Re: [Haifux] Fwd: compiler/tools/architecture conference
On Mon, Oct 19, 2015 at 11:07:14PM +0300, Oron Peled wrote: > [Muly, sorry for hijacking your thread, but it's becoming > interesting...] No worries, I relinquish any and all claims I might have had for this thread. So, it occurs to me that Haifux hasn't met in ages. Would there by interest in a talk about a new profit-maximizing operating system I built called nom? Here's the abstract: In the near future, cloud providers will sell their users virtual machines with CPU, memory, network, and storage resources whose prices constantly change according to market-driven supply and demand conditions. Running traditional operating systems in these virtual machines is a poor fit: traditional operating systems are not aware of changing resource prices and their sole aim is to maximize performance with no consideration of costs. Consequently, they yield low profits. We present nom, a profit-maximizing operating system designed for cloud computing platforms with dynamic resource prices. Applications running on nom aim to maximize profits by optimizing for both performance and resource costs. The nom kernel provides them with direct access to the underlying hardware and full control over their private software stacks. Since nom applications know there is no single “best” software stack, they adapt their stacks’ behavior on the fly according to the current price of available resources and their private valuations of them. We show that in addition to achieving up to 3.9x better throughput and up to 9.1x better latency, nom applications yield up to 11.1x higher profits when compared with the same applications running on Linux and OSv. ObLinux: yes, there is a Linux angle. Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://haifux.org/mailman/listinfo/haifux
Re: [Haifux] Fwd: compiler/tools/architecture conference
Sorry, no idea -- what image are you asking about? ObLinux: who's up for organizing an installation party this semester? :-) Cheers, Muli On Sun, Oct 18, 2015 at 11:31 PM, Diego Iastrubni wrote: > [OT] ... well maybe... > > Muli, > > How did you set up your image in color...? I think this is the > corresponding > code, but I don't think I know how to configure this in KMail :) > > X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; > d=1e100.net; s=20130820; > h=x-gm-message-state:date:from:to:subject:message-id:mime-version > :content-type:content-disposition:user-agent; > bh=henfNDLcQC/XdtIXUVqJ2i2Sr9eyTsBeLL8ZAPOqUak=; > > b=dBLsnYmcf9XysMrIq0v6ivU7drKBEZ9av4tPi9EIBg7Iw4WTwto4x6xGnBIw1Ajql7 > > eSukst/HngHBe4If8Hn/6o8WZQjvagjOTePMCSdMiaguBexbgkhy4AUK4m6AFghICH2C > > guYmOoiuDRj1+CJy7SgNYBzDH5ZZP7gF5qzY0/thv6H7vhvjSc/SX36NwO+25Iv6SiE8 > > 8LpkbBtNpLOg9C00IPTLm/1wvssm/qaXCQ7zLJG8Mc1XIgu+B/ea7v8D1443YkW+z8Io > > Lz545RrV4WsF4E39t2bPpY0VaYhRh1cdrPONPRm9oiARcIsKCtlIYc9xmarEXM9tuhVZ > yx4g== > > > > On יום שלישי, 4 באוגוסט 2015 11:51:13 Muli Ben-Yehuda wrote: > > Might interest some of you. Apologies if you receive multiple copies. > > > > CALL FOR PAPERS > > > > Compiler, Architecture and Tools Conference (CATC) > > Intel Development Center, Haifa, Israel > > November 23rd, 2015 > > > > http://software.intel.com/compilerconf2015 > > > > Endorsed by the HiPEAC Network of Excellence > > > > > > IMPORTANT DATES > > > > * Paper Submission deadline: August 6th, 2015 > > > > * Author Notification: September 10th, 2015 > > > > * Conference date: November 23rd, 2015 > > > > > > > > OVERVIEW > > > > The interaction among advanced compilation techniques, modern > > processor and computing architectures, and associated tools continues > > to face new challenges and opportunities. Traditional demands to > > increase performance, reduce power consumption, and reduce time to > > market now apply to heterogeneous, virtualized and diverse > > user-experience environments. > > > > Extensive data and task parallelism are being exposed by new > > programming environments such as Halide, Swift, and OpenMP version > > 4.0, relying on innovative architectures, compilers, binary > > translation and runtime tools. > > > > This conference will focus on these exciting new directions and how > > they are influencing the architecture and compilation domain. The > > conference will be held in a format similar to that held last year > > leaving time for networking and presentations. > > > > TOPICS > > > > The main focus of this conference is the interaction of compiler > > technologies, processor and computing architectures and tools to > > address the latest programming environments and demands. The topics of > > interest for this conference include, but are not limited to: > > > > * Compilers, runtime and tools for modern server, client, mobile and > > embedded systems > > > > * Compiler/hardware support for hiding memory and I/O latencies > > > > * Compilation to hardware > > > > * Dynamic translation and optimization > > > > * Heterogeneous parallel architectures and computational models > > > > * GPU, accelerator and coprocessor architectures > > > > * Power/Performance/Monitoring tools for application behavior > > understanding > > > > * Parallel programming languages, algorithms and applications > > > > * Internet of Things > > > > INVITED SPEAKERS > > > > We are delighted to have Dr. Youfeng Wu, Principal Research Scientist > > at Intel and Prof. Mooly (Shmuel) Sagiv from Tel-Aviv University as > > our keynote speakers. > > > > VENUE > > > > The conference will take place at Intel Development Center > > (bldg. IDC9) in Matam Industrial Park, Haifa, Israel. Lunch and light > > refreshments will be served. Participation in the conference is free > > of charge, but registration is required. The official language of the > > conference is English. > > > > SUBMISSION > > > > Authors should submit an abstract of their contributions by email to > > Gadi Haber (gadi.ha...@intel.com). Submissions must include an > > abstract of 100-300 words and contact information. The authors are > > encouraged to provide related documents wi
[Haifux] Fwd: compiler/tools/architecture conference
Might interest some of you. Apologies if you receive multiple copies. CALL FOR PAPERS Compiler, Architecture and Tools Conference (CATC) Intel Development Center, Haifa, Israel November 23rd, 2015 http://software.intel.com/compilerconf2015 Endorsed by the HiPEAC Network of Excellence IMPORTANT DATES * Paper Submission deadline: August 6th, 2015 * Author Notification: September 10th, 2015 * Conference date: November 23rd, 2015 OVERVIEW The interaction among advanced compilation techniques, modern processor and computing architectures, and associated tools continues to face new challenges and opportunities. Traditional demands to increase performance, reduce power consumption, and reduce time to market now apply to heterogeneous, virtualized and diverse user-experience environments. Extensive data and task parallelism are being exposed by new programming environments such as Halide, Swift, and OpenMP version 4.0, relying on innovative architectures, compilers, binary translation and runtime tools. This conference will focus on these exciting new directions and how they are influencing the architecture and compilation domain. The conference will be held in a format similar to that held last year leaving time for networking and presentations. TOPICS The main focus of this conference is the interaction of compiler technologies, processor and computing architectures and tools to address the latest programming environments and demands. The topics of interest for this conference include, but are not limited to: * Compilers, runtime and tools for modern server, client, mobile and embedded systems * Compiler/hardware support for hiding memory and I/O latencies * Compilation to hardware * Dynamic translation and optimization * Heterogeneous parallel architectures and computational models * GPU, accelerator and coprocessor architectures * Power/Performance/Monitoring tools for application behavior understanding * Parallel programming languages, algorithms and applications * Internet of Things INVITED SPEAKERS We are delighted to have Dr. Youfeng Wu, Principal Research Scientist at Intel and Prof. Mooly (Shmuel) Sagiv from Tel-Aviv University as our keynote speakers. VENUE The conference will take place at Intel Development Center (bldg. IDC9) in Matam Industrial Park, Haifa, Israel. Lunch and light refreshments will be served. Participation in the conference is free of charge, but registration is required. The official language of the conference is English. SUBMISSION Authors should submit an abstract of their contributions by email to Gadi Haber (gadi.ha...@intel.com). Submissions must include an abstract of 100-300 words and contact information. The authors are encouraged to provide related documents with additional details on their research work, such as an extended abstract, paper, and/or links. Abstracts and eventually presentations will be provided online at the conference webpage. The conference does not include proceedings, however, with author approval, papers may appear online and extended versions of the best papers may be included in a special issue of IJPP. In addition, authors may submit papers recently presented in (or accepted to) top-tier conferences in the areas of interest. ORGANIZERS Ayal Zaks, Dorit Nuzman, Gadi Haber, Leeor Peled, Michal Nir; Intel Development Center, Haifa (first.last at intel.com) Erez Petrank; Computer Science Dept., Technion, Haifa (first at cs.technion.ac.il) Yosi Ben-Asher; Computer Science Dept., University of Haifa (first at cs.haifa.ac.il) PAST CONFERENCE http://software.intel.com/compilerconf2014 ___ Haifux mailing list Haifux@haifux.org http://haifux.org/mailman/listinfo/haifux
Re: [Haifux] Fwd: [YBA] Interested in giving HAIFUX lecture
On Mon, Jan 07, 2013 at 08:04:02AM +0200, Orna Agmon Ben-Yehuda wrote: > Hello Haifuxers, > > YBA is offering to speak in Haifux. Please voice your opinion regarding > your topics of interest. This one: > 1. Bonding and VLAN interfaces - I have a mobile lab consisting of a laptop > and a high-end HP 24 port switch that I use to demonstrate the > configurations. Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] Something fun and geeky to read
Nice idea! Two off the top of my head: The Story of Mel: http://catb.org/jargon/html/story-of-mel.html Some AI Koans: http://catb.org/jargon/html/koans.html Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] OT: Open CVS/SVN servers
On Tue, May 01, 2012 at 02:07:23PM +0300, Orr Dunkelman wrote: > I was wondering whether anyone is aware of open CVS/SVN servers that > allow users to open a small repository. > > The project I have in mind is actually a LaTeX project, and not an > open source one, so not much space is needed. > > Support of 4 users is needed, not much storage (LaTeX files, after > all), and of course, availability (but the project would end in a > year or two, so I am not looking for something that would stay > forever). Apologies in advance for answering a question you didn't ask, but if you can replace "CVS/SVN" with "git or mercurial", then I can strongly recommend bitbucket. They provide free unlimited private repositories for educational users. Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] lecture suggestion: "disk" storage media - state of the art in the enterprise world
On Wed, Apr 04, 2012 at 10:22:13AM +0300, guy keren wrote: > 7. will cover hard disks, SSDs (Solid-State Disks) and the way they > are/will be used in enterprise computing environments. > > if there's interest - it'll take me a few weeks to prepare. I'll try very hard to come to this one. Cheers Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
[Haifux] Fwd: [ceclub] Today: Prof. Dror Feitelson @ EE
This talk may interest many of you. *Today* at 11:30. Sorry for the short notice. Cheers, Muli - Forwarded message from Ittay Eyal - Date: Tue, 15 Nov 2011 23:30:38 +0200 From: Ittay Eyal To: clubne...@listserv.technion.ac.il Subject: [ceclub] Today: Prof. Dror Feitelson @ EE Reply-To: Ittay Eyal You are invited to the next ceClub talk. For details of coming talks, ceClub's online calendar etc., please visit the ceClub site <http://ceclub.technion.ac.il/>. Speaker: Prof. Dror Feitelson Title:Observations on Linux Development Affil.: Hebrew University, CS Time: *Today*, 16/11/11, *11:30* * * * * Location: *Meyer 861* Abstract: Below Linux is used extensively in systems research as a platform for the implementation of new ideas, exploiting its open-source nature. But Linux is also interesting as an object of study about software engineering. In particular, Linux defies common management theories, as it lacks any coherent plan or management structure, but still grows at an ever-increasing rate, while also gaining market share. We will review some previous studies of Linux development and add new observations regarding the lifecycle model and code complexity. The presentation includes results of my students Ayelet Israeli, Ahmad Jbara, and Adam Matan. *bio:* Dror Feitelson has been on the faculty of the School of Computer Science and Engineering at the Hebrew University of Jerusalem since 1995. His major contributions have been in parallel job scheduling (where he co-founded the JSSPP series of workshops) and workload modeling (where he maintains the parallel workloads archive). Recently he has also started to work on software engineering, and in particular the development of open-source systems like Linux. o-o-o-o-o-o-o-o-o-optout-optout-optout-o-o-o-o-o-o-o-o-o To unsubscribe, click the following link: http://listserv.technion.ac.il/cgi-bin/wa?TICKET=NzM0NDg2IG11bGlAQ1MuVEVDSE5JT04uQUMuSUwgQ0xVQk5FVC1MIDhP/bsHa+vZ&c=SIGNOFF - End forwarded message ----- -- Muli Ben-Yehuda | http://www.mulix.org ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] Running 32 bit applications (Firefox?) on 64 bit machines
On Sat, Aug 20, 2011 at 07:32:29PM +0300, Eli Billauer wrote: > It's not like I expect Firefox to address 1 GB of RAM. There's no 1GB of RAM limitation with 32 bit. Perhaps you meant 4GB of RAM? > So it really makes me wonder: Why are the preinstalled binaries on a > 64 bit machine, well, 64 bit executables? I run a 64 bit machine > because I want the *overall* RAM to exceed 4 GB, but except for > virtual machines, I don't expect any application to have problems > with the 32 bit limitation. I speculate that the best reason, from the distribution's point of view, is that it is simpler to only maintain and support one environment (64 bit) rather than maintain and support co-existing 32 bit and 64 bit environments. Distributions will grudgingly carry both 32 bit and 64 bit versions of an application where there 64 bit version is deficient in some sense (e.g., firefox with flash) but would naturally like to minimize these hassles. Cheers, Muli -- Muli Ben-Yehuda | http://www.mulix.org ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] Implementing read() like UNIX guys like it
On Fri, Apr 22, 2011 at 01:31:02PM +0300, Eli Billauer wrote: > The overall package I'm working on is kind-of general purpose, and > it's the package's user who chooses how the hardware input behaves. > I can't know anything about the data flow behavior, and still, I > want the reads from the relevant file descriptor to behave as one > would expect. For example, a user may choose to read from the file > descriptor with scanf or fgets. The user would expect these to > return whenever sufficient data has been fed into the hardware side. > On the other hand, if the data comes slowly into hardware, read()'s > will return with one byte at a time, which is maybe not desirable > either. There are a couple of ways you could look at it. First, ask yourself is the package user more likely to treat this as a file or a socket? And if a file, is it a structured file or a certain size, or is it a file whose contents are not known in advance? Then make your read behave the same way a read on such a file or socket would behave. I don't like this option, personally, because it makes assumptions about the user. The second way is to just support both modes of operation. Check if you were opened with O_NONBLOCK. If yes, don't ever block -- you are allowed to return -EAGAIN if you would otherwise block. If you were not opened with O_NONBLOCK, then blocking is allowed, and you can wait until you have a full quantum of data to return at once. Just out of curiosity -- what does the hardware do? Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] Lecture about PCI?
On Tue, Feb 15, 2011 at 11:33:14PM +0200, Eli Billauer wrote: > This is a pretty heavy lecture to prepare, but if I'll see you guys > here getting crazy about the idea, I'll consider it. :) Count me in. I think that to do these subjects justice you'll need more than 2 hours... Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] lecture (series) proposal: the story of alice and bob - the I/O requests
On Fri, Jan 28, 2011 at 02:59:27AM +0200, c...@actcom.co.il wrote: > > any takers? anyone already has all the needed material (lack of such > material was the initial trigger for this proposal)? Sounds great, I'll definitely come listen to this one. Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] new Technion course: Operating Systems Engineering
On Mon, Jan 24, 2011 at 12:09:19PM +0200, Shachar Shemesh wrote: > Don't get me wrong. Had I still been a student, I'd take that class > in a blink of an eye. I'm just worried that students will take it > without realizing just how much work it is. We intend to clarify up front that this is going to require serious work, and anyone who isn't up to it, is welcome to drop the class. > >>Is the platform Intel, or something saner (say, ARM, PPC, MC680x0 > >>or, come to think of it, just about any other CPU)? > >The platform is good ol' x86 because (1) everyone has one and (2) > >whatever you learn in this course will be immediately applicable in > >the real world. > It also has a horrific assembly, as well as a "too many layers" MMU > which revives that segmented addresses everyone have been working so > hard to forget ever existed. Undoubtedly x86 won't win any points in an ISA beauty contest, but it does what it does well, segmentation can be useful even today (see for example the original Xen protection model on 32 bit or vx32), and its hierarchical page table model is the one the students are familiar with and is heavily explored in the research literature. > ARM, on the other hand, is (1) reasonable and straight forward > assembly, and even reasonable MMU (2) has readily available > emulators for all platforms and (3) is also fairly immediately > applicable in the real world, arguably more so than X86. It also lags quite a bit behind x86 in various interesting ways, if you come from the server space: virtualization support for example, or efficient page table design. Not to mention that ARM apparently is following in the x86 foot steps -- didn't they just introduce the horrors of PAE? > Let's put it this way. The chances a CS graduate will be asked to > write an ARM based BSP are much higher than the chances she'll be > asked to write an X86 based one. Maybe. But the chances he'll end doing kernel hacking on x86 are higher, especially if he lands eventually in our research group. Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] new Technion course: Operating Systems Engineering
Shachar Shemesh wrote: > On the one hand, this is an extremely interesting course. On the > other hand, it sounds like a HUGE amount of work for just 4 points. You know what they say, no pain, no gain. This will be the sort of course that puts a beard on your face. > Is the platform Intel, or something saner (say, ARM, PPC, MC680x0 > or, come to think of it, just about any other CPU)? The platform is good ol' x86 because (1) everyone has one and (2) whatever you learn in this course will be immediately applicable in the real world. Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] Lecture about valgrind
On Wed, Jun 16, 2010 at 10:51:15AM +0300, Shachar Raindel wrote: > Hi all, > > I have been fiddling with the insides of valgrind lately, and > thought that other club members might also be interested to hear > about how it works under the hood. > > For these of you who haven't heard of valgrind yet, it is a dynamic > binary translator, which pin-points common memory-related problems > (most of the buffer overflows, using freed memory block, double > free, using uninitialized variables, memory leaks, etc.). > > If there is enough interest, I will probably be able to get a few > slides together and talk about it for a bit (read - 2 hours). +1 from me. Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
[Haifux] Final Call for Participation: SYSTOR 2010---The 3rd Annual Haifa Experimental Systems Conference
SYSTOR 2010 The 3rd Annual Haifa Experimental Systems Conference https://www.research.ibm.com/haifa/conferences/systor2010/index.shtml 24--26 May 2010 Haifa, Israel Registration deadline: May 22 The IBM Haifa Research Lab and the IBM Systems and Technology Group Lab, in collaboration with Caesarea Rothschild Institute (CRI) at the University of Haifa are organizing SYSTOR 2010, a successor to the highly successful SYSTOR conference and workshops on systems and storage held at the IBM Haifa Research Lab. The goal of the conference is to promote systems research and foster stronger ties between the Israeli and worldwide systems research communities and industry. Therefore, international submissions are specially encouraged. SYSTOR 2010 will be a three-day conference. The first day will begin with a keynote talk by Prof. David Kaeli, IEEE fellow from Northeastern University on "Many-core Computing: A Disruptive Technology Enabling Low-cost, Low-power Desktop Supercomputing". The talk will be followed by sessions on storage and virtualization and workload-driven optimizations. In the afternoon, a student poster session with sweet refreshments will be held. The second day will begin with a keynote talk by Prof. Idit Keidar from the technion on "Reliable Distributed Storage" followed by sessions on parallel programs and virtualization. The morning talk and sessions will be held at the Hech Museum Auditorium in Haifa University. The day will end with an optional social event in a tour to Nazareth and a dinner near the Sea of Galilee. The third day will include sessions on storage and communication and low-level optimizations and a panel about cloud computing: "Cloud Computing: A New Direction or a Passing Trend?" The full program for all three days is available on the conference website: https://www.research.ibm.com/haifa/conferences/systor2010/program.shtml We look forward to seeing you at SYSTOR 2010! -- Muli Ben-Yehuda | m...@il.ibm.com | +972-4-8281080 Manager, Virtualization and Systems Architecture Master Inventor, IBM Research -- Haifa ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
[Haifux] Call for Participation: SYSTOR 2010---The 3rd Annual Haifa Experimental Systems Conference
Call for Participation: SYSTOR 2010 — The Third Annual Haifa Experimental Systems Conference May 24–26, 2010 https://www.research.ibm.com/haifa/conferences/systor2010/ IBM R&D Labs in Israel IBM Research - Haifa and the IBM Systems and Technology Group Lab in Israel, in collaboration with the Caesarea Rothschild Institute (CRI) at the University of Haifa are organizing SYSTOR 2010, a successor to the highly successful SYSTOR conferences on systems and storage held in the past at IBM Research - Haifa. The goal of the conference is to promote systems research and foster stronger ties between the Israeli and worldwide systems research communities and industry. We have a strong program of 18 refereed papers, two keynote talks, and a panel on the future of cloud computing. Full program details are available at https://www.research.ibm.com/haifa/conferences/systor2010/program.shtml The conference will also have a poster session highlighting works in progress. International participation is encouraged. Student Activities As part of the conference, students will have the opportunity to interact with senior researchers and obtain feedback on their research. To promote such interactions, we will have a poster session where students present their work. We will also have a lunch where members of the program committee and IBM researchers will be matched with students for more intensive interaction. Venue and Social Event The conference will be held in the lovely city of Haifa, which overlooks the Mediterranean Sea, and is home to Jews, Muslims, Christians, and Baha’is. SYSTOR 2010 excursions will include trips and museum tours focusing on the historical periods of the Holy Land via special visits to the University of Haifa's Hecht Museum, the old city of Nazareth, and the Sea of Galilee. The conference will take place at IBM Research - Haifa and the University of Haifa. -- Muli Ben-Yehuda | m...@il.ibm.com | +972-4-8281080 Manager, Virtualization and Systems Architecture Master Inventor, IBM Research -- Haifa ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] Kernel oops, so what?
On Fri, Jan 15, 2010 at 04:27:34PM +0200, Eli Billauer wrote: > So, should I just take it cool and wait for a new kernel with this > fixed, ignoring these messages? > > Just for the heck of it, the relevant part from /var/log/messages > follows. I am running on a new Gigabyte motherboard. I don't expect > anyone to dissect this. I mean, a report has been submitted. > > >Eli > > > - log follows, for the fun - > > > Jan 15 14:45:17 test kernel: WARNING: at arch/x86/kernel/hpet.c:390 > hpet_next_event+0x5c/0x81() (Not tainted) Given that this is just a warning and is not fatal, I would either debug it or ignore it. Generally the kernel will warn when something unexpected happens, but will then continue limping merrily along. Cheers, Muli -- Muli Ben-Yehuda | m...@il.ibm.com | +972-4-8281080 Manager, Virtualization and Systems Architecture Master Inventor, IBM Research -- Haifa Second Workshop on I/O Virtualization (WIOV '10): http://sysrun.haifa.il.ibm.com/hrl/wiov2010/ ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
[Haifux] [JOB OFFER] systems researchers at IBM Research - Haifa
[apologies if you've already received this on linux-il] Hello everyone, We have a limited number of open systems research positions at the IBM Research - Haifa lab[1-3]. We are looking for PhD-level[4] researchers and engineers, with proven hands-on experience and expertise in one or more of the following areas: networking, operating systems, virtualization, systems management, grid/cloud computing, and storage. Successful candidates should demonstrate strong background in computer science or computer engineering, ability to perform cutting edge R&D, with a focus on ability to take ideas from idea to materialization. At the lab we combine cutting edge research with advanced development, with a heavy dose of kernel hacking thrown in. We work on and contribute to many open source projects, including the Linux kernel and the KVM hypervisor. If you are interested, or know someone who might be interested, please don't hesitate to send me an email with a CV. [1] http://www.haifa.il.ibm.com/research.html [2] The systems department: http://www.haifa.il.ibm.com/dept/stt/index.html [3] My group: http://www.haifa.il.ibm.com/dept/stt/sas.html [4] If you have what it takes, lack of a PhD is not an issue. Cheers, Muli -- Muli Ben-Yehuda | m...@il.ibm.com | +972-4-8281080 Manager, Virtualization and Systems Architecture Master Inventor, IBM Haifa Research Laboratory ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
[Haifux] [OFFTOPIC?] SYSTOR 2009 Call for Participation
CALL FOR PARTICIPATION SYSTOR 2009---The Israeli Experimental Systems Conference http://www.haifa.il.ibm.com/conferences/systor2009/ 4-6 May 2009 Haifa, Israel Registration deadline: May 2nd SYSTOR 2009, the Israeli Experimental Systems Conference, will be held at IBM Haifa Labs, in Haifa, Israel. The conference program will run over three days, combining the forefront of academic systems research with real-world systems developed in industry. The goal of the conference is to promote systems research and to foster stronger ties between the Israeli and worldwide systems research communities and industry. Conference proceedings will be published by ACM in the ACM Digital Library. There is a limited number of seats available on a first-come-first-served basis upon registration at http://www.haifa.ibm.com/conferences/systor2009/registration.shtml (registration is free of charge). Lunch and refreshments will be served on all three days courtesy of IBM Haifa Labs. The first day of the conference will feature sessions on distributed systems, concurrency, and power management. Marc Snir, University of Illinois at Urbana Champaign, will give a keynote talk, and in the afternoon a student poster session with sweet refreshments will be held. The second day will begin with the keynote "Towards Invisible Storage" by Alain Azagury, Director, XIV Business Executive, IBM, and an invited talk on "The Next Generation Data Center" by Michael Kagan, Mellanox CTO. After the morning talks, there will be paper sessions focusing on data de-duplication and storage issues. The day will end with an optional social event in Caesarea. The third day will conclude the conference with paper sessions on virtualization and system optimizations, and a panel of well-known systems researchers who will debate "What is Systems Research about and is it Relevant?" The full program for all three days is available on the conference website. We look forward to seeing you at SYSTOR 2009! SYSTOR Advisory Committee * Marc Auslander, IBM * Ken Birman, Cornell * Danny Dolev, HUJI * Julian Satran, IBM * Marc Snir, UIUC * Willy Zwaenepoel, EPFL Program Chairs * Michael Factor, IBM * Dror Feitelson, HUJI General Chair * Miriam Allalouf, IBM Publicity Chair * Muli Ben Yehuda, IBM Publication Chair * Gregory Chockler, IBM ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
[Haifux] Fwd: EuroSys Student Travel Grant to attend OSDI 2008
Perhaps one of the PhD students on the list will be interested: EuroSys Student Travel Grants (OSDI 08): Call for Applications == A primary goal of the EuroSys society is to promote excellence in Computer Systems education, research and industry in Europe. As such, we are offering the opportunity for talented European PhD students to attend a top-class systems conference where they can be exposed to high-standard research in the company of peers from around the world. EuroSys therefore offers TWO TRAVEL GRANTS for PhD students to attend the OSDI 2008 conference. OSDI is one of the leading systems conferences and has published many influential results in previous years. The travel grants will fully fund air travel to the conference (up to a preset limit sufficient from most European locations), registration fees and budget accommodation for up to 4 nights, which may involve sharing a room. Other costs, such as meals not already included in the registration fees, will not be reimbursed by EuroSys. Recipients of a Usenix student travel grant are also eligible to apply, but will only receive top-up funding for expenses not already covered. In return, we will ask travel grant recipients to carry out two tasks: - Distribute EuroSys fliers or posters at the conference. - Write a summary of the conference for the EuroSys webpage, or even better, maintain a real-time blog of the sessions such as the one by Derek Murray at SOSP 2007: http://www.mrry.co.uk/blog/2007/10/. The EuroSys travel grant is a competitive award with recipients selected by merit rather than need. Note that it is not a strict requirement to be a member of EuroSys to apply, and free membership for the following year will be included as part of the grant. The submission deadline is October 27th 2008 and no extension will be given. If you wish to apply please address your application to [EMAIL PROTECTED] and include the following: - Your CV. - At most 1 paragraph summarising your current research. - At most 1 page on which paper you think is the most interesting in past proceedings of the OSDI conference and why. - If you are neither a PhD student at a European institution nor a member of EuroSys: explain why you think EuroSys should fund your trip. We also require a recommendation letter from your PhD advisor sent directly to [EMAIL PROTECTED] by the October 27th deadline. Selection committee: - Rebecca Isaacs, Microsoft Research - Guillaume Pierre, VU university Amsterdam http://www.eurosys.org/travel-grant/ -- The First Workshop on I/O Virtualization (WIOV '08) Dec 2008, San Diego, CA, http://www.usenix.org/wiov08/ <-> SYSTOR 2009---The Israeli Experimental Systems Conference http://www.haifa.il.ibm.com/conferences/systor2009/ ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] SSD and linux
On Thu, Sep 18, 2008 at 06:49:58PM +0300, Doron Zuckerman wrote: > Do you have any idea where we can find this? I haven't looked at those bits recently, but it sounds like Gabi is pointing you to the right path. In any case, to be honest, I think what you propose doesn't make sense, even as "research". Look at it this way. When does busy waiting makes sense? When the overhead of sleeping is offset by the useful work that gets done while you sleep (or when you can't sleep). So let's say that the overhead of a context switch is T_c. Switching to some other task and back will cost 2*T_c. Assuming that any work that the task you switch to does is useful, busy waiting makes sense only if you can resume executing the faulting task within 2*T_c time. So, unless you can read the frame from the SSD within 2*T_c time (which I highly doubt...) busy waiting does not make sense. Another point to consider is that if you are running on a UP machine and your kernel isn't preemptible, and the work to submit the I/O to disk happens in some other context than the one you run in, if you busy wait the I/O may never get submitted, and you'll busy wait forever! Cheers, Muli -- Workshop on I/O Virtualization (WIOV '08) Co-located with OSDI '08, Dec 2008, San Diego, CA http://www.usenix.org/wiov08 ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] SSD and linux
On Thu, Sep 18, 2008 at 06:56:09PM +0300, gabik wrote: > Doron > > You can work on 2.6.24 if you prefer. I just picked some version and checked > on it. (for some reason there is no arch/i386 in 2.6.24. Maybe they have > renamed it into x86?) Yes. arch/x86 is now for both 32 and 64 bits. Cheers, Muli -- Workshop on I/O Virtualization (WIOV '08) Co-located with OSDI '08, Dec 2008, San Diego, CA http://www.usenix.org/wiov08 ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] SSD and linux
On Thu, Sep 18, 2008 at 12:27:36PM +0300, Doron Zuckerman wrote: > I'm not sure it will speed up the OS, however I'm doing an academic > research on the matter as part of a project I'm taking, and I plan > to check this point. I'm pretty sure it won't. > The leading thought was that since the SSD is not a mechanical > drive, pages can be brought faster in this way, and there is no need > to context switch, thus, avoiding the overhead included. I suggest a much simpler exercise: (a) time how long it takes to read a block of data from the SSD (b) time how long a context switch takes See that (b) is orders of magnitude faster than (a). > So far I found the function "__generic_make_request" in file > "ll_blk". This function calls a sub function named "might_sleep". > I have deleted the call to this function whenever I'm in a > pagefault, however I'm not sure if this function casuses the sleep, > or is just used for debugging in order to check if we entered a > suspend state. might_sleep() is a debugging aid, which is used by code that might sleep in order to check that it hasn't been called in a context where you can't sleep (non-process context such as an interrupt handler). > My question is if this is the function I should change in order to > accept the change I'm willing to get, or if the change should be > made in > q->make_request_fn which, according to my understanding, > belongs to the specific driver I'm using. Neither. Take a look at the page fault path for a major fault. What it does (from 10,000 feet) is initiate reading the page from disk, and then going t sleep until the page is ready. Going to sleep in the page fault path is what causes the context switch you want to avoid. What you want to do instead of going to sleep is busy-wait for the data. Try to find the place where the faulting process is put to sleep and convert that code to busy wait instead, terminating the busy-wait when the page has been brought in. Cheers, Muli -- Workshop on I/O Virtualization (WIOV '08) Co-located with OSDI '08, Dec 2008, San Diego, CA http://www.usenix.org/wiov08 ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] SSD and linux
On Tue, Sep 16, 2008 at 12:31:09PM +0300, Doron Zuckerman wrote: > Hi all, > > I have a question regarding the linux kernel (for those of you who are > familiar with it). > > I'm looking for a way to add a change to the linux kernel in order to check > if I can make it more compatible with my Asus EEE-PC. > I would like to change the kernel in such way that it will not do a context > switch every time there is a page fault > and will wait for the required page to be brought from the SSD (Solid State > Drive), then continue as usual. We context switch because the task (thread) cannot continue working until the page is paged in from the disk. If we don't context switch, and the thread cannot continue running until the page fault is resolved, what will the OS do in the meantime? Note that even though the EEE has an SSD drive, it's still several orders of magnitude slower than the time the context switch takes. Cheers, Muli -- Workshop on I/O Virtualization (WIOV '08) Co-located with OSDI '08, Dec 2008, San Diego, CA http://www.usenix.org/wiov08 ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] Lightning talk! LEHITPAKED
On Wed, Jul 09, 2008 at 12:16:50PM +0300, Ohad Lutzky wrote: > (I couldn't find any appropriate translation. Anyone know one?) fall in. Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
[Haifux] Technion job fair
Is anyone planning to go to the Technion job fair tomorrow? I'll be slaving away at the IBM booth, and if you'll be around, you should come and say hi. Especially if you'd like to do research into operating systems and virtual machines at IBM's Haifa Research Lab. In fact, if that sounds like something you'd like to do, send me your resume ASAP regardless of whether you'll be at the fair... Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
[Haifux] JOB: senior Linux kernel engineer
IBM is looking for a senior Linux kernel engineer to join the recently acquired FilesX team. To apply, please send your CV to Keren Benes-Yosef . Feel free to shoot me an email if you have any questions. Quoth: Location: Haifa Description: The employee will function as part of a team that develops an advanced Data Protection & Recovery Management solutions. Requirements: Candidate must have five years of experience with programming in C/C++ and at least 3 years experience with Unix/Linux kernel development. Candidate must have experience with developing Unix/Linux kernel modules and storage drivers. Excellent interpersonal skills, ability to work in multi-task environment; good problem solving skills and excellent English are also a must. Education: B.A./B.Sc. in computer science or an IDF computers division graduate. P.S. Keren is also looking for senior Windows kernel engineers. Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
[Haifux] my talk on the 25th
Unfortunately I will not be able to give the talk scheduled on the 25th (have to fly to the US earlier than expected). Is anyone else interested in giving a talk then? Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] [Haifux Lecture] User space syscall tracing andmanipulation - fakeroot-ng by Shachar Shemesh
On Thu, Jan 17, 2008 at 02:12:31PM +0200, Shachar Shemesh wrote: > 1. Automatic manipulation. Unlike strace, fakeroot-ng actually >changes the program while running. Unlike gdb, it does so >automatically. When I did this in the past, it was always intimately tied to what the victim was doing. I'll be very interested in hearing how you got around it. > 2. Syscall generation - program calls one syscall, you make it call >three. Interesting... I assume this is without kernel support (e.g., UML's SKAs patches). > 3. Recursive debuggers support - run strace (or fakeroot-ng, but at > least at the moment not gdb) from within the fakeroot environment. Fun. Looking forward to the talk. Cheres, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
Re: [Haifux] [Haifux Lecture] User space syscall tracing andmanipulation - fakeroot-ng by Shachar Shemesh
On Thu, Jan 17, 2008 at 12:45:10PM +0200, Shachar Shemesh wrote: > Fakeroot-ng is a (as far as I know) first attempt to do the things > usually done with LD_PRELOAD using the ptrace mechanism. It was both > the trigger and the root cause of this lecture. Not sure what you mean by "things usually done with LD_PRELOAD?" Certainly ptrace has been used to both trace and modify running binaries, by gdb, strace, dumpmem[1], memfetch[2] and others. I think I even gave a haifux talk on run-time modification of programs using ptrace for fun an profit a few years ago. > The lecture will look at fakeroot, fakechroot, fakeroot-ng and > strace, at varying degrees of depths, mostly because all four chose > slightly different approaches for solving, fundamentally, the same > problem. They did? Sounds like an interesting talk, will try to attend. [1] http://www.mulix.org/dumpmem.html [2] http://lcamtuf.coredump.cx/soft/memfetch.tgz Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
[Haifux] talk offer: tapping into the fountaion of CPUs
Would there be interest from the esteemed list members in hearing the following paper, to be presented at ASPLOS '08? Tapping into the Fountain of CPUs---On Operating System Support for Programmable Devices, by Yaron Weinsberg, Danny Dolev, Tal Anker, Muli Ben-Yehuda, Pete Wyckoff. Abstract: The constant race for faster and more powerful CPUs is drawing to a close. No longer is it feasible to significantly increase the speed of the CPU without paying a crushing penalty in power consumption and production costs. Instead of increasing single thread performance, the industry is turning to multiple CPU threads or cores (such as SMT and CMP) and heterogeneous CPU architectures (such as the Cell Broadband Engine). While this is a step in the right direction, in every modern PC there is a wealth of untapped compute resources. The NIC has a CPU; the disk controller is programmable; some high-end graphics adapters are already more powerful than host CPUs. Some of these CPUs can perform some functions more efficiently than the host CPUs. Our operating systems and programming abstractions should be expanded to let applications tap into these computational resources and make the best use of them. Therefore, we propose the Hydra framework, which lets application developers use the combined power of every compute resource in a coherent way. Hydra is a programming model and a runtime support layer which enables utilization of host processors as well as various programmable peripheral devices' processors. We present the framework and its application for a demonstrative use-case, as well as provide a thorough evaluation of its capabilities. Using Hydra we were able to cut down the development cost of a system that uses multiple heterogenous compute resources significantly. Cheers, Muli ___ Haifux mailing list Haifux@haifux.org http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
[Haifux] SYSTOR 2007 Call For Participation
[we apologize if you receive multiple copies of this announcement]
CALL FOR PARTICIPATION
***
Haifa Systems and Storage Conference 2007 (SYSTOR '07)
October 29 - 30, 2007, Haifa, Israel
http://www.haifa.il.ibm.com/Workshops/systor2007
***
Registration deadline: October 22.
The first annual Haifa Systems and Storage Conference will be held at
the IBM Haifa Labs, in Haifa, Israel. The conference consists of two
one-day workshop, a virtualization workshop on October 29 and a
storage workshop on October 30.
SYSTOR 2007 is the first high-quality refereed systems and storage
conference organized by IBM Haifa Labs and the Technion, drawing upon
the successful foundation of previous systems and storage
seminars. The purpose of this conference is to forge and nourish
research and working relations within the academic and industrial
community targeting researchers and practitioners alike.
There is a limited number of seats available on a
first-come-first-served basis upon registration at
http://www.haifa.il.ibm.com/Workshops/systor2007/registration.html
(registration is free of charge). Lunch and refreshments will be
served on both days courtesy of IBM Haifa Labs.
---
Virtualization Workshop: the first day of the conference will be a
full-day systems research workshop on the subject of
virtualization. The workshop will present contemporary advancements in
the field of systems virtualization and will feature two renowned
keynote speakers:
Danny Dolev
(Hebrew University of Jerusalem, Israel)
and
Willy Zwaenepoel
(Lausanne Federal Institute of Technology (EPFL), Switzerland)
The workshop will also include a number of talks, a lightning-talk
session and a poster session. Audience participation will be expected
and encouraged. Selected best papers from the workshop will be
published in a special issue of ACM SIGOPS Operating Systems Review.
Storage Workshop: the second day of the conference will include
presentations by IBM contributors and guest speakers on storage market
trends and innovations. A focus will be put on leading R&D trends that
will likely underlie the future of storage technologies. Keynote
speaker:
Pini Cohen
(Senior analyst, STKI)
The full agenda for both days will be available on the conference
website.
---
SYSTOR '07 Organizing Committee:
* Miriam Allalouf, IBM Haifa Research Lab
* Gitit Bar-El, IBM Haifa Development Lab
* Muli Ben-Yehuda, IBM Haifa Research Lab
* Yitzhak (Tsahi) Birk, Technion
* Eliezer Dekel, IBM Haifa Research Lab
* Erez Hadad, IBM Haifa Research Lab
* Arik Kol, IBM Israel, GTU
* Dalit Naor, IBM Haifa Research Lab
* Amir Sasson, IBM Haifa Development Lab
* Assaf Schuster, Technion
* Shmuel Vashdi, IBM Haifa Development Lab
* Yaron Wolfsthal, IBM Haifa Research Lab
* Ben-Ami Yassour, IBM Haifa Research Lab
Virtualization Workshop Program Committee:
* Assaf Schuster, Technion
* Yitzhak (Tsahi) Birk, Technion
* Roy Friedman, Technion
* Sivan Toledo, Tel Aviv University
* Shlomi Dolev, Ben Gurion University
* Leendert Van Doorn, AMD
* Marc E. Fiuczynski, Princeton
* Dilma Da Silva, IBM T. J. Watson Research Center
* Bob Wisniewski, IBM T. J. Watson Research Center
* Michael Factor, IBM Haifa Research Lab
* Julian Satran, IBM Haifa Research Lab
* Muli Ben-Yehuda, IBM Haifa Research Lab
* Ilan Shimony, IBM Haifa Research Lab
* Leah Shalev, IBM Haifa Research Lab
* Avi Mendelson, Intel
* Shimon Shoken, IDC
Storage Workshop Program Committee:
* Michael Rodeh, IBM Corporate
* Aviad Zlotnick, IBM Haifa Development Lab
* Arik Kol, IBM Israel, GTU
* Gitit Bar-El, , IBM Haifa Development Lab
* Amir Sasson, IBM Haifa Development Lab
* Shmuel Vashdi, IBM Haifa Development Lab
SYSTOR '07 local arrangements:
* Yael Hay Karsenty, IBM Haifa Research Lab
* Sigal Feler, IBM Haifa Development Lab
--
SYSTOR 2007 --- 1st Annual Haifa Systems and Storage Conference 2007
http://www.haifa.il.ibm.com/Workshops/systor2007/
Virtualization workshop: Oct 29th, 2007 | Storage workshop: Oct 30th, 2007
___
Haifux mailing list
Haifux@haifux.org
http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux
[Haifux] looking for electrician/comm guy to wire up a new house
Forwarding on behalf of Dan: I'm looking for an electrician/comm guy to wire my house (standard copper ethernet). All recommendations are welcome. I'll report back to the list with feedback if there's interest. Please reply to the list or to [EMAIL PROTECTED] Dan Pelleg - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
[Haifux] recommendations for computer store
I'm in the market for a new computer or two. Its been a while since my last computer purchase - can anyone recommend a good store in Haifa? price is less important than prompt service. Thanks, Muli - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] A lecture suggestion : Linux Kernel Networking Overview
On Sun, Jul 01, 2007 at 05:11:05PM +0300, Rami Rosen wrote: > Hello all, > > I suggest to give a lecture about Linux Kernel Networking; Two thumbs up from me. I'll even attend :-) One comment: you have way too much material there to cover it in sufficient depth in one talk. I suggest picking one (or more) subjects and focusing on those. Cheers, Muli - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] [HAIFUX Lecture] The Price of Safety: Evaluating IOMMU Performance by Muli Ben-Yehuda
On Fri, Jun 08, 2007 at 09:48:23AM +0300, Orr Dunkelman wrote: > Next Monday, 11th of June, at 18:30 the Haifa Linux Club, will > gather to hear Muli Ben-Yehuda speak about > > >The Price of Safety: Evaluating IOMMU Performance Slides are now available at http://www.mulix.org/lectures/price-of-safety/price-of-safety-haifux-jun-2007.pdf Cheers, Muli - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
[Haifux] Re: [HAIFUX Lecture] The Price of Safety: Evaluating IOMMU Performance by Muli Ben-Yehuda
On Sat, Jun 09, 2007 at 02:57:29PM +0300, Chava Leviatan wrote: > Any chance that this interesting lecture will get to Tel-Aviv as > well ? I don't have any plans at the moment to give it in Tel-Aviv. Cheers, Muli - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
[Haifux] [OFFTOPIC?] CFP Virtualization Workshop at SYSTOR 2007
[we apologize if you receive multiple copies of this announcement] The Storage and Systems Department at IBM's Haifa Research Lab (HRL), in collaboration with the Technion - Israel Institute of Technology, cordially invite you to a full-day workshop on the subject of virtualization within the SYSTOR 2007 conference. SYSTOR 2007 is the first high-quality refereed systems and storage conference organized by HRL, drawing upon the successful foundation of previous systems and storage seminars. The purpose of this conference is to forge and nourish research and working relations within the academic and industrial community in Israel and in the world, targeting researchers and practitioners alike. The virtualization workshop in SYSTOR 2007 will present contemporary advancements in the field of systems virtualization and will feature two renowned keynote speakers: Danny Dolev from the Hebrew University of Jerusalem, Israel, and Willy Zwaenepoel from Lausanne Federal Institute of Technology (EPFL), Switzerland. SYSTOR 2007 will be held on Monday and Tuesday, October 29-30, 2007, at the HRL site at the Haifa University campus from 09:15 to 17:15. The virtualization workshop will take place on the first day followed by the Storage: Market Trends and Future R&D seminar for practitioners on the second day. Lunch and refreshments will be served on both days. Participation is free. Topics Relevant topics include, but are not limited to, the following: * Virtualization techniques: native, emulation, paravirtualization, OS-level and application-level * Programming language support for virtualization * Hardware virtualization support: machine, IO, networking, storage * Virtual machine support for parallel and distributed computing models * Applications of virtualization: deployment, security, correctness, and reliability * Virtual-machine scheduling and resource allocation * Image management: storage, semantics, cloning and configuration * Processor and architecture simulators * Experiences with implementing virtualization Submission SYSTOR's virtualization workshop accepts only original work by the respective authors. Authors may submit their work either as an extended abstract (up to 2 pages) or as a full paper (up to 14 pages). Submissions should be in 11 pt font, in .pdf format. Submissions will be evaluated, after which notifications will be sent to the authors. All accepted works will be presented in a single track. No workshop proceedings will be published, but selected full papers will be published in a special issue of ACM Operating Systems Review. Extended abstracts will not be published. Work presented at SYSTOR may have been published elsewhere; in case of such an overlap, authors should add a reference to the additional publication to their submission. To submit, use the web submission system at http://papers.haifa.il.ibm.com/systor2007/ Important dates Submission August 16, 2007 Acceptance Notification September 13, 2007 Camera-ready (full papers) October 11, 2007 SYSTOR 2007 Virtualization Workshop October 29, 2007 Organizing committee * Erez Hadad, IBM Haifa Research Lab * Muli Ben-Yehuda, IBM Haifa Research Lab * Eliezer Dekel, IBM Haifa Research Lab * Miriam Allalouf, IBM Haifa Research Lab * Ben-Ami Yassour, IBM Haifa Research Lab * Yaron Wolfsthal, IBM Haifa Research Lab * Dalit Naor, IBM Haifa Research Lab * Assaf Schuster, Technion * Yitzhak (Tsahi) Birk, Technion Program committee * Assaf Schuster, Technion * Yitzhak (Tsahi) Birk, Technion * Roy Friedman, Technion * Sivan Toledo, Tel-Aviv University * Shlomi Dolev, Ben-Gurion * Leendert Van Doorn, AMD * Marc E. Fiuczynski, Princeton * Dilma Da Silva, IBM T.J. Watson * Bob Wisniewski,IBM T.J. Watson * Michael Factor, IBM Haifa Research Lab * Julian Satran, IBM Haifa Research Lab * Muli Ben-Yehuda, IBM Haifa Research Lab * Ilan Shimony, IBM Haifa Research Lab * Leah Shalev, IBM Haifa Research Lab * Avi Mendelson, Intel Center Israel * Shimon Shoken, Interdisciplinary Center Please feel free to further distribute this invitation to students and fellow researchers / developers. - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] another lecture proposal
On Tue, Apr 10, 2007 at 09:12:11PM +0300, Leon Romanovsky wrote: > I'm interested too. Cool, the lecture is scheduled for June 11th. Cheers, Muli - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Windows in Haifux
On Tue, Apr 10, 2007 at 07:11:09PM +0300, gabik wrote: > Hi > > Clubsys is highly academic and theoretical. They will not allow > lecture on "dirty, system-level" details, which is what interests > us, the engineers, I suppose. :) That's too bad... it's not like there's been a wealth of clubsys lectures. Have you tried talking with the organizers? in my several years of experience organizing such seminars, seminar organizers rarely say no to offers of help, especially if they're accompanied by a suggestion to lecture. I would love to see an "OS club" in the Technion --- but I don't think Haifux should be it. FWIW I am co-organizer of a weekly "Systems and Storage Research Seminar" at IBM's Haifa Research Lab, which has "OS lectures" fairly often. Email me privately if you'd like to be added to the announcement list. Cheers, Muli - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Windows in Haifux
On Tue, Apr 10, 2007 at 06:26:01PM +0300, gabik wrote: > I propose OS club. I think that's what clubsys is supposed to be... - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] lecture proposal
On Thu, Apr 05, 2007 at 07:38:05PM +0300, Kohn Emil Dan wrote: > > AFAIK the latest version of AIX (AIX 5L) is based on Linux (at least this > is what I heard that the 'L' stands for), I think there are a few hundred AIX developers who would be very suprised to hear that! Cheers, Muli - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
[Haifux] another lecture proposal
If there's interest, I'll be happy to give this talk I'll be giving at OLS '07 at Haifux as well. The Price of Safety: Evaluating IOMMU Performance IOMMUs, "IO Memory Management Units", are hardware devices that translate device DMA addresses to machine addresses. Isolation capable IOMMUs perform a valuable system service, preventing rogue devices from performing errant or malicious DMAs, thereby substantially increasing the system's reliability and availability. Without an IOMMU, a peripheral device could be programmed to overwrite any part of the system's memory. An isolation capable IOMMU restricts a device so that it can only access parts of memory it has been explicitly granted access to. Operating systems utilize IOMMUs to isolate device drivers; hypervisors utilize IOMMUs to grant secure direct hardware access to virtual machines. With the imminent publication of the PCI-SIG's IO Virtualization standard, as well as Intel and AMD's introduction of isolation capable IOMMUs in all new servers, IOMMUs will become ubiquitous. IOMMUs can impose a performance penalty due to the extra memory accesses required to perform DMA operations. The exact performance degradation depends on the IOMMU design, its caching architecture, the way it is programmed and the workload. In this paper, we present the performance characteristics of the Calgary and DART IOMMUs in Linux, both on bare metal and hypervisors. We measure the throughput and CPU utilization of several IO workloads with and without an IOMMU and analyze the results. We then discuss potential strategies for mitigating the IOMMU's costs. We conclude by presenting a set of optimizations we have implemented and the resulting performance improvements. Cheers, Muli - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] lecture proposal
On Mon, Apr 02, 2007 at 04:10:08PM +0300, boazg wrote: > i was wondering if there is enough interest for a lecture on AIX? I don't quite see how a talk about a proprietary OS fits Haifux's charter? Cheers, Muli - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Lecture suggestion: Open Solaris
On Thu, Jan 18, 2007 at 08:31:37AM +0200, guy keren wrote: > > yes please! > > > > Cheers, > > Muli > > a "me too" vote ;) While we're at it, if someone feels comfortable talking about MINIX v.3 I'd love to hear it. Cheers, Muli - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Lecture suggestion: Open Solaris
On Thu, Jan 18, 2007 at 08:34:17AM +0200, Rami Rosen wrote: > Hello, > > Sun had released Open Solaris about a year and a half ago ,under > CDDL license; It is probably going to be changed soon to GPLv3 (See: > http://www.linux-watch.com/news/NS8979755794.html). > > I suggest a general overview lecture about Open Solaris, which will > deal with Open Solaris open source model (with comparison to linux) > , Open Solaris advantages and disadvantages, and also with topics > like solaris streams ,solaris-linux interoperability, zones and the > like. > > The lecture can be scheduled from 12/2/07 onward. yes please! Cheers, Muli - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Kernel macro kdev_t_to_nr deprecated?
On Mon, Jan 01, 2007 at 10:20:20PM +0200, Eli Billauer wrote:
> It seems like the kdev_t_to_nr macro has disappeared from the kernel
> somewhere between 2.6.0 and now. There are also indications that saying
>
> #define kdev_t_to_nr(x) (x)
>
> is a good idea. Can anyone shed some light on this? Do I need to update
> my module?
http://lwn.net/Articles/2.6-kernel-api/
http://lwn.net/Articles/driver-porting/
Specifically, see http://lwn.net/Articles/22196/
> For those interested, typical use of the macro for checking which minor
> the inode was opened with:
>
> static int frandom_open(struct inode *inode, struct file *filp)
> {
>
>struct frandom_state *state;
>
>int num =MINOR(kdev_t_to_nr(inode->i_rdev));
>if ((num != frandom_minor) && (num != erandom_minor)) return
>-ENODEV;
int num = iminor(inode); /* untested! */
Cheers,
Muli
-
Haifa Linux Club Mailing List (http://www.haifux.org)
To unsub send an empty message to [EMAIL PROTECTED]
[Haifux] is there a talk today?
- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
[Haifux] meeting tomorrow?
Who's in charge of taking care of the room, now that Orr has become geographically-challenged? can said person confirm we have a room for Avi's excellent talk tomorrow? Cheers, Muli - Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] [HAIFUX LECTURE] DVD Authoring with Linux
On Sat, Nov 04, 2006 at 06:32:03PM +0200, Eli Billauer wrote:
> We are always looking for new lecturers and topics, and are scheduling
> the 2006 season. Got something interesting you wish to talk about? Got
> something new you want to learn, and need the drive of a lecture to make
> you learn it? Talk to us - [EMAIL PROTECTED]
FYI, and just in case the announcement for the next talk ends up going
out 30 minutes before the talk - we are going to have a guest talk on
Nov 13 that many of you will find highly interesting:
KVM (Kernel Based Virtual Machine) - Avi Kivity
Abstract
KVM (for Kernel-based Virtual Machine) is a new virtualization
hardware driver for Linux. kvm allows one to run multiple virtual
machines ("guests") on a single Linux host. Guests are isolated from
each other and from the host. It is similar to Xen and VMware but
differs in some important respects. The talk will cover the kvm
architecture and will include a short demonstration.
Cheers,
Muli
-
Haifa Linux Club Mailing List (http://www.haifux.org)
To unsub send an empty message to [EMAIL PROTECTED]
[Haifux] Xen
We might have an opening soon for a Xen hacker at the IBM Haifa Research Lab. Talented hackers of all persuasions, including students and PhDs, welcome. Please contact me for more details. Cheers, Muli -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Security issues in Linux
On Wed, May 10, 2006 at 12:17:24PM +0300, Ohad Lutzky wrote: > I love the whole live-to-press nature of it all... you'd think the > researchers would have discussed it with Mackall themselves first. I introduced Zvika and Matt over email after Zvika asked me to. I don't know if they actually corresponded. I too would've much preferred to see a patch rather than a press release, but ... *shrug*. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Security issues in Linux
On Wed, May 10, 2006 at 02:14:19AM -0400, Tzafrir Cohen wrote: > One small point that still bothers me: > > On Wed, May 10, 2006 at 01:22:18AM +0300, Michael Vasiliev wrote: > > > For the less security aware, there is the kernel support for hardware > > generators on the motherboard in the current kernel that is about as hard > > to > > get as running "make menuconfig" and enabling an option. (Well, maybe they > > miss it because they analyze the kernel source snapshot of December 2004, > > can > > anyone confirm?) > > Will that work on every motherboard? On every architecture? No, it depends on the existence of the HW RNG on a given board. > Anyway, has there been any discussion of their claims after the article > was published but before it made it to the press? Two monthes is a long > time. I also read somewhere that the authors claimed that they have > brought the problems to the attention of kernel developers but nothing > was done. > > Anybody with more information? I discussed this paper with Matt Mackall, the Linux /dev/random maintainer, a while ago. As far as I can recall, he thought most of the claims were pretty dated (i.e., known). He also thought there was one interesting bit, but we didn't get a chance to discuss it further. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Security issues in Linux
On Tue, May 09, 2006 at 11:09:34AM +0300, Ohad Lutzky wrote: > By the way, why would they have to _reverse engineer_ the kernel's > PRNG? Isn't it GPLd like the rest? Yeah, but apparently, they had trouble reading the code. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] I Wonder [was: Money]
On Mon, Mar 20, 2006 at 09:12:05AM +0200, Rafi Gordon wrote: > Assuming that you mean by this that they are developing (and selling) drivers > for linux, I wonder if anybody knows which drivers (and other > software) they are developing for Linux. On the drivers front Intel is involved at least with wireless (e.g., ipw2100, ipw2200), bridging and of course NICs (e1000). grepping the kernel Changelog will show tons of Intel submissions. Cheers, Muli -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Honey, I shrunk the club
On Tue, Mar 14, 2006 at 02:53:32PM +0200, Eli Billauer wrote: > Your opinions are mostly welcome. And if possible, please talk about why > *you* don't show up. I believe it's more interesting than general > theories about what happens. Thank you for raising this subject. Here are my reasons for not showing up any more except on very special occasions: - about half of the talks aren't interesting enough. Those include roughly anything which deals with "how to use Linux" as opposed to "how to hack Linux". Linux is mainstream. - I no longer feel the great sense of community we used to have. If you've been reading hamakr-discussions, you know all about this. - I'm just too busy working on Linux. Compared to working on it, talking about it holds very little appeal. I miss the days when we had Haifux *projects* - although if we had such a project today, I probably wouldn't have time or energy to work on it. Thanks for the trip down memory lane, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Call for Presentations for the Tel Aviv Linux Club
On Wed, Feb 15, 2006 at 10:36:14AM +0200, Shlomi Fish wrote: > We meet every two weeks on Sundays, in Schreiber building, room 008 of > the Tel Aviv University. Shlomi, kindly, cut the spam. From past experience, it works *much* better to approach potential speakers in private. Public announcements are fine but only every so often. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Ruby tutorial
On Fri, Jan 27, 2006 at 01:47:15AM +0200, guy keren wrote: > i once started reading a tutorial on ruby, just to see what's all the fuss > about, and gave out in the middle - if i see no idea after reading for > half an hour, i dim the whole exercise pointless. Ah, but what do you think of haskell? and the bonus question, was it invented by a japanese? Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Ruby tutorial
On Wed, Jan 25, 2006 at 11:07:13PM +0200, Ohad Lutzky wrote: > All present hackers, you'll much enjoy this - an excellent Ruby tutorial. > Starts > out slow, but proceeds with really cool examples. > > http://tryruby.hobix.com If I may troll a bit: why ruby? what does it have that (pick your favorite language) doesn't have? Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] compiling
On Tue, Dec 27, 2005 at 02:02:06PM +0200, yakouba abaya wrote: > suppose i have : ABC.h , A.c , B.c , C.c > is it possible to build all three source files to one object file ? > > gcc -c A.c | gcc -c B.c | gcc -c C.c > gcc A.o B.o C.o -o ABC.o ??? Not without extensive trickery. Why would you want to do that? > inorder to do : gcc ABC.o app.c -o app > instead of : gcc A.o B.o C.o app.c -o app If that's what you want, add the seperate object files into a single static library. I fail to see what's the point, though. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Suggestion for a haifux lecture
On Wed, Dec 07, 2005 at 12:19:52PM +0200, Paramahansa Polo wrote: > For a description of udev, and what its original design goals were > http://www.kroah.com/linux/talks/ols_2003_udev_paper/Reprint-Kroah-Hartman-OLS2003.pdf > > Some slides material for a talk > http://www.kroah.com/linux/talks/ols_2003_udev_talk/ Sounds great, looking forward to it. You might also wish to submit it to OSDC's kernel and hypervisor track (http://www.osdc.org.il). Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
[Haifux] Re: [W2L] Re: [Haifux] Re: [W2L] w2l - some numbers, good notes about yesterday's lecture and a request
On Tue, Dec 06, 2005 at 02:55:39PM +0200, Alon Altman wrote: > This one: http://haifux.org/lectures/86-sil/kernel-modules-drivers/ Thanks. Looks good. My only comment is that new (2.6 based) module utilities usemodprobe.conf rather than modules.conf, with a simplified subset of the syntax. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Re: [W2L] w2l - some numbers, good notes about yesterday's lecture and a request
On Tue, Dec 06, 2005 at 11:47:09AM +0200, Alon Altman wrote: > I think my "permissions preview" in the previous lecture was enough for > this, but I think it interests most people. Another option is guy's kernel > lecture, as I got A LOT of questions about drivers. Actually, I think I like > this option (kernel) better, even though I probably won't finish it all in > one meeting. Which kernel lecture? Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
[Haifux] IBM HRL storage and system, compilers and architecture seminars
The following seminars at IBM HRL might interest you, considering the non-negligilbe overlap with past and future Haifux lectures. Storage and Systems Dept Seminar: http://www.haifa.il.ibm.com/Workshops/systems-and-storage2005/agenda.html Compilers and Architecture Dept Seminar: http://www.haifa.il.ibm.com/Workshops/compiler2005/agenda.html Entrance is free to the public, but please register in advance. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] IOU - from the lecture today
On Tue, Nov 15, 2005 at 09:29:47AM +0200, Alon Altman wrote: > Please limit the spread of misinformation to your lecture. Kindly discourse politely, or not at all. > The true reason for these messages is the fact that valgrind does not > suppress certain errors generated by glibc and the linux dynamic loader. You > have probably upgraded your standard libraries without upgrading valgrind's > suppression files. This might or might not be the reason, but what Orna said is absolutely correct: upgrading Valgrind caused the warning messages to go away. > More importantly, the reason that the memory leaks were not detected was > the fact that valgrind was run without the --leak-check=full flag, and thus > leaks were only counted but not reported. They were detected. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Lecture idea - WiFi (802.11) in GNU/Linux
On Wed, Nov 02, 2005 at 11:55:14AM +0200, Ohad Lutzky wrote: > Two components I was thinking of for this lecture - > - An introduction to how WiFi is used in GNU/Linux > (distro-independant, but with linux-wireless-utils, so not relevant > BSD or Mac) - tools like iwconfig, iwlist, along with some information > about the availability of wireless NIC modules - especially the > ipw2200, which is included in all/most new Centrino laptops > -- User-friendliness in these utilities - a summary of the problems > with using the commandline for this (tedious and daunting for new > users, not automated), a survey of current available solutions (UIs > like netapplet, wifi_radar and my own netdlg, along with automated > solutions such as hotplug, ifplugd, waproamd), pros and cons to each > and what basically needs to be done to improve them. Sounds good, especially if it was accompanied by another lecture on the wifi support in the kernel, different wifi standards, etc. Is anyone interested in preparing such a lecture? Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Re: [W2L] unanswered questions
On Tue, Sep 13, 2005 at 03:59:12PM +0300, Adir Abraham wrote: > I also don't see anybody who is *really against* it. OK, the ones who are > truly agaisnt it are known - you, Orna and Alon. SO? Since you're counting, I think having an instaparty became pointless a couple of years ago. 5 installations in a whole day should tell you something. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Real-time write on *ANY* filesystem
On Wed, Jun 22, 2005 at 03:08:45PM +0200, Eli Billauer wrote: > As for the results you posted: It's the peaks I'm after, not the tail. > The peaks appear anywhere in the list. So the best thing is to draw a > graph of these numbers. This is the tail of the distribution - i.e the peak. (generated via sort -n $file | tail -15) Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Real-time write on *ANY* filesystem
On Wed, Jun 22, 2005 at 11:07:51AM +0200, Eli Billauer wrote: > Muli Ben-Yehuda wrote: > > >>mknod mypipe p > >>mbuffer -i mypipe -o /fatfs/output-file & > >>./writefat mypipe > listfile > >Until 2.6.mumble, pipes only used a single page in memory. Since > >2.6.mumble we're using up to 16 pages and flipping between consumer > >and producer, which should give much better pipe utilization for large > >writes. > Note that mbuffer is the RAM FIFO, and it was empty all the time (as one > could expect). Since mbuffer never blocked, I don't think it matters how > good the pipe between them is. This is why I found it weird that I got > delays at all, using a RAM FIFO. Where can I find the sourcve for mbuffer? > >- use write, not fwrite!!! > >- use DIRECT_IO to bypass kernel caching > >- use the appropriate IO elevator > > > Are these general guidelines for writing fast I/O, or are there good > reasons to suspect that one of these cause occasional long blocks? All of these are candidates. > It would be nice if some of you tried the program I sent, and let's see > if you get the same results. Note that the real action begins when the > partition you write to is getting full. I ran four tests, 2 on 2.4.30 and 2 on 2.6.11-rc2 on my slow SCSI test machine. Each was run against an ext3 fs (.orig) and against my swap fs (.raw) I'm attaching the tail end of the distribution for each case. What this tells me is: - that the file system definitely gets in the way - that 2.6 is much better (no suprises there) - that adding both DIRECT_IO and (especially) kernel AIO will do wonder to improve the results. I might take a stab at adding both later and measuring. Which kernel are you using? listfile.2.4.orig: 1033934 1039532 1050982 1062859 1072628 1074470 1095608 1117049 1117059 1130159 --- listfile.2.4.raw: 982159 989523 992505 1001441 1002068 1017735 1021283 1023135 1064356 1089229 --- listfile.2.6.orig: 56340 70775 118317 149468 154185 172567 174960 271780 359151 1113761 ------- listfile.2.6.raw: 94354 94550 94693 98262 99790 112159 124361 147720 186502 192103 --- -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] Real-time write on *ANY* filesystem
On Wed, Jun 22, 2005 at 02:41:39AM +0200, Eli Billauer wrote: > And finally: Does an RAM FIFO help? Surprisingly, the answer is no. I > did the following: > > mknod mypipe p > mbuffer -i mypipe -o /fatfs/output-file & > ./writefat mypipe > listfile Until 2.6.mumble, pipes only used a single page in memory. Since 2.6.mumble we're using up to 16 pages and flipping between consumer and producer, which should give much better pipe utilization for large writes. > Insights, anybody? Yeah, how about you cut out the various middlemen from the code? at least it's not in Java... - use write, not fwrite!!! - use DIRECT_IO to bypass kernel caching - use the appropriate IO elevator - verify that your disk drivers are tuned for whatever you want to do (is DMA on?) - what else is the system doing? is it idle? busy? is anything else interfering with the scheduling of your program? Linux is a general purpose OS, which means it's good for a lot of things and optimal for none. If you want it to be optimal for your specific usage, you should spend some time optimizing and tuning it for it. And that's true regardless of what your usage happens to be. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] ACPI on IBM R40
On Fri, May 06, 2005 at 06:15:03PM +0300, Mark Silberstein wrote: > Suspend to disk is a good idea, but it doesn't really work for my > Radeon7500 video card, i.e. it suspends, but comes back with garbaged > display, impossible to see anything. Can't help you with the ACPI problem (I've been trying to steer clear of it for years now and am still in apm land) but switching to a text console before suspending might do the trick, using e.g. chvt(1). Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
[Haifux] Fwd: [IBM HRL TALK] Towards Hard Realtime Response from the Linux Kernel: Adapting RCU to Hard Realtime
Reminder, this talk will take place tomorrow and is highly recommended to Linux kernel afficiandos. - Forwarded message from Muli Ben-Yehuda <[EMAIL PROTECTED]> - Date: Sun, 17 Apr 2005 22:34:35 +0300 From: Muli Ben-Yehuda <[EMAIL PROTECTED]> User-Agent: Mutt/1.5.6+20040907i To: haifux@haifux.org Subject: [Haifux] [IBM HRL TALK] Towards Hard Realtime Response from the Linux Kernel: Adapting RCU to Hard Realtime Paul McKenney of the IBM Linux Technology Center will be visiting IBM's Haifa Research Lab on May 3rd and giving a talk. I highly recommend attending - Paul's talks are insightful and entertaining, and it's a very interesting subject, IMENHO. For those who want to attend, this is a regular HRL seminar, so there's no need to contact me in advance. the seminars are open to the public. IBM Haifa Research Labs are located on the Haifa University campus, Mt. Carmel. Parking will be available only for those who reserve a place by calling 04 - 8296100 at least one day in advance. Be sure to bring a valid picture ID, to be presented at reception. Contact me if you have any questions. Title: Towards Hard Realtime Response from the Linux Kernel: Adapting RCU to Hard Realtime Speaker: Paul McKenney, IBM Beaverton, Oregon May 3rd, 11:00 - 12:00, IBM Haifa Labs Auditorium, Haifa University Campus Traditionally, realtime response has been designed into operating systems offering it. Retrofitting hard realtime response into an existing general-purpose OS is not impossible, but is quite difficult: all non-preemptive code paths in the OS must have deterministic execution time. Realtime capabilities are nonetheless being added to Linux: the preemptible-kernel facility added to the 2.6 kernel enables surprisingly good soft realtime, and Ingo Molnar's CONFIG_PREEMPT_RT patch is producing amazing results: 1-microsecond average scheduling latency with 20-microsecond measured maximum latency. This is still soft realtime, but it is good enough for all but the most demanding applications. Other approaches have been proposed, and are summarized in this talk. The advent of aggressively multithreaded CPUs and multi-core dies brings a new challenge: can Linux offer realtime response on multiprocessor systems? In the past, one obstacle to realtime response on SMP systems has been RCU, which disables preemption throughout read-side critical sections. In addition, RCU's deferral of freeing can cause problems for memory-constrained systems. This talk describes some novel implementations of RCU that address these problems while still permitting reasonable performance and scalability. -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ ----- End forwarded message - -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] RAM FIFO
On Wed, Apr 27, 2005 at 02:08:55PM +0200, Eli Billauer wrote: > Now, since I use a common UNIX pipe to store the file (tee, actually), > the disk blocks, and the stream is broken. So I thought: Hmmm, how > about a simple application, which allocates, say, 16 MB of > non-pageable RAM (peanuts in today's terms), behaving like a simple > data FIFO, which I put in the pipeline? This would give plenty of time > solve whatever is more urgent than writing to the disk. I am not familiar with such an application but it should be trivial to write. Two other things you might want to consider: - tee sucks for performance. Unless you have particular reason to use tee, I recommend doing stdio -> file, and tail -f on the file seperately. - 2.6.11 and onwards kernels have a much improved pipe handling code. If the disk isn't your only bottleneck, you might want to try this. Also, you can play various games with the disk and application in question to reduce (or eliminate) the disk problems. Things you can try include hdparm, elevator settings, AIO and DIO. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
[Haifux] [IBM HRL TALK] Towards Hard Realtime Response from the Linux Kernel: Adapting RCU to Hard Realtime
Paul McKenney of the IBM Linux Technology Center will be visiting IBM's Haifa Research Lab on May 3rd and giving a talk. I highly recommend attending - Paul's talks are insightful and entertaining, and it's a very interesting subject, IMENHO. For those who want to attend, this is a regular HRL seminar, so there's no need to contact me in advance. the seminars are open to the public. IBM Haifa Research Labs are located on the Haifa University campus, Mt. Carmel. Parking will be available only for those who reserve a place by calling 04 - 8296100 at least one day in advance. Be sure to bring a valid picture ID, to be presented at reception. Contact me if you have any questions. Title: Towards Hard Realtime Response from the Linux Kernel: Adapting RCU to Hard Realtime Speaker: Paul McKenney, IBM Beaverton, Oregon May 3rd, 11:00 - 12:00, IBM Haifa Labs Auditorium, Haifa University Campus Traditionally, realtime response has been designed into operating systems offering it. Retrofitting hard realtime response into an existing general-purpose OS is not impossible, but is quite difficult: all non-preemptive code paths in the OS must have deterministic execution time. Realtime capabilities are nonetheless being added to Linux: the preemptible-kernel facility added to the 2.6 kernel enables surprisingly good soft realtime, and Ingo Molnar's CONFIG_PREEMPT_RT patch is producing amazing results: 1-microsecond average scheduling latency with 20-microsecond measured maximum latency. This is still soft realtime, but it is good enough for all but the most demanding applications. Other approaches have been proposed, and are summarized in this talk. The advent of aggressively multithreaded CPUs and multi-core dies brings a new challenge: can Linux offer realtime response on multiprocessor systems? In the past, one obstacle to realtime response on SMP systems has been RCU, which disables preemption throughout read-side critical sections. In addition, RCU's deferral of freeing can cause problems for memory-constrained systems. This talk describes some novel implementations of RCU that address these problems while still permitting reasonable performance and scalability. -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
[Haifux] Re: [HAIFUX LECTURE] The Xen hypervisor, by Muli Ben-Yehuda
On Sun, Apr 10, 2005 at 08:51:39AM +0300, Muli Ben-Yehuda wrote: > Next Monday (11/04/2005), 18:30, the Haifa Linux Club will once again > meet to hear Muli Ben-Yehuda (aka me) talk about > > The Xen hypervisor > > You are all invited! More info on how to get to the meeting etc. at > the haifux website: http://www.haifux.org. Slides are now available at http://www.mulix.org/lectures/xen-haifux-apr-2005/xen-haifux.pdf. I highly recommend going to the source at http://www.cl.cam.ac.uk/Research/SRG/netos/xen/. Cheers, Muli signature.asc Description: Digital signature
[Haifux] [HAIFUX LECTURE] The Xen hypervisor, by Muli Ben-Yehuda
Next Monday (11/04/2005), 18:30, the Haifa Linux Club will once again meet to hear Muli Ben-Yehuda (aka me) talk about The Xen hypervisor You are all invited! More info on how to get to the meeting etc. at the haifux website: http://www.haifux.org. Topics to be covered include a general introduction, design and implementation of Xen, the new Xen 2.0 IO model and future plans. [1] http://www.cl.cam.ac.uk/Research/SRG/netos/xen/index.html [2] A short overview of Xen, from http://www.mulix.org/lectures/OLS2004.html: Xen is a virtual machine monitor, developed by the University of Cambridge Computer Laboratory. Unlike VMWare, which provides complete virtualization, guest operating systems need to be ported to the Xen environment. So far, Linux 2.4 and 2.6 have been ported, as well as NetBSD, FreeBSD and Plan9, and Windows XP. The Windows XP port was done in collaboration with MS Research, and took much longer than the Linux port... Xen works by letting the monitor (hypervisor) run in ring 0, and the guest OS run in ring 1. Userspace runs in ring 3, as usual. From a Linux point of view, porting Linux to Xen (refereed to as XenoLinux) is just a matter of implementing the arch specific hooks in Linux - no core kernel files are modified! Xen provides secure protection between VMs (unlike e.g. coLinux), allows flexible partitioning of resources, and supports seamless low-latency migration of running VMs(!). They also claims impressive performance numbers, within 3% of the host performance. -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
Re: [Haifux] [IBM HRL TALK] Overview of AIX Kernel
On Mon, Apr 04, 2005 at 11:31:58PM +0200, Sivan Green wrote: > Will there be anything similar like this session in Tel Aviv ? I'm afaid not, unless someone from the lab chooses to organize it. Haifa is not that far away ;-) Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
Re: [Haifux] [IBM HRL TALK] Overview of AIX Kernel
On Mon, Apr 04, 2005 at 02:53:18AM +0300, guy keren wrote: > > Title: Overview of AIX kernel > > is this a continuation of the meeting that took place a month or two > back? Yes, it is. We'll continue from the point we left off at the time. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
[Haifux] [IBM HRL TALK] Overview of AIX Kernel
The next IBM Haifa Research Lab Linux Study Group talk will take place on Tuesday, 05/04/2005, at 1400 at IBM HRL (http://www.haifa.il.ibm.com). Anyone interested in attending should contact me privately for arrangements. Title: Overview of AIX kernel By: Kalman Meth Abstract: We will briefly discuss the following topics, time permitting, as far as we get: AIX System kernel components PowerPC hardware overview AIX Address translation AIX Process management AIX File system AIX Kernel extentions and device drivers Comparison to Linux kernel will be made in an open discussion manner. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] system stuck after moving to kernel2.6.11
On Tue, Mar 29, 2005 at 08:25:04AM +0200, Aviv Goll wrote: > hi, > I am using ubuntu 5.04 an some time ago i tried to move to kernel > 2.6.11 using the synaptic application to install the 2.6.11 kernel. > Synaptic even set up my boot loader and all i had left to do was > reset. the problem was that after reset gnome started and a few > seconds later (a few seconds past the splash screen) the system > completely stuck leaving me with only the option of phisical reboot. > yesterday i tried compiling the 2.6.11 kernel and got the exect same > resaults. > > I don't know where to find the logs of the booting so i don't know > where does it fail... Are there any messages on screen? if yes, where does it stop and what does it say? if not, you need to hook up a serial console and get the boot messages on a remote machine. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
[Haifux] [IBM HRL TALK] Infiniband Support on Linux - Status and Directions
The next IBM Haifa Research Lab Linux Study Group talk will take place on Tuesday, 22/03/2005, at 1400 at IBM HRL (http://www.haifa.il.ibm.com). Anyone interested in attending should contact me privately for arrangements. Lecture Title: InfiniBand support on Linux: status and direction By : Michael S. Tsirkin, Mellanox Technologies Ltd Abstract InfiniBand is a switched interconnect fabric standard featuring high bandwidth and low latency, with RDMA capabilities. It can run at link speeds up to 120 gigabit per second (Gb/s). Current implementations support up to 20Gb/s wire speed on Host Channel Adapters (HCAs), and 60Gb/s on switches. InfiniBand is finding wide usage in connecting clusters of servers and as a SAN (storage-area-networking) technology. It is also widely used in the high performance computing community, powering e.g. the Columbia, #2 Supercomputer in the Top500 Supercomputer Sites (Nov 2004 list), peaking at 51.87 TFlops. This talk will provide an overview of the interconnect/adapter capabilities, and of the openib.org Linux InfiniBand stack. We will discuss benefits and challenges of utilizing, under Linux, hardware features such as: - Hardware reliability - Zero-copy support - OS bypass support - Remote memory access (RDMA) Furthermore, the status of various ULPs (Upper Layer Protocols) such as IP over IB, SDP, iSCSI, uverbs, and others - under Linux - will be described. -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
Re: [Haifux] Ubuntu Linux Giveaway Event
On Sat, Mar 05, 2005 at 06:49:53PM +0200, Sivan Green wrote: > Are there any organization that would like to backup this event? IBM-HRL > etc? (I would love to do it in IBM-HRL, if this is interesting) Not speaking for IBM-HRL, but I don't see the connection (feel free to enlighten me off-list). I think you should contact Hamamkor. > I would like to hear if this sounds interesting, and suggestion where > would such event be best presented? For starters, a (technical) talk about your work on Ubuntu in Haifux would be interesting. It sounds like you want to do an "evangelism" sort of thing, though. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
Re: [Haifux] Lecture proposals
On Mon, Feb 14, 2005 at 08:19:41PM +0200, Alon Altman wrote: > Lecture proposals for Haifux: > 1. Software Suspend II. May evolve to a "linux on laptop" lecture. > Software Suspend is a mechanism to allow you to retain your system state > when your system power is off. It allows you to quickly power off and on > your system. Especially useful for laptops. Yes > 2. lirc - Linux Infrared Remote Control - Controlling your PC with a > standard remote - how and why. Not really, but don't mind me ;-) > 3. Firewal Piercing, or how to bypass draconian restrictions and do what > YOU want with your internet connection. I will discuss ppp over SSH and > running everything over HTTP and/or DNS. Sure Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
[Haifux] [IBM HRL TALK] Overview of AIX kernel
The next IBM Haifa Research Lab Linux Study Group talk will take place on Tuesday, 08/02/2006, at 1400 at IBM HRL (http://www.haifa.il.ibm.com). Anyone interested in attending should contact me privately for arrangements. Title: Overview of AIX kernel By: Kalman Meth Abstract: We will briefly discuss the following topics, time permitting, as far as we get: AIX System kernel components PowerPC hardware overview AIX Address translation AIX Process management AIX File system AIX Kernel extentions and device drivers Comparison to Linux kernel will be made in an open discussion manner. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
Re: [Haifux] device driver
On Sun, Jan 30, 2005 at 11:42:29AM +0200, yakoub abaya wrote: > struct file repesents a proccess operating on a struct inode. > each file has a private_data , as i've read this field will > containe the struct we declare in the module code That's the usual way of doing it. > 1) but what if we declared two structs ? what about other >static variables we might declare ? It is up to you to decide what (if anything) to open into filp->private_data at open() time. > 2) is it possible for all file operations methods > to operate on the same global struct declared in >the module code without using private_data ? Yes, but it is not considered good design to have global data. Chees, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
[Haifux] Introduction to Linux Device Drivers talk
I will be giving a rerun of the "Introduction to Linux Device Drivers"[0] talk to the Operating Systems class in the Technion this Wednesday, 1030-1230, Taub 1 in the CS faculty. This is the same talk that was given to Haifux and to the OS class in previous sememsters, with minor updates[1]. Everyone is invited... [0] http://www.mulix.org/lectures/intro_to_linux_device_drivers/intro_linux_device_drivers.pdf [1] The code is updated to 2.6.11-rc1-bk9. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
Re: [Haifux] A lecture, maybe, if you like?
On Wed, Jan 19, 2005 at 12:11:30AM +0200, Oron Peled wrote: > Hi, > > Does anybody want to hear a (yet unwritten) lecture about > > Gnucash: How I manage my money and stay alive to > tell the story... > > In which I'll present a great piece of free software which > is not (yet?) famous. Definitely! Sooner is better, too ;-) Cheers, Mull -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
Re: [Haifux] linux ipc
On Sun, Jan 16, 2005 at 09:29:26PM +0200, Oron Peled wrote: > And to make his life more interesting: NETLINK sockets... Let's not forget relayfs, if you're using 2.6.11-rc1-mm1 ;-) (http://kerneltrap.org/node/4546) Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
Re: [Haifux] linux ipc
On Sun, Jan 16, 2005 at 01:40:28PM +0200, yakoub abaya wrote: > Is there a special interface for communication > between : device driver <-> daemon , > other than conventional ipc ? > ( i know of pipes , signal & kill ) read and write on a device file, a new file system, /sysfs, /procfs, ioctl (not a good idea), add your own syscalls, etc, etc. A better question is what do you want to do? Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ -- Haifa Linux Club Mailing List (http://www.haifux.org) To unsub send an empty message to [EMAIL PROTECTED]
Re: [Haifux] signal
On Thu, Jan 13, 2005 at 08:35:20AM +0200, yakoub abaya wrote: > when doing man signal : > there is : > typdef void (*sighandler_t)(int); > Q:what is this typdef ?, i don't understand it It says that sighandler_t has the type "pointer to a function that accepts an int and returns void". Consult a good C book for more details. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
Re: [Haifux] module directives ?
On Thu, Jan 13, 2005 at 08:31:30AM +0200, yakoub abaya wrote: > i've been reading about driver porting to kerne2.6 > the section about block device . > > they write : >static int __init sbd_init(void); <__init >static void __exit sbd_exit(void); < __exit > >module_init(sbd_init); >module_exit(sbd_exit); > Q: what is __init and __exit ? are they preprocessor directives ? Yes. They say that this code can be discarded in some cases, to make a slightly smaller kernel. __init says that this is initialization code. Once the kernel or module has finished initializing, it can be discarded. __exit says that this is finalization code. If module unloading is disabled, it can be discarded. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
[Haifux] Xen and the art of giving talks
Dear Haifuxians, Would you care to hear a talk on virtualization, hypervisors and Xen[1][2]? Topics to be covered include a general introduction, design and implementation of Xen, the new Xen 2.0 IO model and future plans. [1] http://www.cl.cam.ac.uk/Research/SRG/netos/xen/index.html [2] A short overview of Xen, from http://www.mulix.org/lectures/OLS2004.html: Xen is a virtual machine monitor, developed by the University of Cambridge Computer Laboratory. Unlike VMWare, which provides complete virtualization, guest operating systems need to be ported to the Xen environment. So far, Linux 2.4 and 2.6 have been ported, as well as NetBSD, FreeBSD and Plan9, and Windows XP. The Windows XP port was done in collaboration with MS Research, and took much longer than the Linux port... Xen works by letting the monitor (hypervisor) run in ring 0, and the guest OS run in ring 1. Userspace runs in ring 3, as usual. From a Linux point of view, porting Linux to Xen (refereed to as XenoLinux) is just a matter of implementing the arch specific hooks in Linux - no core kernel files are modified! Xen provides secure protection between VMs (unlike e.g. coLinux), allows flexible partitioning of resources, and supports seamless low-latency migration of running VMs(!). They also claims impressive performance numbers, within 3% of the host performance. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
[Haifux] Xen and the art of giving talks
Dear Haifuxians, Would you care to hear a talk on virtualization, hypervisors and Xen[1][2]? Topics to be covered include a general introduction, design and implementation of Xen, the new Xen 2.0 IO model and future plans. [1] http://www.cl.cam.ac.uk/Research/SRG/netos/xen/index.html [2] A short overview of Xen, from http://www.mulix.org/lectures/OLS2004.html: Xen is a virtual machine monitor, developed by the University of Cambridge Computer Laboratory. Unlike VMWare, which provides complete virtualization, guest operating systems need to be ported to the Xen environment. So far, Linux 2.4 and 2.6 have been ported, as well as NetBSD, FreeBSD and Plan9, and Windows XP. The Windows XP port was done in collaboration with MS Research, and took much longer than the Linux port... Xen works by letting the monitor (hypervisor) run in ring 0, and the guest OS run in ring 1. Userspace runs in ring 3, as usual. From a Linux point of view, porting Linux to Xen (refereed to as XenoLinux) is just a matter of implementing the arch specific hooks in Linux - no core kernel files are modified! Xen provides secure protection between VMs (unlike e.g. coLinux), allows flexible partitioning of resources, and supports seamless low-latency migration of running VMs(!). They also claims impressive performance numbers, within 3% of the host performance. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
Re: [Haifux] BSD lecture
On Thu, Dec 30, 2004 at 02:22:51PM +0200, Ido Barnea wrote: > Is the lecture still wanted? Yes. > Are there any special requests about the contents of the second > part? Nope. > Anything else. Sounds great! if you prefer to split it into two lectures, that might be even better. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
[Haifux] [HAIFUX LECTURE] C Without a Spoon- Programming in C in a free software environment - Orna Agmon
Next Monday (27/12/2004), 18:30, the Haifa Linux Club will once again meet to hear Orna Agmon talk about: C Without a Spoon Programming in C in a free software environment In this series of lectures we aim to introduce good C programming in the environment of free software, to people who have already programmed in some language using procedural programming (In other words, we assume you know of some form of the contruct if-then-else, you know what variables and functions/subroutines are). The series is composed of units, each of 1-2 hours. The first is intended for people who are not familiar with C, and the rest of it is set according to topics, such that people who already program in C might learn something new from them, and still those who were introduced to C only at the first lecture will not get lost. The lectures will include Homework. These will be food for thought in most cases, or reading a man page, and writing a program of about 15 lines at some cases. We will go over the homework at the beginning of each lesson, and further lectures may assume having prepared the homework. Since this is a long series of lectures, we will only carry it through as long as there is an interest. We will begin by bringing everyone to the same level: the first lecture will include a quick description of the basic C constructs (if, while, for, switch, ternary if). We will discuss standards, where to get help, blocks and scopes, and we will be able to compile and run a "Hello World" program at the end of it. On the second unit we will learn about the CPP - C preprocessor, and about gcc - the free C compiler, and about using its options cleverly. Slides and course material are available at: http://haifux.org/~ladypine/115-sil/ We meet in the Technion, Taub 3. See http://www.haifux.org/where.html for arrival details. Attendance is free, and you are all invited! We are always looking for interesting lecture ideas, and we have already begun scheduling the 2005 season! Future lectures include: 116 IPv6Dani Arbel 3/1/2005 117 gcc and vectorization Dorit Naishlos 17/1/2005 118 SNMP and OpenNMSZeev Halevi 31/01/2005 119 The VFS of the Linux-2.4 kernel - a play in 5 acts Guy Keren 14/02/2005 Happy St. Valentine's Day! 120 Writing a STAM disk-based file-system for the Linux-2.4 kernel Guy Keren 28/02/2005 Prior understanding required . Have a subject you want to talk about? Or a subject you'd like to hear someone else talk about? email us. -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
[Haifux] Re: [IBM HRL TALK] The Pentium Architecture - Support for Multiprocessing Operating Systems
On Fri, Nov 05, 2004 at 08:59:52PM +0200, Muli Ben-Yehuda wrote: Haifuxers, Edi has graciously agreed to give the this talk to Haifux, if there's interest. I think that those of you interested in the low level OS details would enjoy it quite a lot. Interested? Cheers, Muli > Title : The Pentium Architecture - Support for Multiprocessing > Operating Systems > > By : Edi Shmueli > > Abstract > General purpose processors supply a set of mechanisms to support > multiprocessing operating systems. Among these mechanisms are the > memory management facilities, protection mechanism, interrupts > handling mechanisms and even mechanisms for process management. > > We will cover these mechanisms as they are provided by the Pentium > processor family, and briefly discuss their usage by the Linux > operating system. -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
[Haifux] [IBM HRL TALK] The Pentium Architecture - Support for Multiprocessing Operating Systems
The next talk will take place on Tuesday, 09/11/2004, at 1400 at IBM HRL (http://www.haifa.il.ibm.com). Anyone interested in attending should contact me privately for details. Title : The Pentium Architecture - Support for Multiprocessing Operating Systems By : Edi Shmueli Abstract General purpose processors supply a set of mechanisms to support multiprocessing operating systems. Among these mechanisms are the memory management facilities, protection mechanism, interrupts handling mechanisms and even mechanisms for process management. We will cover these mechanisms as they are provided by the Pentium processor family, and briefly discuss their usage by the Linux operating system. Keywords: paging, segmentation, privilege/protection rings, interrupts and exceptions, context switching. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
[Haifux] IBM Linux Study Group announcements on or off topic?
Hellow, Haifuxers, IBM's Haifa Research Lab (http://www.haifa.il.ibm.com) holds more or less regular biweekly Linux study sessions, modeled in part after Haifux, and occasionally sharing speakers and themes. I am one of the organizers, so you can guess what are our usual topics. Would you be interested in receiving announcements of these biweekly sessions? Let me know. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
Re: [Haifux] Proposed Lecture - Linux USB Subsystem Layer (in Video4Linux Context)
On Thu, Oct 28, 2004 at 02:21:40PM +0200, guy keren wrote: > yes, yes, yes! we _want_ technical lectures. dying for technical lectures. > if you need two lectures to supply the material - we'll be very happy > (e.g. if you can give a 2nd talk about actually developing USB drivers, > for kernel programmers, that will be lovely). Seconded. Looking forward to the talk. Cheers, Muli -- Muli Ben-Yehuda http://www.mulix.org | http://mulix.livejournal.com/ signature.asc Description: Digital signature
