greg/xv6
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
xv6/BUGS

51 lines
1.1 KiB
Text
Raw Normal View History

accumulate bugs
2006-09-17 19:19:01 +00:00
proc.c:
as a consequence of the implementation of proc_kill,
any loop calling sleep should check for p->killed
and be able to break out with an error return.
notes
2007-08-08 10:32:32 +00:00
it is better if you check *before* sleep.
accumulate bugs
2006-09-17 19:19:01 +00:00
can swap procdump up after proc_kill
and then have proc_exit and proc_wait on same sheet
a few other nits
2006-09-17 22:01:12 +00:00
sched -> switch2scheduler? or just switch?
a few notes
2006-10-13 19:55:07 +00:00
factor out switching and scheduling code from process code
Various cleanup: - Got rid of dummy proc[0]. Now proc[0] is init. - Added initcode.S to exec /init, so that /init is just a regular binary. - Moved exec out of sysfile to exec.c - Moved code dealing with fs guts (like struct inode) from sysfile.c to fs.c. Code dealing with system call arguments stays in sysfile.c - Refactored directory routines in fs.c; should be simpler. - Changed iget to return *unlocked* inode structure. This solves the lookup-then-use race in namei without introducing deadlocks. It also enabled getting rid of the dummy proc[0].
2007-08-21 19:22:08 +00:00
shuffle for formatting
accumulate bugs
2006-09-17 19:19:01 +00:00
Various cleanup: - Got rid of dummy proc[0]. Now proc[0] is init. - Added initcode.S to exec /init, so that /init is just a regular binary. - Moved exec out of sysfile to exec.c - Moved code dealing with fs guts (like struct inode) from sysfile.c to fs.c. Code dealing with system call arguments stays in sysfile.c - Refactored directory routines in fs.c; should be simpler. - Changed iget to return *unlocked* inode structure. This solves the lookup-then-use race in namei without introducing deadlocks. It also enabled getting rid of the dummy proc[0].
2007-08-21 19:22:08 +00:00
syscall.c:
cannot convince runoff1 to split the extern lists to fill previous page completely.
a few notes
2006-10-13 19:55:07 +00:00
Various cleanup: - Got rid of dummy proc[0]. Now proc[0] is init. - Added initcode.S to exec /init, so that /init is just a regular binary. - Moved exec out of sysfile to exec.c - Moved code dealing with fs guts (like struct inode) from sysfile.c to fs.c. Code dealing with system call arguments stays in sysfile.c - Refactored directory routines in fs.c; should be simpler. - Changed iget to return *unlocked* inode structure. This solves the lookup-then-use race in namei without introducing deadlocks. It also enabled getting rid of the dummy proc[0].
2007-08-21 19:22:08 +00:00
fs.c: split all name operations off in name.c? (starting with namei but
record tim's bug and patch
2006-10-18 14:58:27 +00:00
wdir keep in fs.c)
Various cleanup: - Got rid of dummy proc[0]. Now proc[0] is init. - Added initcode.S to exec /init, so that /init is just a regular binary. - Moved exec out of sysfile to exec.c - Moved code dealing with fs guts (like struct inode) from sysfile.c to fs.c. Code dealing with system call arguments stays in sysfile.c - Refactored directory routines in fs.c; should be simpler. - Changed iget to return *unlocked* inode structure. This solves the lookup-then-use race in namei without introducing deadlocks. It also enabled getting rid of the dummy proc[0].
2007-08-21 19:22:08 +00:00
locking?
shuffle for formatting
notes
2007-08-08 10:32:32 +00:00
pipe.c:
more comments?
comment how functions get called?
sysfile.c:
is the sys_exec picture upside down?
can sys_open and sys_exec be simplified any?
general:
sizeof parens?
Various cleanup: - Got rid of dummy proc[0]. Now proc[0] is init. - Added initcode.S to exec /init, so that /init is just a regular binary. - Moved exec out of sysfile to exec.c - Moved code dealing with fs guts (like struct inode) from sysfile.c to fs.c. Code dealing with system call arguments stays in sysfile.c - Refactored directory routines in fs.c; should be simpler. - Changed iget to return *unlocked* inode structure. This solves the lookup-then-use race in namei without introducing deadlocks. It also enabled getting rid of the dummy proc[0].
2007-08-21 19:22:08 +00:00
bio.c:
decide odd or even
bwrite doesn't need a second argument
file.c:
move fileincref onto page 1?
add note
2007-08-10 17:19:15 +00:00
L=$HOME/mit/l
(for i in *.c; do xoc -x xgnu -x ./nodecleq.zeta --typesonly $i; done) 2>&1 | grep warning
Various cleanup: - Got rid of dummy proc[0]. Now proc[0] is init. - Added initcode.S to exec /init, so that /init is just a regular binary. - Moved exec out of sysfile to exec.c - Moved code dealing with fs guts (like struct inode) from sysfile.c to fs.c. Code dealing with system call arguments stays in sysfile.c - Refactored directory routines in fs.c; should be simpler. - Changed iget to return *unlocked* inode structure. This solves the lookup-then-use race in namei without introducing deadlocks. It also enabled getting rid of the dummy proc[0].
2007-08-21 19:22:08 +00:00
saw random sharedfd failure.
why does fdalloc consume reference?
why mkdir and create?
Reference in a new issue Copy permalink