Date: Thu, 27 Oct 2011 17:37:00 GMT From: John Baldwin <jhb@FreeBSD.org> To: Perforce Change Reviews <perforce@freebsd.org> Subject: PERFORCE change 200834 for review Message-ID: <201110271737.p9RHb0IG054758@skunkworks.freebsd.org>
next in thread | raw e-mail | index | archive | help
http://p4web.freebsd.org/@@200834?ac=10 Change 200834 by jhb@jhb_jhbbsd on 2011/10/27 17:36:23 Optimize vm page locking ala vm_unhold_pages(). Requested by: kib Affected files ... .. //depot/projects/fadvise/sys/vm/vm_object.c#4 edit Differences ... ==== //depot/projects/fadvise/sys/vm/vm_object.c#4 (text+ko) ==== @@ -1881,6 +1881,7 @@ void vm_object_page_cache(vm_object_t object, vm_pindex_t start, vm_pindex_t end) { + struct mtx *mtx, *new_mtx; vm_page_t p, next; VM_OBJECT_LOCK_ASSERT(object, MA_OWNED); @@ -1895,13 +1896,24 @@ * Here, the variable "p" is either (1) the page with the least pindex * greater than or equal to the parameter "start" or (2) NULL. */ + mtx = NULL; for (; p != NULL && (p->pindex < end || end == 0); p = next) { next = TAILQ_NEXT(p, listq); - vm_page_lock(p); + /* + * Avoid releasing and reacquiring the same page lock. + */ + new_mtx = vm_page_lockptr(p); + if (mtx != new_mtx) { + if (mtx != NULL) + mtx_unlock(mtx); + mtx = new_mtx; + mtx_lock(mtx); + } vm_page_try_to_cache(p); - vm_page_unlock(p); } + if (mtx != NULL) + mtx_unlock(mtx); } /*
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201110271737.p9RHb0IG054758>