MCP ksched doesn't need to hold the lock forever
authorBarret Rhoden <brho@cs.berkeley.edu>
Wed, 9 May 2012 23:03:15 +0000 (16:03 -0700)
committerBarret Rhoden <brho@cs.berkeley.edu>
Wed, 5 Sep 2012 21:43:57 +0000 (14:43 -0700)
commit5909cdd49a760dadd60758f3b2247a4e98b59c2f
tree4e7f80058a64fbcd7182414061883d58f01403f5
parentc8c5b0f0d1ee82e12510510c3e1d3893892a407e
MCP ksched doesn't need to hold the lock forever

__core_request() can now let go of the ksched lock.  We only have that
lock held in __run_mcp_ksched() when we need to muck with things the
lock explicitly protects: proc list integrity and membership in this
case.
kern/src/schedule.c