There are a few spots where we call memory_notifier. This doesn't need to be done under lock since memory_notify is just a blocking_notifier_call_chain with it's own locking mechanism.
This RFC is a follow on to the one I submitted earlier. (move register_memory_resource out of the lock_memory_hotplug, commit ac13c46) Most of our time is being spend under the memory hotplug lock in particular online_pages() so it makes sense to move out everything that can be easily moved out. However perf pointed me to a spot to work on, setup_zone_migrate_reserve. In fact most of the time is spent there. Since that is going to require more reading and time I will start by whittling out some easy pieces. cc: Andrew Morton <[email protected]> cc: Tang Chen <[email protected]> cc: Wen Congyang <[email protected]> cc: Toshi Kani <[email protected]> cc: Yasuaki Ishimatsu <[email protected]> cc: Xishi Qiu <[email protected]> cc: Cody P Schafer <[email protected]> cc: "Rafael J. Wysocki" <[email protected]> cc: David Rientjes <[email protected]> cc: Jiang Liu <[email protected]> Cc: Hedi Berriche <[email protected]> Cc: Mike Travis <[email protected]> [email protected] (open list:MEMORY MANAGEMENT) [email protected] (open list) --- mm/memory_hotplug.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index 62a0cd1..a3cbd14 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -985,12 +985,12 @@ int __ref online_pages(unsigned long pfn, unsigned long nr_pages, int online_typ if (need_zonelists_rebuild) zone_pcp_reset(zone); mutex_unlock(&zonelists_mutex); + unlock_memory_hotplug(); printk(KERN_DEBUG "online_pages [mem %#010llx-%#010llx] failed\n", (unsigned long long) pfn << PAGE_SHIFT, (((unsigned long long) pfn + nr_pages) << PAGE_SHIFT) - 1); memory_notify(MEM_CANCEL_ONLINE, &arg); - unlock_memory_hotplug(); return ret; } @@ -1016,9 +1016,10 @@ int __ref online_pages(unsigned long pfn, unsigned long nr_pages, int online_typ writeback_set_ratelimit(); + unlock_memory_hotplug(); + if (onlined_pages) memory_notify(MEM_ONLINE, &arg); - unlock_memory_hotplug(); return 0; } @@ -1601,8 +1602,8 @@ repeat: vm_total_pages = nr_free_pagecache_pages(); writeback_set_ratelimit(); - memory_notify(MEM_OFFLINE, &arg); unlock_memory_hotplug(); + memory_notify(MEM_OFFLINE, &arg); return 0; failed_removal: -- 1.8.2.1 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
