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