qemu/include
Longpeng(Mike) 14c985cffa target-i386: present virtual L3 cache info for vcpus
Some software algorithms are based on the hardware's cache info, for example,
for x86 linux kernel, when cpu1 want to wakeup a task on cpu2, cpu1 will trigger
a resched IPI and told cpu2 to do the wakeup if they don't share low level
cache. Oppositely, cpu1 will access cpu2's runqueue directly if they share llc.
The relevant linux-kernel code as bellow:

	static void ttwu_queue(struct task_struct *p, int cpu)
	{
		struct rq *rq = cpu_rq(cpu);
		......
		if (... && !cpus_share_cache(smp_processor_id(), cpu)) {
			......
			ttwu_queue_remote(p, cpu); /* will trigger RES IPI */
			return;
		}
		......
		ttwu_do_activate(rq, p, 0); /* access target's rq directly */
		......
	}

In real hardware, the cpus on the same socket share L3 cache, so one won't
trigger a resched IPIs when wakeup a task on others. But QEMU doesn't present a
virtual L3 cache info for VM, then the linux guest will trigger lots of RES IPIs
under some workloads even if the virtual cpus belongs to the same virtual socket.

For KVM, there will be lots of vmexit due to guest send IPIs.
The workload is a SAP HANA's testsuite, we run it one round(about 40 minuates)
and observe the (Suse11sp3)Guest's amounts of RES IPIs which triggering during
the period:
        No-L3           With-L3(applied this patch)
cpu0:	363890		44582
cpu1:	373405		43109
cpu2:	340783		43797
cpu3:	333854		43409
cpu4:	327170		40038
cpu5:	325491		39922
cpu6:	319129		42391
cpu7:	306480		41035
cpu8:	161139		32188
cpu9:	164649		31024
cpu10:	149823		30398
cpu11:	149823		32455
cpu12:	164830		35143
cpu13:	172269		35805
cpu14:	179979		33898
cpu15:	194505		32754
avg:	268963.6	40129.8

The VM's topology is "1*socket 8*cores 2*threads".
After present virtual L3 cache info for VM, the amounts of RES IPIs in guest
reduce 85%.

For KVM, vcpus send IPIs will cause vmexit which is expensive, so it can cause
severe performance degradation. We had tested the overall system performance if
vcpus actually run on sparate physical socket. With L3 cache, the performance
improves 7.2%~33.1%(avg:15.7%).

Signed-off-by: Longpeng(Mike) <longpeng2@huawei.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2016-09-09 20:58:34 +03:00
..
block drive-backup: added support for data compression 2016-09-05 19:06:48 +02:00
crypto crypto: add support for querying parameters for block encryption 2016-07-26 17:46:37 +02:00
disas disas: Fix ATTRIBUTE_UNUSED define clash with ALSA headers 2016-07-19 16:40:39 +01:00
exec tcg: Reorg TCGOp chaining 2016-08-05 21:44:18 +05:30
fpu softfloat: Fix warn about implicit conversion from int to int8_t 2016-08-15 16:15:38 +01:00
hw target-i386: present virtual L3 cache info for vcpus 2016-09-09 20:58:34 +03:00
io Clean up ill-advised or unusual header guards 2016-07-12 16:20:46 +02:00
libdecnumber Clean up ill-advised or unusual header guards 2016-07-12 16:20:46 +02:00
migration Clean up decorations and whitespace around header guards 2016-07-12 16:20:46 +02:00
monitor monitor: fix crash when leaving qemu with spice audio 2016-08-08 14:16:11 +02:00
net net: Use correct type for bool flag 2016-07-19 20:18:27 +02:00
qapi blockjob: Update description of the 'id' field 2016-07-13 13:26:02 +02:00
qemu coroutine: Assert that no locks are held on termination 2016-09-05 19:06:48 +02:00
qom exec: Set cpu_index only if it's not been explictly set 2016-07-26 15:32:01 -03:00
standard-headers linux-headers: update 2016-09-05 15:15:16 +02:00
sysemu Block layer patches 2016-09-06 17:18:17 +01:00
ui clang: Fix warning reg. expansion to 'defined' 2016-08-09 22:57:36 +02:00
elf.h Clean up ill-advised or unusual header guards 2016-07-12 16:20:46 +02:00
glib-compat.h glib-compat: add g_(s)list_free_full() 2016-09-08 17:57:32 +04:00
qemu-common.h Update ancient copyright string in -version output 2016-08-11 16:24:53 +01:00
qemu-io.h
trace-tcg.h Clean up decorations and whitespace around header guards 2016-07-12 16:20:46 +02:00
trace.h Clean up decorations and whitespace around header guards 2016-07-12 16:20:46 +02:00