No longer disables irqs when mucking with pcpui
authorBarret Rhoden <brho@cs.berkeley.edu>
Mon, 15 Oct 2012 21:16:51 +0000 (14:16 -0700)
committerBarret Rhoden <brho@cs.berkeley.edu>
Tue, 16 Oct 2012 21:42:20 +0000 (14:42 -0700)
commit83b6bf1d3dd603b1d2a59c0dbc3c3d5b06cd7059
treee8ee648f3c2765b131941f7b2ac1347bc060bb1f
parent7f88edba3ad525a138819c45875cdef17bd5169e
No longer disables irqs when mucking with pcpui

For things like cur_proc, owning_{proc,vcoreid}, and cur_tf, we don't
disable interrupts when mucking with them in proc code.  We used to need
this when proc mgmt KMSGs were immediate.  So long as we continue to not
muck with that state from IRQ handlers or IMMED KMSGs, we're okay.
kern/arch/i686/trap.c
kern/src/process.c