diff options
author | kmacy <kmacy@FreeBSD.org> | 2010-04-30 00:46:43 +0000 |
---|---|---|
committer | kmacy <kmacy@FreeBSD.org> | 2010-04-30 00:46:43 +0000 |
commit | 1dc1263413262d13f45f42d89c554d0ddc167ca2 (patch) | |
tree | 250705d42bcb9b364f4fd2233c38faaadadf7ae0 /UPDATING | |
parent | a43425e8835983e7c93a15a2d6a42fdb292c1676 (diff) | |
download | FreeBSD-src-1dc1263413262d13f45f42d89c554d0ddc167ca2.zip FreeBSD-src-1dc1263413262d13f45f42d89c554d0ddc167ca2.tar.gz |
On Alan's advice, rather than do a wholesale conversion on a single
architecture from page queue lock to a hashed array of page locks
(based on a patch by Jeff Roberson), I've implemented page lock
support in the MI code and have only moved vm_page's hold_count
out from under page queue mutex to page lock. This changes
pmap_extract_and_hold on all pmaps.
Supported by: Bitgravity Inc.
Discussed with: alc, jeffr, and kib
Diffstat (limited to 'UPDATING')
-rw-r--r-- | UPDATING | 8 |
1 files changed, 8 insertions, 0 deletions
@@ -22,6 +22,14 @@ NOTE TO PEOPLE WHO THINK THAT FreeBSD 9.x IS SLOW: machines to maximize performance. (To disable malloc debugging, run ln -s aj /etc/malloc.conf.) + +20100429: + 'vm_page's are now hashed by physical address to an array of mutexes. + Currently this is only used to serialize access to hold_count. Over + time the page queue mutex will be peeled away. This changes the size + of pmap on every architecture. And requires all callers of vm_page_hold + and vm_page_unhold to be updated. + 20100402: WITH_CTF can now be specified in src.conf (not recommended, there are some problems with static executables), make.conf (would also |