3 * Test pintos with GSX.
5 * Need targets for checking single tests or single subdirs
7 - `make check' should say how to get more information
9 - `make recheck', `make clean-check'
11 * Need to put time limits on tests.
13 * Add a way for students to prints messages ignored by the autograder.
15 - Ignore text between /* and */.
17 - Allow for a comment on `exit' lines?
19 * Improve automatic interpretation of exception messages.
21 * debug_backtrace() in kernel should only dereference valid pointers.
23 * We need better and more example programs.
25 - Need an mmap example program as a replacement for the crappy mmap FAQ
28 - How about `diff' and `cmp' programs?
32 - mlfqs tests suck. They aren't even correct, e.g. the amarv
33 submission from win0405 is graded incorrectly.
37 - Don't emphasize that stuff needs to be copied from user space to
38 kernel space. Instead, emphasize validation and suggest that
39 copying is a common solution and that it will be necessary in
40 project 3 and in real OSes. Also revise the grading criteria to
43 - Mark read-only pages as actually read-only in the page table. Or,
44 since this was consistently rated as the easiest project by the
45 students, require them to do it.
47 - Don't provide per-process pagedir implementation but only
48 single-process implementation and require students to implement
49 the separation? This project was rated as the easiest after all.
50 Alternately we could just remove the synchronization on pid
51 selection and check that students fix it.
55 - Discuss the perils of mixing dirty bits between kernel and user virtual
60 - Update grading/vm to reflect new mmap, munmap forms.
64 - Increase maximum disk size from 8 MB to something that actually
65 requires doubly indirect nodes. There is a negative pressure here
66 from the bitmap object--perhaps we need a specialized bitmap that
67 doesn't have to be all in-memory at once.
69 Alternatively, shrink the inode size.
71 - Add option to disable buffer cache.
75 - Need a better way to measure performance improvement of buffer
76 cache. Some students reported that their system was slower with
77 cache--likely, Bochs doesn't simulate a disk with a realistic
80 - Clarify effect of remove(cwd).
84 - Finish writing tour.
86 - Add "Digging Deeper" sections that describe the nitty-gritty x86
87 details for the benefit of those interested.
89 - Add explanations of what "real" OSes do to give students some
92 - GNU make 3.80 or later is required.
98 . Low-level x86 stuff, like paged page tables.
100 . Specifics on how to implement sbrk, malloc.
104 . opendir/readdir/closedir
106 . everything needed for getcwd()
110 - Release some of them.
112 - The threads, userprog, vm test source files could use
113 factorization and cleanup along the lines of fslib in the filesys
116 - The p1-4.c testcase needs significant tuning. Currently it takes
117 too long (especially when SHOW_PROGRESS is turned on) and doesn't
118 show significant improvement.
122 - Need an optimization barrier and an explanation of it in the
125 - Need to check the wait system call more thoroughly.