kernel - Reduce atomic ops in switch code
commite1bff962bff1ed6f5d4d19d3c6867d58fda7cfee
authorMatthew Dillon <dillon@apollo.backplane.com>
Tue, 26 Jul 2016 19:56:31 +0000 (26 12:56 -0700)
committerMatthew Dillon <dillon@apollo.backplane.com>
Wed, 27 Jul 2016 00:21:20 +0000 (26 17:21 -0700)
tree80ac1c58acc3a4e8ad150b6005c7ced213a6c3f0
parentbc2be5ee4c521f03988d6d8d155c2fb41a1a8150
kernel - Reduce atomic ops in switch code

* Instead of using four atomic 'and' ops and four atomic 'or' ops, use
  one atomic 'and' and one atomic 'or' when adjusting the pmap->pm_active.

* Store the array index and simplified cpu mask in the globaldata structure
  for the above operation.
sys/kern/init_main.c
sys/kern/lwkt_thread.c
sys/platform/pc64/x86_64/genassym.c
sys/platform/pc64/x86_64/global.s
sys/platform/pc64/x86_64/swtch.s
sys/sys/globaldata.h