Hi Deepak Thanks very much for your help. It really works. But what's "do_interrupt" for?
Best regards. 2014-02-09 1:00 GMT+08:00 <[email protected]>: > Send Marss86-Devel mailing list submissions to > [email protected] > > To subscribe or unsubscribe via the World Wide Web, visit > https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel > 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 Marss86-Devel digest..." > > > Today's Topics: > > 1. Re: Can the MARSS86 be trace driven? (John, Johnsy) > 2. Cloud, Hadoop, Facebook, DB, Web browser (Sun Spyder, > Chrome), SpecJVM, SpecWeb benchmarks on MARSS86? (John, Johnsy) > 3. Re: Can the MARSS86 be trace driven? (Brendan Fitzgerald) > 4. Re: Cloud, Hadoop, Facebook, DB, Web browser (Sun Spyder, > Chrome), SpecJVM, SpecWeb benchmarks on MARSS86? > ([email protected]) > 5. How are system calls realized in MARSSx86 ? (beilei sun) > 6. Re: How are system calls realized in MARSSx86 ? (Deepak Kathayat) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Fri, 7 Feb 2014 19:28:12 +0000 > From: "John, Johnsy" <[email protected]> > To: "[email protected]" > <[email protected]>, > "[email protected]" > <[email protected]> > Subject: Re: [marss86-devel] Can the MARSS86 be trace driven? > Message-ID: > <[email protected]> > Content-Type: text/plain; charset="us-ascii" > > Could someone on the development side, please comment on this? > Thanks > Johnsy > > From: John, Johnsy > Sent: Monday, February 03, 2014 2:08 PM > To: [email protected] > Cc: John, Johnsy > Subject: Can the MARSS86 be trace driven? > > Hi, > I have a question regarding the modes (execution driven, trace driven) in > which MARSS86 can be run. > Can it be trace driven? > If so, please provide the trace format. > > Thanks > Johnsy > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: < > https://www.cs.binghamton.edu/mailman/private/marss86-devel/attachments/20140207/ae601885/attachment-0001.html > > > > ------------------------------ > > Message: 2 > Date: Fri, 7 Feb 2014 19:34:31 +0000 > From: "John, Johnsy" <[email protected]> > To: "[email protected]" > <[email protected]>, > "[email protected]" > <[email protected]> > Subject: [marss86-devel] Cloud, Hadoop, Facebook, DB, Web browser (Sun > Spyder, Chrome), SpecJVM, SpecWeb benchmarks on MARSS86? > Message-ID: > <[email protected]> > Content-Type: text/plain; charset="us-ascii" > > Hi All, > I have a question regarding the various new server benchmarks. > Did anyone try Cloud, Hadoop, Facebook, DB, Web browser (Sun Spyder, > Chrome), SpecJVM, SpecWeb benchmarks on MARSS86? > > Thanks > Johnsy > > > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: < > https://www.cs.binghamton.edu/mailman/private/marss86-devel/attachments/20140207/9295cbf5/attachment-0001.html > > > > ------------------------------ > > Message: 3 > Date: Fri, 7 Feb 2014 14:56:05 -0500 > From: Brendan Fitzgerald <[email protected]> > To: "John, Johnsy" <[email protected]> > Cc: "[email protected]" > <[email protected]>, > "[email protected]" > <[email protected]> > Subject: Re: [marss86-devel] Can the MARSS86 be trace driven? > Message-ID: > < > cakv1wmbsqz0rfgts4smf82veyufijvsx0ypxcqv+9atux9s...@mail.gmail.com> > Content-Type: text/plain; charset="iso-8859-1" > > http://thread.gmane.org/gmane.comp.emulators.marss86/634/focus=636 > > > On Fri, Feb 7, 2014 at 2:28 PM, John, Johnsy <[email protected]> wrote: > > > Could someone on the development side, please comment on this? > > > > Thanks > > > > Johnsy > > > > > > > > *From:* John, Johnsy > > *Sent:* Monday, February 03, 2014 2:08 PM > > *To:* [email protected] > > *Cc:* John, Johnsy > > *Subject:* Can the MARSS86 be trace driven? > > > > > > > > Hi, > > > > I have a question regarding the modes (execution driven, trace driven) in > > which MARSS86 can be run. > > > > Can it be trace driven? > > > > If so, please provide the trace format. > > > > > > > > Thanks > > > > Johnsy > > > > > > > > _______________________________________________ > > http://www.marss86.org > > Marss86-Devel mailing list > > [email protected] > > https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel > > > > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: < > https://www.cs.binghamton.edu/mailman/private/marss86-devel/attachments/20140207/b61122f3/attachment-0001.html > > > > ------------------------------ > > Message: 4 > Date: Fri, 7 Feb 2014 15:57:52 -0500 > From: [email protected] > To: "John, Johnsy" <[email protected]> > Cc: "[email protected]" > <[email protected]>, > "[email protected]" > <[email protected]> > Subject: Re: [marss86-devel] Cloud, Hadoop, Facebook, DB, Web browser > (Sun Spyder, Chrome), SpecJVM, SpecWeb benchmarks on MARSS86? > Message-ID: > <[email protected]> > Content-Type: text/plain;charset=iso-8859-1 > > Anything Java-based will not work. If I understand correctly, SMC is not > handled correctly and it almost immediately breaks JIT/dynarec. So, that > knocks out Hadoop, SpecJVM, Android images, etc. > > I, personally, can't comment on much else as I typically only use SPEC, > PARSEC, etc.; we have a listing of which applications within these suites > work on the website, but not anything else. > > Tyler > > > Hi All, > > I have a question regarding the various new server benchmarks. > > Did anyone try Cloud, Hadoop, Facebook, DB, Web browser (Sun Spyder, > > Chrome), SpecJVM, SpecWeb benchmarks on MARSS86? > > > > Thanks > > Johnsy > > > > > > > > _______________________________________________ > > http://www.marss86.org > > Marss86-Devel mailing list > > [email protected] > > https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel > > > > > > > > ------------------------------ > > Message: 5 > Date: Sat, 8 Feb 2014 17:49:27 +0800 > From: beilei sun <[email protected]> > To: [email protected] > Subject: [marss86-devel] How are system calls realized in MARSSx86 ? > Message-ID: > <CAGvAz9iwW6UJcjnCXyDBrMj1dG=WDUuAWRHZ1utZYx7mi=_ > [email protected]> > Content-Type: text/plain; charset="iso-8859-1" > > Hi, > > *How are system calls realized in MARSSx86 ?* > > I didn't find the relevant realization in the ptlsim source code. > > So I tried the Qemu code. > According to the comments in the "sim_cpu_exec" function, "do_interrupt" > function simulates a real cpu exception. > "do_interrupt" function that is realized in /qemu/target-i386/op_helper.c. > Its comments said that this function is to begin execution of an > interruption. So I captured the parameter "intno" of the "do_interrupt" > function once it is called. > > But the value of this parameter was obvioulsy wrong compared with the > system call Linux actually called. When the system is busy, the "intno" > (interrupt NO.) may be 14, 7, 62, etc, even though the system is busy on > reading files. If the system is free, the value is 239. > > Any one have any ideas? > > Thanks very much. > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: < > https://www.cs.binghamton.edu/mailman/private/marss86-devel/attachments/20140208/60d50623/attachment-0001.html > > > > ------------------------------ > > Message: 6 > Date: Sat, 8 Feb 2014 18:01:37 +0800 > From: Deepak Kathayat <[email protected]> > To: beilei sun <[email protected]> > Cc: [email protected] > Subject: Re: [marss86-devel] How are system calls realized in MARSSx86 > ? > Message-ID: > < > cab8gmotsvjqsgcjbbsrhxcxia7fehguz2qv4myhmfakuuqz...@mail.gmail.com> > Content-Type: text/plain; charset="iso-8859-1" > > System calls are realized in QEMU. x86-64 uses syscall instruction to fire > up system calls. You are looking for helper_syscall in > target-i386/op_helper.c. > > Regards, > Deepak > > > On Sat, Feb 8, 2014 at 5:49 PM, beilei sun <[email protected]> wrote: > > > Hi, > > > > *How are system calls realized in MARSSx86 ?* > > > > I didn't find the relevant realization in the ptlsim source code. > > > > So I tried the Qemu code. > > According to the comments in the "sim_cpu_exec" function, "do_interrupt" > > function simulates a real cpu exception. > > "do_interrupt" function that is realized in > /qemu/target-i386/op_helper.c. > > Its comments said that this function is to begin execution of an > > interruption. So I captured the parameter "intno" of the "do_interrupt" > > function once it is called. > > > > But the value of this parameter was obvioulsy wrong compared with the > > system call Linux actually called. When the system is busy, the "intno" > > (interrupt NO.) may be 14, 7, 62, etc, even though the system is busy on > > reading files. If the system is free, the value is 239. > > > > Any one have any ideas? > > > > Thanks very much. > > > > _______________________________________________ > > http://www.marss86.org > > Marss86-Devel mailing list > > [email protected] > > https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel > > > > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: < > https://www.cs.binghamton.edu/mailman/private/marss86-devel/attachments/20140208/97efb3cf/attachment-0001.html > > > > ------------------------------ > > _______________________________________________ > http://www.marss86.org > Marss86-Devel mailing list > [email protected] > https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel > > > End of Marss86-Devel Digest, Vol 48, Issue 3 > ******************************************** >
_______________________________________________ http://www.marss86.org Marss86-Devel mailing list [email protected] https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel
