Received-SPF: pass (google.com: domain of linux-kernel+bounces-15232-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
From: Bibo Mao <maobibo@loongson.cn>
To: Huacai Chen <chenhuacai@kernel.org>,
	Tianrui Zhao <zhaotianrui@loongson.cn>,
	Juergen Gross <jgross@suse.com>
Cc: loongarch@lists.linux.dev,
	linux-kernel@vger.kernel.org,
	virtualization@lists.linux.dev,
	kvm@vger.kernel.org
Subject: [PATCH 0/5] LoongArch: Add pv ipi support on LoongArch VM
Date: Wed,  3 Jan 2024 15:16:10 +0800
Message-Id: <20240103071615.3422264-1-maobibo@loongson.cn>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit

This patchset adds pv ipi support for VM. On physical machine, ipi HW
uses IOCSR registers, however there will be trap into hypervisor with
IOCSR registers access. This patch uses SWI interrupt for IPI
notification. During ipi sending with hypercall method, there is still
one trap; however with SWI interrupt handler there is no trap.

This patch passes to runltp testcases, and unixbench score is 99% of
that on physical machine on 3C5000 single way machine. Here is unixbench
score with 16 cores on 3C5000 single way machine.

----------------UnixBench score on 3C5000 machine with 16 cores --------
Dhrystone 2 using register variables         116700.0  339749961.8  29113.1
Double-Precision Whetstone                       55.0      57716.9  10494.0
Execl Throughput                                 43.0      33563.4   7805.4
File Copy 1024 bufsize 2000 maxblocks          3960.0    1017912.5   2570.5
File Copy 256 bufsize 500 maxblocks            1655.0     260061.4   1571.4
File Copy 4096 bufsize 8000 maxblocks          5800.0    3216109.4   5545.0
Pipe Throughput                               12440.0   18404312.0  14794.5
Pipe-based Context Switching                   4000.0    3395856.2   8489.6
Process Creation                                126.0      55684.8   4419.4
Shell Scripts (1 concurrent)                     42.4      55901.8  13184.4
Shell Scripts (8 concurrent)                      6.0       7396.5  12327.5
System Call Overhead                          15000.0    6997351.4   4664.9
System Benchmarks Index Score                                        7288.6

----------------UnixBench score on VM with 16 cores -----------------
Dhrystone 2 using register variables         116700.0  341649555.5  29275.9
Double-Precision Whetstone                       55.0      57490.9  10452.9
Execl Throughput                                 43.0      33663.8   7828.8
File Copy 1024 bufsize 2000 maxblocks          3960.0    1047631.2   2645.5
File Copy 256 bufsize 500 maxblocks            1655.0     286671.0   1732.2
File Copy 4096 bufsize 8000 maxblocks          5800.0    3243588.2   5592.4
Pipe Throughput                               12440.0   16353087.8  13145.6
Pipe-based Context Switching                   4000.0    3100690.0   7751.7
Process Creation                                126.0      51502.1   4087.5
Shell Scripts (1 concurrent)                     42.4      56665.3  13364.4
Shell Scripts (8 concurrent)                      6.0       7412.1  12353.4
System Call Overhead                          15000.0    6962239.6   4641.5
System Benchmarks Index Score                                        7205.8

Bibo Mao (5):
  LoongArch: KVM: Add hypercall instruction emulation support
  LoongArch: KVM: Add cpucfg area for kvm hypervisor
  LoongArch/smp: Refine ipi ops on LoongArch platform
  LoongArch: Add paravirt interface for guest kernel
  LoongArch: Add pv ipi support on LoongArch system

 arch/loongarch/Kconfig                        |   7 +
 arch/loongarch/include/asm/Kbuild             |   1 -
 arch/loongarch/include/asm/hardirq.h          |   5 +
 arch/loongarch/include/asm/inst.h             |   1 +
 arch/loongarch/include/asm/irq.h              |  10 +-
 arch/loongarch/include/asm/kvm_para.h         | 157 ++++++++++++++++++
 arch/loongarch/include/asm/loongarch.h        |  10 ++
 arch/loongarch/include/asm/paravirt.h         |  27 +++
 .../include/asm/paravirt_api_clock.h          |   1 +
 arch/loongarch/include/asm/smp.h              |  31 ++--
 arch/loongarch/include/uapi/asm/Kbuild        |   2 -
 arch/loongarch/kernel/Makefile                |   1 +
 arch/loongarch/kernel/irq.c                   |  24 +--
 arch/loongarch/kernel/paravirt.c              | 144 ++++++++++++++++
 arch/loongarch/kernel/perf_event.c            |  14 +-
 arch/loongarch/kernel/setup.c                 |   2 +
 arch/loongarch/kernel/smp.c                   |  61 ++++---
 arch/loongarch/kernel/time.c                  |  12 +-
 arch/loongarch/kvm/exit.c                     | 118 +++++++++++--
 19 files changed, 527 insertions(+), 101 deletions(-)
 create mode 100644 arch/loongarch/include/asm/kvm_para.h
 create mode 100644 arch/loongarch/include/asm/paravirt.h
 create mode 100644 arch/loongarch/include/asm/paravirt_api_clock.h
 delete mode 100644 arch/loongarch/include/uapi/asm/Kbuild
 create mode 100644 arch/loongarch/kernel/paravirt.c


base-commit: 610a9b8f49fbcf1100716370d3b5f6f884a2835a
-- 
2.39.3