Re: [gem5-users] Compiling stats at multi-intervals.

[Surya Narayanan]

I see that we can dump statistics in certain intervals i,e periodically. 

In /m5/src/sim/pseudo_inst.hh
void dumpstats(ThreadContext *tc, Tick delay, Tick period);

In /m5/util/m5/m5.c
 { "dumpstats",      do_dump_stats,       "[delay [period]]" },

But I don't know how to enable this in gem5. I tried modifying the .rcs script 
like the following by passing the parameter but the program executed the same 
way.

/sbin/m5 dumpstats[10000]
/sbin/m5 resetstats

Is there any command line technique to do this? 

If I stat dump at regular intervals will there be multiple stat.txt file formed 
in m5out folder?



----- Original Message -----
> From: gem5-users-requ...@gem5.org
> To: gem5-users@gem5.org
> Sent: Tuesday, November 15, 2011 5:10:47 AM
> Subject: gem5-users Digest, Vol 64, Issue 30
> Send gem5-users mailing list submissions to
> gem5-users@gem5.org
> 
> To subscribe or unsubscribe via the World Wide Web, visit
> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
> or, via email, send a message with subject or body 'help' to
> gem5-users-requ...@gem5.org
> 
> You can reach the person managing the list at
> gem5-users-ow...@gem5.org
> 
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of gem5-users digest..."
> 
> 
> Today's Topics:
> 
> 1. Re: thread-pthread,pid,contex_switch (Steve Reinhardt)
> 2. SPEC 2006 Benchmarks Configuration Script (Marios Michael)
> 3. statistics of cpu idle intervals distribution of PERSEC
> Benchmarks (Tong Xu)
> 4. Re: thread-pthread,pid,contex_switch (kopar...@uth.gr)
> 5. Compile problem with gcc4.6.2 and swig2.0.2 (yanke)
> 6. Re: Effect of Operating System policies -FS or SE mode?
> (Hasina Khatoon)
> 
> 
> ----------------------------------------------------------------------
> 
> Message: 1
> Date: Mon, 14 Nov 2011 10:33:09 -0800
> From: Steve Reinhardt <ste...@gmail.com>
> To: gem5 users mailing list <gem5-users@gem5.org>
> Subject: Re: [gem5-users] thread-pthread,pid,contex_switch
> Message-ID:
> <CAHgMoh8=mritgyyj+newg2ojgdc3fnojgf+go4kjatapqwt...@mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
> 
> LiveProcess is only used in SE mode, not FS mode.
> 
> I don't understand the rest of your question.
> 
> Steve
> 
> On Mon, Nov 14, 2011 at 2:55 AM, <kopar...@uth.gr> wrote:
> 
> > Hi,
> > I dont know if i am refering to the right list. I m writing a module
> > on
> > ALPHA_FS
> > and i need to know if there is a connection between the thread's on
> > the
> > cpu and
> > the thread's of the pthread library.
> > Moreover is the varable on clas::LiveProcess pid unique to each
> > process on
> > the
> > system? Is there any function that is called on the contex switch?
> > Thanc you
> > _______________________________________________
> > gem5-users mailing list
> > gem5-users@gem5.org
> > http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
> >
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> <http://m5sim.org/cgi-bin/mailman/private/gem5-users/attachments/20111114/195edff6/attachment-0001.html>
> 
> ------------------------------
> 
> Message: 2
> Date: Mon, 14 Nov 2011 22:39:44 +0200
> From: Marios Michael <mmich...@cs.ucy.ac.cy>
> To: gem5-users@gem5.org
> Cc: Marios Michael <mmich...@cs.ucy.ac.cy>
> Subject: [gem5-users] SPEC 2006 Benchmarks Configuration Script
> Message-ID: <2cd4e628-ebc3-44e3-ba2d-02fb8f3d4...@cs.ucy.ac.cy>
> Content-Type: text/plain; charset="us-ascii"
> 
> I am trying to run the SPEC 2006 benchmarks on se mode without
> success. Basically i have downloaded the system configuration python
> file for the M5 SE mode "#cmp.py" and when i run it in order to check
> if its ok i have the following error message:
> 
> 
> 
> elsi@ubuntu:~/gem5-stable$ ./build/ALPHA_SE/m5.opt
> configs/example/cmp.py
> gem5 Simulator System. http://gem5.org
> gem5 is copyrighted software; use the --copyright option for details.
> 
> gem5 compiled Nov 14 2011 14:03:25
> gem5 started Nov 14 2011 22:29:16
> gem5 executing on ubuntu
> command line: ./build/ALPHA_SE/m5.opt configs/example/cmp.py
> Traceback (most recent call last):
> File "<string>", line 1, in <module>
> File "/home/elsi/gem5-stable/src/python/m5/main.py", line 356, in main
> exec filecode in scope
> File "configs/example/cmp.py", line 7, in <module>
> m5.AddToPath('./configs')
> AttributeError: 'module' object has no attribute 'AddToPath'
> 
> 
> 
> 
> Can anyone help me with this or give me some instructions in order to
> run the benchmarks?
> 
> I am new here
> 
> Thanks
> 
> Best regards,
> 
> Marios Michael
> mmich...@cs.ucy.ac.cy
> 
> 
> 
> 
> 
> 
> 
> 
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> <http://m5sim.org/cgi-bin/mailman/private/gem5-users/attachments/20111114/a5951a3c/attachment-0001.html>
> 
> ------------------------------
> 
> Message: 3
> Date: Mon, 14 Nov 2011 15:07:46 -0600
> From: Tong Xu <xuton...@gmail.com>
> To: gem5-users@gem5.org
> Subject: [gem5-users] statistics of cpu idle intervals distribution of
> PERSEC Benchmarks
> Message-ID:
> <cac4bm4c2o4yfxkgabj7a-nynrzju83nvtxcwu9rf1gxxjdy...@mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
> 
> Hi, Folks
> 
> I need to get the statistics of idle cycels distribution of PERSEC
> Benchmarks. For example, I want to know how many CPU idle intervals
> are
> shorter than 100 cycles and how many idle intervals are shorter than
> 200
> cycles and longer than 100 cycles.
> 
> I have built up and install full system files according to the
> instructions
> from University of Texas at Austin (
> http://www.cs.utexas.edu/~parsec_m5/). I have noticed that in
> stats.txt there are some statistic results about
> cpu idle cycles and idle fraction. But they cannot give the
> distribution
> information.
> 
> How can I get the idle intervals distribution? Which source files
> should I
> modify?
> 
> Thank you
> Regards
> 
> Tong
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> <http://m5sim.org/cgi-bin/mailman/private/gem5-users/attachments/20111114/e862042a/attachment-0001.html>
> 
> ------------------------------
> 
> Message: 4
> Date: Tue, 15 Nov 2011 01:25:42 +0200
> From: kopar...@uth.gr
> To: gem5-users@gem5.org
> Subject: Re: [gem5-users] thread-pthread,pid,contex_switch
> Message-ID: <20111115012542.if3pwm0en90gc...@webmail.uth.gr>
> Content-Type: text/plain; charset=UTF-8; format="flowed"
> 
> If so then on class process there is an variavle M5_pid is this
> variable unique
> to each process running on FS.
> 
> Now concernig my question about pthreads maybe an example will help :
> If i have a program using 2 pthreads and my ALPHA system simulates a 2
> core cpu.
> If pthread 1 is running on core 1 and pthread 2 is running on core 2
> is it
> possible for pthread2 to run on core 1? If yes is there anyway to
> indentify
> different pthreads of the same application while executing the
> application?
> 
> Now about the contex switch assume we have a workload with 2
> applications
> running on the ALPHA single-core system.If for some reason application
> 1 is
> running and a contex switch is happening can i understand that by some
> event or
> something like that
> 
> Thank you
> 
> 
> Quoting Steve Reinhardt <ste...@gmail.com>:
> 
> > LiveProcess is only used in SE mode, not FS mode.
> >
> > I don't understand the rest of your question.
> >
> > Steve
> >
> > On Mon, Nov 14, 2011 at 2:55 AM, <kopar...@uth.gr> wrote:
> >
> >> Hi,
> >> I dont know if i am refering to the right list. I m writing a
> >> module on
> >> ALPHA_FS
> >> and i need to know if there is a connection between the thread's on
> >> the
> >> cpu and
> >> the thread's of the pthread library.
> >> Moreover is the varable on clas::LiveProcess pid unique to each
> >> process on
> >> the
> >> system? Is there any function that is called on the contex switch?
> >> Thanc you
> >> _______________________________________________
> >> gem5-users mailing list
> >> gem5-users@gem5.org
> >> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
> >>
> >
> 
> 
> 
> 
> ------------------------------
> 
> Message: 5
> Date: Tue, 15 Nov 2011 11:22:19 +0800 (CST)
> From: yanke <ykya...@126.com>
> To: gem5-users@gem5.org
> Subject: [gem5-users] Compile problem with gcc4.6.2 and swig2.0.2
> Message-ID: <7c8ca039.4ccd.133a53f28e1.coremail.ykya...@126.com>
> Content-Type: text/plain; charset="gbk"
> 
> Hello? When I compiling gem5 with gcc4.6.2 and swig2.0.2 ?I get the
> following errors?
> 
> [ CXX] ALPHA_FS/python/swig/pyobject.cc -> .do
> build/ALPHA_FS/python/swig/pyobject.cc: In function 'int
> connectPorts(SimObject*, const string&, int, SimObject*, const
> string&, int)':
> build/ALPHA_FS/python/swig/pyobject.cc:97:16: error: variable 'mo1'
> set but not used [-Werror=unused-but-set-variable]
> build/ALPHA_FS/python/swig/pyobject.cc:97:22: error: variable 'mo2'
> set but not used [-Werror=unused-but-set-variable]
> 
> Could anyone help me?
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> <http://m5sim.org/cgi-bin/mailman/private/gem5-users/attachments/20111115/c4374bcf/attachment-0001.html>
> 
> ------------------------------
> 
> Message: 6
> Date: Tue, 15 Nov 2011 09:11:01 +0500
> From: Hasina Khatoon <hasina.khat...@nu.edu.pk>
> To: gem5-users@gem5.org
> Subject: Re: [gem5-users] Effect of Operating System policies -FS or
> SE mode?
> Message-ID:
> <CAGfJSkuAqQDJN8RCXvT2SAQ=m8+ms8a254raoursuyja4-h...@mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
> 
> Thanks. Can you please let me know how I can access the OS source
> files of gem5? I am unable to access the source code for the OS.
> 
> Hasina Khatoon
> 
> On Mon, Nov 14, 2011 at 5:01 PM, <gem5-users-requ...@gem5.org> wrote:
> > Send gem5-users mailing list submissions to
> > ? ? ? ?gem5-users@gem5.org
> >
> > To subscribe or unsubscribe via the World Wide Web, visit
> > ? ? ? ?http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
> > or, via email, send a message with subject or body 'help' to
> > ? ? ? ?gem5-users-requ...@gem5.org
> >
> > You can reach the person managing the list at
> > ? ? ? ?gem5-users-ow...@gem5.org
> >
> > When replying, please edit your Subject line so it is more specific
> > than "Re: Contents of gem5-users digest..."
> >
> >
> > Today's Topics:
> >
> > ? 1. Effect of Operating System policies -FS or SE mode?
> > ? ? ?(Hasina Khatoon)
> > ? 2. Re: Effect of Operating System policies -FS or SE mode? (Nilay)
> > ? 3. Re: Guidance on modifying ruby code in Gem5 (Nilay Vaish)
> > ? 4. Re: Problem with building ALPHA FS (kopar...@uth.gr)
> > ? 5. Re: Problem with building ALPHA FS (kopar...@uth.gr)
> > ? 6. thread-pthread,pid,contex_switch (kopar...@uth.gr)
> > ? 7. when prefetcher is called? (Mahmood Naderan)
> > ? 8. Re: when prefetcher is called? (biswabandan panda)
> > ? 9. Re: when prefetcher is called? (Mahmood Naderan)
> >
> >
> > ----------------------------------------------------------------------
> >
> > Message: 1
> > Date: Mon, 14 Nov 2011 11:30:47 +0500
> > From: Hasina Khatoon <hasina.khat...@nu.edu.pk>
> > To: gem5-users@gem5.org
> > Subject: [gem5-users] Effect of Operating System policies -FS or SE
> > ? ? ? ?mode?
> > Message-ID:
> > ? ? ?
> > ?<cagfjsktui8_wad2f_tkwfwrocb_4+ypc0xjhfdynq-7oln3...@mail.gmail.com>
> > Content-Type: text/plain; charset=ISO-8859-1
> >
> > I would like to run a few experiments and observe cache performance
> > after changing the operating system features like page mapping,
> > scheduling,etc. What is a better option to use? FS or SE mode for
> > gem5?
> >
> > Thanks in advance
> >
> > Hasina Khatoon
> >
> > --
> > Department of Computer Science
> > National University of Computer and Emerging Sciences (FAST-NU)
> > Karachi Campus (Main)
> >
> >
> > ------------------------------
> >
> > Message: 2
> > Date: Mon, 14 Nov 2011 00:38:44 -0600
> > From: "Nilay" <ni...@cs.wisc.edu>
> > To: "gem5 users mailing list" <gem5-users@gem5.org>
> > Subject: Re: [gem5-users] Effect of Operating System policies -FS or
> > ? ? ? ?SE mode?
> > Message-ID:
> > ? ? ?
> > ?<9ada347bf832bc7cdb5c1e238b3a8643.squir...@webmail.cs.wisc.edu>
> > Content-Type: text/plain;charset=iso-8859-1
> >
> > On Mon, November 14, 2011 12:30 am, Hasina Khatoon wrote:
> >> I would like to run a few experiments and observe cache performance
> >> after changing the operating system features like page mapping,
> >> scheduling,etc. What is a better option to use? FS or SE mode for
> >> gem5?
> >>
> >
> > Certainly the FS mode. In SE mode, the OS is not running.
> >
> > --
> > Nilay
> >
> >
> >
> > ------------------------------
> >
> > Message: 3
> > Date: Mon, 14 Nov 2011 01:14:58 -0600 (CST)
> > From: Nilay Vaish <ni...@cs.wisc.edu>
> > To: gem5 users mailing list <gem5-users@gem5.org>
> > Subject: Re: [gem5-users] Guidance on modifying ruby code in Gem5
> > Message-ID: <pine.lnx.4.64.1111140108180.32...@ribera.cs.wisc.edu>
> > Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed
> >
> > First a disclaimer, I might wrong in my understanding of gem5. But I
> > will still try to answer your questions. The answers are inline.
> >
> > On Fri, 11 Nov 2011, tejasi pimpalkhute wrote:
> >
> >> Hi Team,
> >>
> >> We are trying to implement An SDRAM aware flow control on Gem5 for
> >> which we
> >> need to add some intelligence(i.e., the router schedules the
> >> requests based
> >> on Row Address and Bank Address) to the router and thus reduce the
> >> complexity of Memory controller in the Memory Subsystem.
> >>
> >> It would be great if some one can give us insight about how to
> >> start making
> >> changes in Gem5, what are things to be considered when we try to
> >> create a
> >> new variable or new code file? For example, we would like to
> >> schedule the
> >> memory requests based on priority for which we have to add a new
> >> variable
> >> that specifies it.
> >
> > I would say, just take a stab at changing the code and adding your
> > own
> > stuff. In most cases, it should just work out. There is nothing
> > special
> > that you would need to do. In case you face some particular problem,
> > you
> > can always ask more to the point questions.
> >
> >>
> >> At the router how do we distinguish between a memory request and a
> >> non-Memory request based on the destination address i.e., what is
> >> the
> >> destination address of memory?
> >>
> > If I understand correctly, non-memory requests actually do not flow
> > through Ruby's network. So I don't think a router would get to see
> > an
> > address that does not belong to the memory.
> >
> > --
> > Nilay
> >
> >
> > ------------------------------
> >
> > Message: 4
> > Date: Mon, 14 Nov 2011 11:41:40 +0200
> > From: kopar...@uth.gr
> > To: gem5-users@gem5.org
> > Subject: Re: [gem5-users] Problem with building ALPHA FS
> > Message-ID: <20111114114140.k6vy3evkihwgg...@webmail.uth.gr>
> > Content-Type: text/plain; ? ? ? charset=UTF-8; ?format="flowed"
> >
> > I have installed build-essential but same problem occurs.Any other
> > ideas?
> >
> > Quoting Gabe Black <gbl...@eecs.umich.edu>:
> >
> >> I'm assuming you're on Ubuntu? Try installing build-essential. It
> >> might
> >> be named slightly differently.
> >>
> >> Gabe
> >>
> >> On 11/13/11 03:31, kopar...@uth.gr wrote:
> >>> Hi,
> >>> I'm trying to compile ALPHA_FS and im getting this error message :
> >>> Error: Don't know what compiler options to use for your compiler.
> >>> ? ? ? ?Please fix SConstruct and src/SConscript and try again.
> >>> I m using gcc 4.5.2 and python 2.7.1+
> >>> 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
> >>
> >
> >
> >
> >
> > ------------------------------
> >
> > Message: 5
> > Date: Mon, 14 Nov 2011 12:28:10 +0200
> > From: kopar...@uth.gr
> > To: gem5-users@gem5.org
> > Subject: Re: [gem5-users] Problem with building ALPHA FS
> > Message-ID: <20111114122810.n00kn5d2500cc...@webmail.uth.gr>
> > Content-Type: text/plain; ? ? ? charset=UTF-8; ?format="flowed"
> >
> > problem solved by running : scons CC=gcc CXX=g++
> > build/ALPHA_FS/gem5.debug
> >
> > Quoting kopar...@uth.gr:
> >
> >> I have installed build-essential but same problem occurs.Any other
> >> ideas?
> >>
> >> Quoting Gabe Black <gbl...@eecs.umich.edu>:
> >>
> >>> I'm assuming you're on Ubuntu? Try installing build-essential. It
> >>> might
> >>> be named slightly differently.
> >>>
> >>> Gabe
> >>>
> >>> On 11/13/11 03:31, kopar...@uth.gr wrote:
> >>>> Hi,
> >>>> I'm trying to compile ALPHA_FS and im getting this error message
> >>>> :
> >>>> Error: Don't know what compiler options to use for your compiler.
> >>>> ? ? ? ?Please fix SConstruct and src/SConscript and try again.
> >>>> I m using gcc 4.5.2 and python 2.7.1+
> >>>> 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 mailing list
> >> gem5-users@gem5.org
> >> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
> >>
> >
> >
> >
> >
> > ------------------------------
> >
> > Message: 6
> > Date: Mon, 14 Nov 2011 12:55:38 +0200
> > From: kopar...@uth.gr
> > To: gem5-users@gem5.org
> > Subject: [gem5-users] thread-pthread,pid,contex_switch
> > Message-ID: <20111114125538.5ud7rfr5d2dcw...@webmail.uth.gr>
> > Content-Type: text/plain; ? ? ? charset=UTF-8
> >
> > Hi,
> > I dont know if i am refering to the right list. I m writing a module
> > on ALPHA_FS
> > and i need to know if there is a connection between the thread's on
> > the cpu and
> > the thread's of the pthread library.
> > Moreover is the varable on clas::LiveProcess pid unique to each
> > process on the
> > system? Is there any function that is called on the contex switch?
> > Thanc you
> >
> >
> > ------------------------------
> >
> > Message: 7
> > Date: Mon, 14 Nov 2011 15:07:21 +0330
> > From: Mahmood Naderan <mahmood...@gmail.com>
> > To: gem5 users mailing list <gem5-users@gem5.org>
> > Subject: [gem5-users] when prefetcher is called?
> > Message-ID:
> > ? ? ?
> > ?<cada2p2wxlczgrzhsoqvmyahec5-cj9js8c32mtqhxn+18b5...@mail.gmail.com>
> > Content-Type: text/plain; charset="iso-8859-1"
> >
> > Hi,
> > Just for a double check, is this statement true?
> >
> > The prefetcher is called on of the following cases:
> > 1- cache block is accesses
> > or
> > 2- that block was a prefetched block
> > or
> > 3- cache miss
> >
> > I said that because of the following code.
> >
> > if (satisfied) {
> > ? ? ? ?if (prefetcher && (prefetchOnAccess || (blk &&
> > blk->wasPrefetched()))) {
> > ? ? ? ? ? ?if (blk)
> > ? ? ? ? ? ? ? ?blk->status &= ~BlkHWPrefetched;
> > ? ? ? ? ? ?next_pf_time = prefetcher->notify(pkt, time);
> > ? ? ? ?}
> > ? ? ? ?....
> > } else {
> > ? ? ? ?// miss
> > ? ? ? ?if (mshr) {
> > ? ? ? ? .....
> > ? ? ? ?} else {
> > ? ? ? ? ? ?// no MSHR
> > ? ? ? ? ? ? .....
> > ? ? ? ? ? ?if (prefetcher) {
> > ? ? ? ? ? ? ? ?next_pf_time = prefetcher->notify(pkt, time);
> > ? ? ? ? ? ?}
> > ? ? ? ?}
> > }
> >
> >
> > Thanks and sorry for taking up your time
> > --
> > // Naderan *Mahmood;
> > -------------- next part --------------
> > An HTML attachment was scrubbed...
> > URL:
> > <http://m5sim.org/cgi-bin/mailman/private/gem5-users/attachments/20111114/142914fb/attachment-0001.html>
> >
> > ------------------------------
> >
> > Message: 8
> > Date: Mon, 14 Nov 2011 17:18:47 +0530
> > From: biswabandan panda <biswa....@gmail.com>
> > To: gem5 users mailing list <gem5-users@gem5.org>
> > Subject: Re: [gem5-users] when prefetcher is called?
> > Message-ID:
> > ? ? ?
> > ?<CADFuLAtDgRNZ_qTrXJtQx+2rngwU25dm=sf4-ekjjjcviar...@mail.gmail.com>
> > Content-Type: text/plain; charset="iso-8859-1"
> >
> > there is option to make it run only for misses check in the pythin
> > files of
> > cache and prefetch.
> >
> > On Mon, Nov 14, 2011 at 5:07 PM, Mahmood Naderan
> > <mahmood...@gmail.com>wrote:
> >
> >> Hi,
> >> Just for a double check, is this statement true?
> >>
> >> The prefetcher is called on of the following cases:
> >> 1- cache block is accesses
> >> or
> >> 2- that block was a prefetched block
> >> or
> >> 3- cache miss
> >>
> >> I said that because of the following code.
> >>
> >> if (satisfied) {
> >> ? ? ? ? if (prefetcher && (prefetchOnAccess || (blk &&
> >> blk->wasPrefetched()))) {
> >> ? ? ? ? ? ? if (blk)
> >> ? ? ? ? ? ? ? ? blk->status &= ~BlkHWPrefetched;
> >> ? ? ? ? ? ? next_pf_time = prefetcher->notify(pkt, time);
> >> ? ? ? ? }
> >> ? ? ? ? ....
> >> } else {
> >> ? ? ? ? // miss
> >> ? ? ? ? if (mshr) {
> >> ? ? ? ? ?.....
> >> ? ? ? ? } else {
> >> ? ? ? ? ? ? // no MSHR
> >> ? ? ? ? ? ? ?.....
> >> ? ? ? ? ? ? if (prefetcher) {
> >> ? ? ? ? ? ? ? ? next_pf_time = prefetcher->notify(pkt, time);
> >> ? ? ? ? ? ? }
> >> ? ? ? ? }
> >> }
> >>
> >>
> >> Thanks and sorry for taking up your time
> >> --
> >> // Naderan *Mahmood;
> >>
> >> _______________________________________________
> >> gem5-users mailing list
> >> gem5-users@gem5.org
> >> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
> >>
> >
> >
> >
> > --
> >
> > *thanks&regards
> > *
> > *BISWABANDAN*
> > -------------- next part --------------
> > An HTML attachment was scrubbed...
> > URL:
> > <http://m5sim.org/cgi-bin/mailman/private/gem5-users/attachments/20111114/f3deb727/attachment-0001.html>
> >
> > ------------------------------
> >
> > Message: 9
> > Date: Mon, 14 Nov 2011 15:31:18 +0330
> > From: Mahmood Naderan <mahmood...@gmail.com>
> > To: gem5 users mailing list <gem5-users@gem5.org>
> > Subject: Re: [gem5-users] when prefetcher is called?
> > Message-ID:
> > ? ? ?
> > ?<cada2p2urkebsg25d8mbipmdmuoz4pd1rg1meu+g_o3ijk8v...@mail.gmail.com>
> > Content-Type: text/plain; charset="iso-8859-1"
> >
> > you mean
> > ? ?prefetch_on_access = Param.Bool(False,
> > ? ? ? ? "notify the hardware prefetcher on every access (not just
> > misses)")
> >
> > If that is False (which is default) then prefetch is done on misses.
> > However if it is True, then prefetch is done on
> >
> > (access AND miss) OR (access) OR (miss)
> >
> > --
> > // Naderan *Mahmood;
> >
> >
> > On Mon, Nov 14, 2011 at 3:18 PM, biswabandan panda
> > <biswa....@gmail.com>wrote:
> >
> >> there is option to make it run only for misses check in the pythin
> >> files
> >> of cache and prefetch.
> >>
> >> On Mon, Nov 14, 2011 at 5:07 PM, Mahmood Naderan
> >> <mahmood...@gmail.com>wrote:
> >>
> >>> Hi,
> >>> Just for a double check, is this statement true?
> >>>
> >>> The prefetcher is called on of the following cases:
> >>> 1- cache block is accesses
> >>> or
> >>> 2- that block was a prefetched block
> >>> or
> >>> 3- cache miss
> >>>
> >>> I said that because of the following code.
> >>>
> >>> if (satisfied) {
> >>> ? ? ? ? if (prefetcher && (prefetchOnAccess || (blk &&
> >>> blk->wasPrefetched()))) {
> >>> ? ? ? ? ? ? if (blk)
> >>> ? ? ? ? ? ? ? ? blk->status &= ~BlkHWPrefetched;
> >>> ? ? ? ? ? ? next_pf_time = prefetcher->notify(pkt, time);
> >>> ? ? ? ? }
> >>> ? ? ? ? ....
> >>> } else {
> >>> ? ? ? ? // miss
> >>> ? ? ? ? if (mshr) {
> >>> ? ? ? ? ?.....
> >>> ? ? ? ? } else {
> >>> ? ? ? ? ? ? // no MSHR
> >>> ? ? ? ? ? ? ?.....
> >>> ? ? ? ? ? ? if (prefetcher) {
> >>> ? ? ? ? ? ? ? ? next_pf_time = prefetcher->notify(pkt, time);
> >>> ? ? ? ? ? ? }
> >>> ? ? ? ? }
> >>> }
> >>>
> >>>
> >>> Thanks and sorry for taking up your time
> >>> --
> >>> // Naderan *Mahmood;
> >>>
> >>> _______________________________________________
> >>> gem5-users mailing list
> >>> gem5-users@gem5.org
> >>> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
> >>>
> >>
> >>
> >>
> >> --
> >>
> >> *thanks&regards
> >> *
> >> *BISWABANDAN*
> >>
> >>
> >>
> >>
> >>
> >>
> >> _______________________________________________
> >> gem5-users mailing list
> >> gem5-users@gem5.org
> >> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
> >>
> > -------------- next part --------------
> > An HTML attachment was scrubbed...
> > URL:
> > <http://m5sim.org/cgi-bin/mailman/private/gem5-users/attachments/20111114/208fbc3d/attachment.html>
> >
> > ------------------------------
> >
> > _______________________________________________
> > gem5-users mailing list
> > gem5-users@gem5.org
> > http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
> >
> > End of gem5-users Digest, Vol 64, Issue 29
> > ******************************************
> >
> 
> 
> 
> --
> Department of Computer Science
> National University of Computer and Emerging Sciences (FAST-NU)
> Karachi Campus (Main)
> 
> 
> ------------------------------
> 
> _______________________________________________
> gem5-users mailing list
> gem5-users@gem5.org
> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
> 
> End of gem5-users Digest, Vol 64, Issue 30
> ******************************************

-- 
Regards,
Surya
_______________________________________________
gem5-users mailing list
gem5-users@gem5.org
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users