______________________________________________ gem5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/gem5-dev End of gem5-dev Digest, Vol 60, Issue 97 **************************************** _______________________________________________ gem5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/gem5-dev
