[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-devel] [PULL 03/11] target-arm: Use Neon for zero checking
From: |
Peter Maydell |
Subject: |
[Qemu-devel] [PULL 03/11] target-arm: Use Neon for zero checking |
Date: |
Thu, 14 Jul 2016 17:29:35 +0100 |
From: Vijay <address@hidden>
Use Neon instructions to perform zero checking of
buffer. This is helps in reducing total migration time.
Use case: Idle VM live migration with 4 VCPUS and 8GB ram
running CentOS 7.
Without Neon, the Total migration time is 3.5 Sec
Migration status: completed
total time: 3560 milliseconds
downtime: 33 milliseconds
setup: 5 milliseconds
transferred ram: 297907 kbytes
throughput: 685.76 mbps
remaining ram: 0 kbytes
total ram: 8519872 kbytes
duplicate: 2062760 pages
skipped: 0 pages
normal: 69808 pages
normal bytes: 279232 kbytes
dirty sync count: 3
With Neon, the total migration time is 2.9 Sec
Migration status: completed
total time: 2960 milliseconds
downtime: 65 milliseconds
setup: 4 milliseconds
transferred ram: 299869 kbytes
throughput: 830.19 mbps
remaining ram: 0 kbytes
total ram: 8519872 kbytes
duplicate: 2064313 pages
skipped: 0 pages
normal: 70294 pages
normal bytes: 281176 kbytes
dirty sync count: 3
Signed-off-by: Vijaya Kumar K <address@hidden>
Signed-off-by: Suresh <address@hidden>
Acked-by: Paolo Bonzini <address@hidden>
Message-id: address@hidden
Reviewed-by: Peter Maydell <address@hidden>
Signed-off-by: Peter Maydell <address@hidden>
---
util/cutils.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/util/cutils.c b/util/cutils.c
index 5830a68..7505fda 100644
--- a/util/cutils.c
+++ b/util/cutils.c
@@ -184,6 +184,13 @@ int qemu_fdatasync(int fd)
#define SPLAT(p) _mm_set1_epi8(*(p))
#define ALL_EQ(v1, v2) (_mm_movemask_epi8(_mm_cmpeq_epi8(v1, v2)) == 0xFFFF)
#define VEC_OR(v1, v2) (_mm_or_si128(v1, v2))
+#elif defined(__aarch64__)
+#include "arm_neon.h"
+#define VECTYPE uint64x2_t
+#define ALL_EQ(v1, v2) \
+ ((vgetq_lane_u64(v1, 0) == vgetq_lane_u64(v2, 0)) && \
+ (vgetq_lane_u64(v1, 1) == vgetq_lane_u64(v2, 1)))
+#define VEC_OR(v1, v2) ((v1) | (v2))
#else
#define VECTYPE unsigned long
#define SPLAT(p) (*(p) * (~0UL / 255))
--
1.9.1
- [Qemu-devel] [PULL 00/11] target-arm queue, Peter Maydell, 2016/07/14
- [Qemu-devel] [PULL 10/11] ast2400: pretend DMAs are done for U-boot, Peter Maydell, 2016/07/14
- [Qemu-devel] [PULL 11/11] ast2400: externalize revision numbers, Peter Maydell, 2016/07/14
- [Qemu-devel] [PULL 09/11] ast2400: replace aspeed_smc_is_implemented(), Peter Maydell, 2016/07/14
- [Qemu-devel] [PULL 08/11] hw/misc: fix typo in Aspeed SCU hw-strap2 property name, Peter Maydell, 2016/07/14
- [Qemu-devel] [PULL 04/11] gic: provide defines for v2/v3 targetlist sizes, Peter Maydell, 2016/07/14
- [Qemu-devel] [PULL 07/11] m25p80: Fix QIOR/DIOR handling for Winbond, Peter Maydell, 2016/07/14
- [Qemu-devel] [PULL 05/11] hw/arm/virt: tcg: adjust MPIDR like KVM, Peter Maydell, 2016/07/14
- [Qemu-devel] [PULL 06/11] target-arm: Add missed AArch32 TLBI sytem registers, Peter Maydell, 2016/07/14
- [Qemu-devel] [PULL 03/11] target-arm: Use Neon for zero checking,
Peter Maydell <=
- [Qemu-devel] [PULL 02/11] Revert "hw/ptimer: Perform counter wrap around if timer already expired", Peter Maydell, 2016/07/14
- [Qemu-devel] [PULL 01/11] virtio-mmio: format transport base address in BusClass.get_dev_path, Peter Maydell, 2016/07/14
- Re: [Qemu-devel] [PULL 00/11] target-arm queue, Peter Maydell, 2016/07/14