1 / 38

Research log (last update 2004/11/10)

Research log (last update 2004/11/10) . Yao-Chung Chang. Half-year Plan . Weekly Plan (2004/11/8). Plan for last week Discussing with Yi-Neng about the new architecture Preparing for the midterm exam Plan for this week Arranging the slides Soc lab of microblaze Fooling around Notes

Download Presentation

Research log (last update 2004/11/10)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Research log (last update 2004/11/10) Yao-Chung Chang

  2. Half-year Plan

  3. Weekly Plan (2004/11/8) • Plan for last week • Discussing with Yi-Neng about the new architecture • Preparing for the midterm exam • Plan for this week • Arranging the slides • Soc lab of microblaze • Fooling around • Notes • Finally, after converging the issues of my thesis in the research meeting last week, I’ve had a bit of assurance. In this week, I plan to think about the problem that whether fragmentation is necessary and arrange the slides as well.

  4. Weekly Plan (2004/10/18) • Plan for last week • Implementation and simulation of simple receive-process-transmit model on IXP 2400 • SoC lab & verilog learning • Plan for this week • Design the dispatcher for pool of threads • SoC lab & verilog learning • Notes • After discussion with Yi-Neng, the dispatcher is going to be shaped up. I have to well describe the algorithm of the dispatcher and meanwhile arrange the chapters that would be in my thesis. Regards 國防役, I’ve decide to apply for entering the IRTI, so that I can have a regular life and time for learning interesting stuff.

  5. Weekly Plan (2004/10/11) • Plan for last week • Modify existing string matching algorithm for IXP 2400 • ACA study • SoC lab & verilog learning • Writing resume for 國防役 • Plan for this week • Implementation and simulation of simple receive-process-transmit model on IXP 2400 • SoC lab & verilog learning • Notes • In order to observe the behavior of aho-coraisck algorithm on network processor, I’ve started to modify existing sample codes, say receive, count and transmit. It’s a hard work nonetheless. Besides understanding the hardware specification for further development, I need to design a mechanism, say a dispatcher, for handling flows by assigning threads from pool of threads. Suddenly, I feel the days passed quickly, as regards research, I have to pull my weight as possible as I can.

  6. Weekly Plan (2004/10/04) • Plan for last week • Converge the results of discussion • List the hardware feature of IXP 2400 • Book reading • Plan for this week • Modify existing string matching algorithm for IXP 2400 • ACA study • SoC lab & verilog learning • Writing resume for 國防役 • Notes • In this week, I get started implementing aho-corasic algorithm on IXP 2400. After discussion with Yi-Neng in last week, I decide to focus on how to take advantage of hardware feature in order to speed up the aho-corasic algorithm. Meanwhile, I’ll try to get familiar with the development environment as soon as possible.

  7. Weekly Plan (2004/09/20) • Plan for this week • Book reading • The GUI of Wall • Notes • Lately, a new semester has been beginning. In this week, I plan to concentrate on the research besides all the chores, say Wall GUI. In regard to wall GUI, there are still many parts needed to be done, and hope I can “terminate” it as soon as I can.

  8. Weekly Plan (2004/08/23) • Plan for last week • Reading IXP2400/2800 Programming Guide • Defining the architecture as well as the criteria of threads and microengines allocation • Excellent project demo • Plan for this week • Giving a presentation of thesis proposal • The GUI of Wall • Notes • The group meeting this week is a show time for me to give a presentation of proposal, though there are many ideas I need to make them more concrete. In this week, I plan to discuss with my mentor, Yi-Neng, about the guidelines of thread and microengine allocation, so that I can make my proposal more logical and reasonable.

  9. Weekly Plan (2004/08/16) • Plan for last week • Reading IXP2400/2800 Programming Guide • Understating how packet inspection is done in Snort • Plan for this week • Reading IXP2400/2800 Programming Guide • Defining the architecture as well as the criteria of threads and microengines allocation • Excellent project demo • Notes • In this week, there is a demonstration of excellent project in T.W.T.C. Besides I have to define the architecture as well as the criteria of threads and microengines allocation. Gradually, I’ve felt it’s not easy to finish my thesis, however, I’ll do my best and try to be well-organized.

  10. Weekly Plan (2004/08/09) • Plan for last week • Reading IXP2400/2800 Programming Guide • Source code reading of Snort • Plan for this week • Reading IXP2400/2800 Programming Guide • Understating how packet inspection is done in Snort • Notes • Earlier this month, I’ve read several papers about implementation of IDS on network processors. But in vain, lots of details are still unknown. Although I’ve pointed out the most important parts of Snort, but how packet inspection is done is needed to find out in this week.

  11. Weekly Plan (2004/08/02) • Plan for last week • Thesis pre-study • Plan for this week • Reading IXP2400/2800 Programming Guide • Source code reading of Snort • Notes • As regards to thesis pre-study, I’ve decide to concentrate on IDS. Therefore, I can analyze the critical parts of Snort, say memory and computation intensive, and off-load them to the micro-engines. In long term, I’ll try to think an elegant way to allocate resource on IXP2400. Hope I can propose a workable proposal in the end of summer vocation.

  12. Weekly Plan (2004/07/26) (1/3) • Plan for last week • Test the functionality of freeswan and tc • Thesis pre-study • Plan for this week • Thesis pre-study Notes – two aspects • Research • Last week, I’ve discussed with my mentor, Yi-Neng, and talked about the parallelism of Aho-Chorasick on NPs. In this week, I must extend the idea derived from someone’s paper, and meanwhile answering the following questions.

  13. Weekly Plan (2004/07/26) (2/3) • Why not let threads themselves pick up the next unprocessed slots ? new threshold being processed unprocessed t0 t1 t2 t3 t8 t9 When (TL/R) <= A(B-T), buffer overflow occurs MEac MEac MEac MErx ….

  14. Pattern 2 packet 1 packet 2 packet 3 t0 t0 t1 t1 MEac TIMEA TIME B Weekly Plan (2004/07/26) (3/3) • How to inspect interleaving packets belong to different flows? • Development • Set up the development environment of IXP2400.

  15. Weekly Plan (2004/07/19) • Plan for this week • Test the functionality of freeswan and tc • Thesis pre-study Notes • A brand-new semester is coming up; I need to spend much time on thesis pre-study and practicing my English writing. In this week, I should find out the answer to the question about the parallelism of Aho-Corasick on network processors, and also improve my weakness on this domain.

  16. Weekly Plan (2004/06/21) • Plan for last week • Preparing for the final exam of formal language • sw/hw co-design final project • Network experiments • Plan for this week • Preparing for the presentation of thesis pre-study • sw/hw co-design final project • Network experiments Notes • I’m tied up in this week, because there are many things to do. Although it’s great that the summer vocation is beginning, I still need to finish the sw/hw co-design final project and pass the network experiment final exam. God bless me!

  17. Weekly Plan (2004/06/14) • Plan for last week • Preparing for the final exam of formal language • sw/hw co-design final project • Network experiments • Plan for this week • Preparing for the final exam of formal language • sw/hw co-design final project • Network experiments Notes • Soon, I’m going to be a graduate student in the second year. It implies I should concentrate on my thesis and plan well for the upcoming new semester. Although it’ll be a hard time to go through, it’s worthwhile. By the way, I wish the 2nd-year graduates will have a good time in their future career. Hey, Y.C, don’t be fool around, just keep your eyes on your target!!

  18. Weekly Plan (2004/06/07) • Plan for last week • Studying formal language • sw/hw co-design final project • Testing functionality of Wall • Plan for this week • Preparing for the final exam of formal language • sw/hw co-design final project • Network experiments Notes • There are many jobs coming up in this week. I have to gear up for the sw/hw co-design final project, which is a VPN gateway with a DES accelerator. So far, I have verified the DES IP core from opencores.org, and it still needs to be integrated with the MCU interface, though. Besides, I’ll concentrate myself on the final exam of formal language. Hope I can pass through this busy week!

  19. Weekly Plan (2004/05/31) • Plan for last week • Studying formal language • Learning VHDL • Plan for this week • Studying formal language • sw/hw co-design final project • Testing functionality of Wall Notes • Unfortunately, we failed in the contest of MOE. No matter what results were, we still need to gear up and prepare for the next time. In this week, I’ll make sure that the functionality of Wall can work, and studying is required as well.

  20. Weekly Plan (2004/05/24) • Plan for last week • Studying formal language • Learning VHDL • Plan for this week • Studying formal language • Learning VHDL Notes • Soon, A new week begins, I should gear up and prepare for the mid-term exam right off the bat.

  21. Weekly Plan (2004/05/17) • Plan for last week • Computer network experiments • Building Wall image • Studying formal language • Learning VHDL • Plan for this week • Studying formal language • Learning VHDL Notes • It’s time to finish the Wall work in this week, and I have to finish my embedded lab before Friday. In last week, I have learned much about the bus interface between ARM and FPGA, and I am supposed to keep on designing a pi accelerator. About the idea of thesis, I have talked to Yi-Neng and Kuo-Kuan, and they gave me some comments and some topics. As a result, I am getting understood what can I do about my thesis.

  22. Weekly Plan (2004/05/10) • Plan for last week • Computer network experiments • Embedded firmware hands-on – MPEG4 decoder • Building snapgear image • Studying formal language • Plan for this week • Computer network experiments • Building Wall image • Studying formal language • Learning VHDL Notes • Without exception, I intend to finish some school assignments this week, among which is a pi accelerator, and it seems a hard work. Therefore I need to study hard to increase my muscle in the domain, and meanwhile think about what can I do about my thesis. By the way, I am getting tired of studying everyday, perhaps I need to take a rest, say go traveling, and appreciate the beauty of nature, blah blah. Sometimes I always think I am not good enough, and very disappointed, hope I can pass through my bad mood in this week.

  23. Weekly Plan (2004/05/03) • Plan for last week • Doing computer network experiments • Embedded hands-on • Plan for this week • Computer network experiments • Embedded firmware hands-on – MPEG4 decoder • Building snapgear image • Studying formal language Notes • I have tried hard to help Szu-Hao to integrate perl, but unfortunately there are still many problems which I need to solve in this week. Besides, I’ll spend lots of time doing a lab of MPEG4 decoder on ecos. Hope I can pass through this busy week!

  24. Weekly Plan (2004/04/26) • Plan for last week • Doing computer network experiments • Learning VHDL • Plan for this week • Doing computer network experiments • Embedded hands-on Notes • There are many upcoming works to do, say network experiments, embedded team project, besides I need to build a workable image for the wall.

  25. Weekly Plan (2004/04/20) • Plan for last week • Building uclinux kernel and integrate filesystem for testing • Doing computer network experiments • Learning VHDL • Plan for this week • Doing computer network experiments • Learning VHDL Notes • In Last week, I have tried to install SnapGear, the embbeded linux distribution, and see what’s going on inside. I found SnapGear is really a nice distribution, which includes the functionalities we want, say VPN. Besides, it provides a user-friendly configurable menu, so that we can choose many alternatives without much effort. Hence, I think we can try it.

  26. Weekly Plan (2004/04/12) • Plan for last week • Preparing for the exam • Plan for this week • Building uclinux kernel and integrate filesystem for testing • Doing computer network experiments • Learning VHDL Notes • Finally, the midterm exam is over, and I have to spend more time on wall project and learning VHDL. My short-term goal is to learn how to design an ASIC, say a simple CRC ASIC, so that it can co-work with the ARM core and offload the overhead of network processing. Moreover, I’ll try to write a simple flash utility which is my hobby. Here, I would like to appreciate my dad that he supports the equipments.

  27. Weekly Plan (2004/04/05) • Plan for last week • Embedded hardware-and-software co-design hands on • Computer network experiments • Plan for this week • Preparing for the exam Notes • In this week, I’ll focus on formal language. Although I have tried to study hard, I think this course is not easy to get high grade. Any way, I’ll do as best as I can.

  28. Weekly Plan (2004/03/29) • Plan for last week • Surveying embedded Linux variants and preparing for the presentation • Learning VHDL • Book reading • Plan for this week • Embedded hardware-and-software co-design hands on • Computer network experiments Notes • In this week, there are many jobs to do. Although I have tried to build a uClinux distribution from the scratch successfully, I need to find a systematic way for us to develop the new wall. Besides, I’ll spend much time porting the uHAL used both in ARM boot monitor and Angel debug monitor, thus I can write a BSL module for the Lab 3 of the embedded hardware-and-software co-design course. By the way, the midterm exam is coming, and I should gear up for the exam.

  29. Weekly Plan (2004/03/22) • Plan for last week • Embedded hardware-and-software co-design hands on • Learning VHDL • Book reading • Plan for this week • Surveying embedded Linux variants and preparing for the presentation • Learning VHDL • Book reading Notes • Recently, I have been aware of the important of English conversation since I feel that I am poor in conversation when talking to a foreigner, besides writing skill is needed to be improved as well. In this week, my plan is as usual, as regards the wall plan, I have to spend much time preparing for the presentation coming up next week.

  30. Weekly Plan (2004/03/15) • Plan for last week • Trying to install Linux on ARM platform • Book reading • Plan for this week • Embedded hardware-and-software co-design hands on • Learning VHDL • Book reading Notes • In this semester, I take the course, called formal language, and I have found that it’s a good training for my logic reasoning. Also, it inspires me to understand the problem statements, and try to solve and prove them. As far as embedded course is concerned, it always interests me. From the lab of last homework, I have learned how to do profiling and code optimization on embedded software, and the lab in this week will be a piece of cake that asks us to write a timer interrupt handler in ARM assembly language. Hoping I will learn much from the experience!

  31. Weekly Plan (2004/03/08) • Plan for last week • Tracing the makefile of the new wall • Book reading • Plan for this week • Trying to install Linux on ARM platform • Book reading Notes • Since we have considered to migrate our new wall to the Linux operating system, I have to survey which variant of Linux is suitable for embedded system, of course, uClinux is the best choice. In this week, I plan to spend much time focusing on VHDL, which will be used as the HDL for hardware-and-software codesign, as a result, it’ll be helpful for my direction of theses.

  32. Weekly Plan (2004/03/01) • Plan for last week • Keeping on porting the new wall to ARM platform • Book reading • Plan for this week • Tracing the makefile of the new wall • Book reading Notes • Before starting porting applications to the new platform, there are some stuffs I need to know, say what parts of the kernel are patched, makefile rules and so on. Besides, I need to make sure the original modification can be applied to the new one, because there are many differences between NetBSD 1.5 and NetBSD 1.6. By the way, I think it is a big function, and hope I can finish the job as soon as I can.

  33. Weekly Plan (2004/02/23) • Plan for last week • Preparing for the paper presentation • Plan for this week • Keeping on porting the new wall to ARM platform • Book reading Notes • Finally, I have finished my presentation about the system research paper, although I didn’t perform very well, I’ll do my best in the next show time. In this semester, I’ll take the course, called “hardware and software co-desgin”, talking about how to design and program your own ASIC. From the courses, I think I can learn much and make my muscle stronger.

  34. Weekly Plan (2004/02/16) • Plan for last week • Building the development environment for Wall on ARM platform • Paper reading • Plan for this week • Preparing for the paper presentation • Notes • As far as porting is concerned, I have found a DBPXA250 emulator called Virtex, therefore, I can do anything on the platform as same as the real one. Soon, it’s the beginning of a new semester, and many courses are very interesting and challenging. Hope I can learn much from them.

  35. Weekly Plan (2004/02/09) • Plan for last week • Fixing antivirus and antispam functionality • Paper reading • Plan for this week • Building the development environment for Wall on ARM platform • Paper reading • Notes • Last week, we spent lots of time finding the problem of anti-spam, finally the problem was solved. However, the machine running the Wall has many hardware problems, say network interface timeout, and I suggest that we should buy new machines. • In this week, I have to try to run NetBSD on Intel DBPX250 platform, and think how to port the original source tree to the new platform without many changes.

  36. Weekly Plan (2004/02/02) • Plan for last week • Helping Chih-Wei to solve the problem of spamassasin • Setup the environment of computer network mini-porject • Preparing for the final exam of computer network • Plan for this week • Fixing antivirus and antispam functionality • Paper reading • Notes • In this week, I have to solve the problem of antispam, since there are some subtle problems when doing package integration. As far as thesis issues are concerned, ASIC design is a work of art for me, and I need to make an effort to learn as much as I can.

  37. Weekly Plan (2004/01/12) • Plan for last week • Helping Chih-Wei to solve the problem of spamassasin • Setup the environment of computer network mini-porject • Preparing for the final exam of computer network • Plan for this week • Computer network mini-porject testing • Notes • Woo hoo! Finally, the semester has ended, and it’s worth to take a rest. In this week, I have to finish the testing of computer network mini-project, besides a system research paper reading is required. Hope I can find my topic of thesis as soon as possible.

  38. Weekly Plan (2004/01/06) • Plan for last week • Preparing for the final exam • Preparing for the computer network mini-project • Plan for this week • Helping Chih-Wei to solve the problem of spamassasin • Setup the environment of computer network mini-porject • Preparing for the final exam of computer network • Notes • Last week, we have encountered some problems about the anti-spam functionality caused by spamassasin, which evaluates every mail as a result of the score, zero, and we are still trying hard to solve the problem. In this week, there are many reports and exams, and hope I can pass in all courses.

More Related