ok time for bed

run now do you get fegfaults with BB?

pat look at lines 276-285 does the math seem right?

minor changes

oInfo.cwd is now an array no longer a pointer..

hmm..

chg: modified UbixFS::vfs_format() to properly init the device ...

chg: modified UbixFS::vfs_init() to verify that it's a ubixfs partition
add: added BAT blockRun in the superBlock

Make endtask loop forever if _current->id == pid

Fixed sched to work with gcc 3.3

Fixed a few problems

Fix ip#

Fix DEAD problem

Fix race conditions in net thread and scheduler

here you go irq 10 io 240 for your ne2k nic

CHRIS: Don't mess with scheduler structures outside the schedulergvim endtask.c Finally fixed the race condition.

Fix race condition (multiple unlock)

Add locking get GetFreePage

Fix order in adjustCowCounter to prevent potential race condition

memset newly allocated tasks

Oops why did i disable irq 1 irq 0 is the timer...

ok

ok enabled networking

Make a FD_DEVICE entry in Makefile.inc

we can now use all versions of gcc thanks to A.P.Williams

Removed kernel/file.c all in the vfs now

fork: cleaned up comments in fork and made fork_copyProcess static

Make spinlock lock function sched_yield. Add spinlockLock_scheduler for use in smp-scheduler (avoids recursive call to sched).

Fixed file syscalls so fdisk does not kernel panic

add: device.h, types.h. Temporarily moved custom ubix typedefs to types.h ...

(<ubixos/types.h> supercedes this file when this code is included into
     the kernel)
chg: changed dev_t to device_interface so as to not conflict with bsd/linux
     definitions
chg: fixed up compiler warnings in btree.cpp

We now initialize all needed daemons