> For the complete documentation index, see [llms.txt](https://nju-ics.gitbook.io/ics2017-programming-assignment/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://nju-ics.gitbook.io/ics2017-programming-assignment/pa0/0.7.md). # Acquiring Source Code for PAs ## Acquiring Source Code for PAs ### Getting Source Code Go back to the home directory by ``` cd ~ ``` Usually, all works unrelated to system should be performed under the home directory. Other directories under the root of file system (`/`) are related to system. Therefore, do NOT finish your PAs and Labs under these directories by `sudo`. Now acquire source code for PA by the following command: ``` git clone -b 2017 https://github.com/NJU-ProjectN/ics-pa.git ics2017 ``` A directory called `ics2017` will be created. This is the project directory for PAs. Details will be explained in PA1. Issue the following commands to perform `git` configuration: ``` git config --global user.name "161220000-Zhang San" # your student ID and name git config --global user.email "zhangsan@foo.com" # your email git config --global core.editor vim # your favorite editor git config --global color.ui true ``` You should configure `git` with your student ID, name, and email. Before continuing, please read [this](/ics2017-programming-assignment/blank/git.md) `git` tutorial to learn some basics of `git`. Enter the project directory `ics2017`, then run ``` git branch -m master bash init.sh ``` to initialize all the subprojects. This script will pull 4 subprojects from github. We will explain them later. Besides, the script will also add some environment variables into the bash configuration file `~/.bashrc`. These variables are defined by absolute path to support the compilation of the subprojects. Therefore, DO NOT move your project to another directory once the initialization finishes, else these variables will become invalid. Particularly, if you use shell other than `bash`, please set these variables in the corresponding configuration file manually. ### Git usage We will use the `branch` feature of `git` to manage the process of development. A branch is an ordered list of commits, where a commit refers to some modifications in the project. You can list all branches by ``` git branch ``` You will see there is only one branch called "master" now. ``` * master ``` To create a new branch, use `git checkout` command: ``` git checkout -b pa0 ``` This command will create a branch called `pa0`, and check out to it. Now list all branches again, and you will see we are now at branch `pa0`: ``` master * pa0 ``` From now on, all modifications of files in the project will be recorded in the branch `pa0`. Now have a try! Modify the `STU_ID` variable in `nemu/Makefile.git`: ``` STU_ID=161220000 # your student ID ``` Run ``` git status ``` to see those files modified from the last commit: ``` On branch pa0 Changes not staged for commit: (use "git add ..." to update what will be committed) (use "git checkout -- ..." to discard changes in working directory) modified: nemu/Makefile.git no changes added to commit (use "git add" and/or "git commit -a") ``` Run ``` git diff ``` to list modifications from the last commit: ```diff diff --git a/nemu/Makefile.git b/nemu/Makefile.git index c9b1708..b7b2e02 100644 --- a/nemu/Makefile.git +++ b/nemu/Makefile.git @@ -1,4 +1,4 @@ -STU_ID = 161220000 +STU_ID = 161221234 # DO NOT modify the following code!!! ``` You should see the `STU_ID` is modified. Now add the changes to commit by `git add`, and issue `git commit`: ``` git add . git commit ``` The `git commit` command will call the text editor. Type `modified my STU_ID` in the first line, and keep the remaining contents unchanged. Save and exit the editor, and this finishes a commit. Now you should see a log labeled with your student ID and name by ``` git log ``` Now switch back to the `master` branch by ``` git checkout master ``` Open `nemu/Makefile.git`, and you will find that `STU_ID` is still unchanged! By issuing `git log`, you will find that the commit log you just created has disappeared! Don't worry! This is a feature of branches in `git`. Modifications in different branches are isolated, which means modifying files in one branch will not affect other branches. Switch back to `pa0` branch by ``` git checkout pa0 ``` You will find that everything comes back! At the beginning of PA1, you will merge all changes in branch `pa0` into `master`. The workflow above shows how you will use branch in PAs: * before starting a new PA, new a branch `pa?` and check out to it * coding in the branch `pa?` (this will introduce lot of modifications) * after finish the PA, merge the branch `pa?` into `master`, and check out back to `master` ### Compiling and Running NEMU Now enter `nemu/` directory, and compile the project by `make`: ``` make ``` If nothing goes wrong, NEMU will be compiled successfully. To perform a fresh compilation, type ``` make clean ``` to remove the old compilation result, then `make` again. To run NEMU, type ``` make run ``` However, you will see an error message: ``` nemu: nemu/src/cpu/reg.c:21: reg_test: Assertion `(cpu.gpr[check_reg_index(i)]._16) == (sample[i] & 0xffff)' failed. ``` This message tells you that the program has triggered an assertion fail at line 21 of the file `nemu/src/cpu/reg.c`. If you do not know what is assertion, blame the 程序设计基础 course. If you go to see the line 21 of `nemu/src/cpu/reg.c`, you will discover the failure is in a test function. This failure is expected, because you have not implemented the register structure correctly. Just ignore it now, and you will fix it in PA1. To debug NEMU with gdb, type ``` make gdb ``` ### Development Tracing Once the compilation succeeds, the change of source code will be traced by `git`. Type ``` git log ``` If you see something like ``` commit 4072d39e5b6c6b6837077f2d673cb0b5014e6ef9 Author: tracer-ics2017 Date: Sun Jul 26 14:30:31 2017 +0800 > compile NEMU 161220000 user Linux debian 3.16.0-4-686-pae #1 SMP Debian 3.16.7-3 i686 GNU/Linux 14:30:31 up 3:44, 2 users, load average: 0.28, 0.09, 0.07 3860572d5cc66412bf85332837c381c5c8c1009f ``` this means the change is traced successfully. If you see the following message while executing make, this means the tracing fails. ``` fatal: Unable to create '/home/user/ics2017/.git/index.lock': File exists. If no other git process is currently running, this probably means a git process crashed in this repository earlier. Make sure no other git process is running and remove the file manually to continue. ``` Try to clean the compilation result and compile again: ``` make clean make ``` If the error message above always appears, please contact us as soon as possible. ### Local Commit Although the development tracing system will trace the change of your code after every successful compilation, the trace record is not suitable for your development. This is because the code is still buggy at most of the time. Also, it is not easy for you to identify those bug-free traces. Therefore, you should trace your bug-free code manually. When you want to commit the change, type ``` git add . git commit --allow-empty ``` The `--allow-empty` option is necessary, because usually the change is already committed by development tracing system. Without this option, `git` will reject no-change commits. If the commit succeeds, you can see a log labeled with your student ID and name by ``` git log ``` To filter out the commit logs corresponding to your manual commit, use `--author` option with `git log`. For details of how to use this option, RTFM. ### Submission Finally, you should submit your project to the submission website. To submit PA0, put your report file (ONLY `.pdf` file is accepted) under the project directory. ``` ics2017 ├── 161220000.pdf # put your report file here ├── init.sh ├── Makefile ├── nanos-lite ├── navy-apps ├── nemu └── nexus-am ``` Then go back to the project directory, issue ``` make submit ``` This command does 3 things: 1. Cleanup unnecessary files for submission 1. Cleanup unnecessary files in git 1. Create an archive containing the source code and your report. The archive is located in the father directory of the project directory, and it is named by your student ID set in Makefile. If nothing goes wrong, transfer the archive to your host. Open the archive to double check whether everything is fine. And you can manually submit this archive to the submission website. ## RTFSC and Enjoy If you are new to GNU/Linux and finish this tutorial by yourself, congratulations! You have learn a lot! The most important, you have learn searching the Internet and RTFM for using new tools and trouble-shooting. With these skills, you can solve lots of troubles by yourself during PAs, as well as in the future. In PA1, the first thing you will do is to [RTFSC](http://i.linuxtoy.org/docs/guide/ch48s06.html). If you have troubles during reading the source code, go to RTFM: * If you can not find the definition of a function, it is probably a library function. Read `man` for more information about that function. * If you can not understand the code related to hardware details, refer to the i386 manual. By the way, you will use C language for programming in all PAs. [Here](http://docs.huihoo.com/c/linux-c-programming) is an excellent tutorial about C language. It contains not only C language (such as how to use `printf()` and `scanf()`), but also other elements in a computer system (data structure, computer architecture, assembly language, linking, operating system, network...). It covers most parts of this course. You are strongly recommended to read this tutorial. Finally, enjoy the journey of PAs, and you will find hardware is not mysterious, so does the computer system! But remember: * The machine is always right. * Every line of untested code is always wrong. * RTFM.