From 95a7d76897c1e7243d4137037c66d15cbf2cce76 Mon Sep 17 00:00:00 2001 From: Konrad Rzeszutek Wilk Date: Wed, 31 Oct 2012 12:38:31 -0400 Subject: xen/mmu: Use Xen specific TLB flush instead of the generic one. As Mukesh explained it, the MMUEXT_TLB_FLUSH_ALL allows the hypervisor to do a TLB flush on all active vCPUs. If instead we were using the generic one (which ends up being xen_flush_tlb) we end up making the MMUEXT_TLB_FLUSH_LOCAL hypercall. But before we make that hypercall the kernel will IPI all of the vCPUs (even those that were asleep from the hypervisor perspective). The end result is that we needlessly wake them up and do a TLB flush when we can just let the hypervisor do it correctly. This patch gives around 50% speed improvement when migrating idle guest's from one host to another. Oracle-bug: 14630170 CC: stable@vger.kernel.org Tested-by: Jingjie Jiang Suggested-by: Mukesh Rathor Signed-off-by: Konrad Rzeszutek Wilk --- include/trace/events/xen.h | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'include/trace') diff --git a/include/trace/events/xen.h b/include/trace/events/xen.h index 15ba03bdd7c6..d06b6da5c1e3 100644 --- a/include/trace/events/xen.h +++ b/include/trace/events/xen.h @@ -377,6 +377,14 @@ DECLARE_EVENT_CLASS(xen_mmu_pgd, DEFINE_XEN_MMU_PGD_EVENT(xen_mmu_pgd_pin); DEFINE_XEN_MMU_PGD_EVENT(xen_mmu_pgd_unpin); +TRACE_EVENT(xen_mmu_flush_tlb_all, + TP_PROTO(int x), + TP_ARGS(x), + TP_STRUCT__entry(__array(char, x, 0)), + TP_fast_assign((void)x), + TP_printk("%s", "") + ); + TRACE_EVENT(xen_mmu_flush_tlb, TP_PROTO(int x), TP_ARGS(x), -- cgit v1.2.2 From 82b212f40059bffd6808c07266a942d444d5558a Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Mon, 26 Nov 2012 16:29:45 -0800 Subject: Revert "mm: remove __GFP_NO_KSWAPD" With "mm: vmscan: scale number of pages reclaimed by reclaim/compaction based on failures" reverted, Zdenek Kabelac reported the following Hmm, so it's just took longer to hit the problem and observe kswapd0 spinning on my CPU again - it's not as endless like before - but still it easily eats minutes - it helps to turn off Firefox or TB (memory hungry apps) so kswapd0 stops soon - and restart those apps again. (And I still have like >1GB of cached memory) kswapd0 R running task 0 30 2 0x00000000 Call Trace: preempt_schedule+0x42/0x60 _raw_spin_unlock+0x55/0x60 put_super+0x31/0x40 drop_super+0x22/0x30 prune_super+0x149/0x1b0 shrink_slab+0xba/0x510 The sysrq+m indicates the system has no swap so it'll never reclaim anonymous pages as part of reclaim/compaction. That is one part of the problem but not the root cause as file-backed pages could also be reclaimed. The likely underlying problem is that kswapd is woken up or kept awake for each THP allocation request in the page allocator slow path. If compaction fails for the requesting process then compaction will be deferred for a time and direct reclaim is avoided. However, if there are a storm of THP requests that are simply rejected, it will still be the the case that kswapd is awake for a prolonged period of time as pgdat->kswapd_max_order is updated each time. This is noticed by the main kswapd() loop and it will not call kswapd_try_to_sleep(). Instead it will loopp, shrinking a small number of pages and calling shrink_slab() on each iteration. The temptation is to supply a patch that checks if kswapd was woken for THP and if so ignore pgdat->kswapd_max_order but it'll be a hack and not backed up by proper testing. As 3.7 is very close to release and this is not a bug we should release with, a safer path is to revert "mm: remove __GFP_NO_KSWAPD" for now and revisit it with the view to ironing out the balance_pgdat() logic in general. Signed-off-by: Mel Gorman Cc: Zdenek Kabelac Cc: Seth Jennings Cc: Valdis Kletnieks Cc: Jiri Slaby Cc: Rik van Riel Cc: Robert Jennings Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/trace/events/gfpflags.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/trace') diff --git a/include/trace/events/gfpflags.h b/include/trace/events/gfpflags.h index 9391706e9254..d6fd8e5b14b7 100644 --- a/include/trace/events/gfpflags.h +++ b/include/trace/events/gfpflags.h @@ -36,6 +36,7 @@ {(unsigned long)__GFP_RECLAIMABLE, "GFP_RECLAIMABLE"}, \ {(unsigned long)__GFP_MOVABLE, "GFP_MOVABLE"}, \ {(unsigned long)__GFP_NOTRACK, "GFP_NOTRACK"}, \ + {(unsigned long)__GFP_NO_KSWAPD, "GFP_NO_KSWAPD"}, \ {(unsigned long)__GFP_OTHER_NODE, "GFP_OTHER_NODE"} \ ) : "GFP_NOWAIT" -- cgit v1.2.2 From a50915394f1fc02c2861d3b7ce7014788aa5066e Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Thu, 29 Nov 2012 13:54:27 -0800 Subject: revert "Revert "mm: remove __GFP_NO_KSWAPD"" It apepars that this patch was innocent, and we hope that "mm: avoid waking kswapd for THP allocations when compaction is deferred or contended" will fix the final kswapd-spinning cause. Cc: Zdenek Kabelac Cc: Seth Jennings Cc: Valdis Kletnieks Cc: Jiri Slaby Cc: Rik van Riel Cc: Robert Jennings Cc: Mel Gorman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/trace/events/gfpflags.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include/trace') diff --git a/include/trace/events/gfpflags.h b/include/trace/events/gfpflags.h index d6fd8e5b14b7..9391706e9254 100644 --- a/include/trace/events/gfpflags.h +++ b/include/trace/events/gfpflags.h @@ -36,7 +36,6 @@ {(unsigned long)__GFP_RECLAIMABLE, "GFP_RECLAIMABLE"}, \ {(unsigned long)__GFP_MOVABLE, "GFP_MOVABLE"}, \ {(unsigned long)__GFP_NOTRACK, "GFP_NOTRACK"}, \ - {(unsigned long)__GFP_NO_KSWAPD, "GFP_NO_KSWAPD"}, \ {(unsigned long)__GFP_OTHER_NODE, "GFP_OTHER_NODE"} \ ) : "GFP_NOWAIT" -- cgit v1.2.2