My changes have a race with re-used bufs and the code doesn't seem to get shorter
Keep the changes that fixed ip->off race

This reverts commit 3a5fa7ed.

Conflicts:

	defs.h
	file.c
	file.h

Remove I_BUSY, B_BUSY, and intrans defs and usages
One spinlock per buf to avoid ugly loop in bget
fix race in filewrite (don't update f->off after releasing lock)

usertest for big write()s
push begin_trans/commit_trans down into syscalls

pdf has very good page breaks now.
would be a good copy for fall 2009.

group locks into structs they protect.
few naming nits.

Remove struct uinode.
Remove type arg to mknod (assume T_DEV).

PDF at http://am.lcs.mit.edu/~rsc/xv6.pdf

Various changes made while offline.

 + bwrite sector argument is redundant; use b->sector.
 + reformatting of files for nicer PDF page breaks
 + distinguish between locked, unlocked inodes in type signatures
 + change FD_FILE to FD_INODE
 + move userinit (nee proc0init) to proc.c
 + move ROOTDEV to param.h
 + always parenthesize sizeof argument

renumber system calls (run gmake clean!)
skeleton for dup system call

don't hold fd table lock across idecref() (latter does block i/o)
idecref calls iput() in case last ref -> freeing inode
dir size is 512 * # blocks, so readi/writei &c work
unlink deletes dirent even if ip->nlink > 0

primitive ls

so fast interrupts overflow the kernel stack
fix: cli() before lapic_eoi()

convert userfs to use printf
bfree
ifree
writei
start on unlink

checkpoint: write(fd,"hello\n",6) where fd is a console dev almost works

give cpu1 a TSS and gdt for when it enters scheduler()
and a pseudo proc[] entry for each cpu
cpu0 waits for each other cpu to start up
read() for files

Rename fd_reference to more suggestive fd_incref.
(Fd_reference sounds like it might just return the ref count.)