Differences between revisions 15 and 16
Revision 15 as of 2014-03-29 15:53:05
Size: 3237
Editor: ?Sphinx Jiang
Revision 16 as of 2014-03-29 15:55:28
Size: 2805
Editor: ?Sphinx Jiang
Deletions are marked like this. Additions are marked like this.
Line 25: Line 25:

---- /!\ '''Edit conflict - other version:''' ----
   1. package patches
   *All the deliverables must be accepted by official repository.

---- /!\ '''Edit conflict - your version:''' ----
Line 36: Line 30:

---- /!\ '''End of edit conflict''' ----
Line 43: Line 35:

---- /!\ '''Edit conflict - other version:''' ----
Line 46: Line 36:

---- /!\ '''Edit conflict - your version:''' ----
      *Write mid-term work summary.

---- /!\ '''End of edit conflict''' ----

Student Application: Sphinx Jiang

  • Name:Sphinx Jiang

  • Contact/Email: yishanj13@gmail.com

  • Background:

    • I am studying software engineering as a postgraduate At Beijing Jiaotong University.Before that I was major in Electrical Engineering at the same university. Debian user for years. The reasons why I suit the project:
    • I am familiar with Debian packaging, Multi-Arch.
    • I have some experience of cross toolchains.
    • I am familiar with C,Shell,know a little assembly language.
  • Project title:

    • Improve Debian on mips64el
  • Project details:

    1. Port debian kernel package to support mips64el for loongson 2/3 and other possible hardware
    2. Port debian-installer to mips64el and build images for loongson 2/3 and other possible hardware
    3. Fix some packages that failed to build on mips64el, now 1000+ success rate reach 94%, from packages as dependence, reach the rate to 97% at least.
  • Synopsis:

    • Make Debian run on mips64el device
  • Benefits to Debian:

    • This project will make Debian run on mips64el architecture advice, such as loongson 2/3.
  • Deliverables:

    1. kernel package
    2. debian-installer
    3. package patches
    • For the question from mentor:
      • The most challenging technical problem is how to rise the package success rate from 94% to 97%. The reason is first, for numbers there are hundreds packages, and none of them are as simple as the packages that are successfully compiled. I will come to solve them one by one, to find out what is the real reason of fail. That may be code not perfect, package compiling problem, or even hardware problem. Each of the reasons will be taken into consideration, I will kill some important failed packages in this way.
  • Project schedule:

    1. May 1 - May 15: Read relevant documentations, basic architecture of mips64el.
    2. May 8 - June 8:Port the kernel package to mips64el.
    3. June 10 - July 15:
      • Port Installer to mips64el.
      • Write mid-term work summary.
    4. July 17 - Aug 20:
      • Rise the success rate of packages on mips64el from 94% to 97% at least.
      • Compare performance with different compile optimize option.
    5. Aug 16: - Aug 22:
      • Deal with any issue that aren't finished yet.
      • Write final summary.
  • Exams and other commitments:

    • I won't take exam, but may need a week to prepare for onboard delay at July.
  • Why Debian?:

    • Debian is my daily used operation system also one of the largest open source community in the world, any improvement of it will affect on huge group of users. Especially the port of some new device.
  • Are you applying for other projects in SoC?:

    • No