Re: [PATCH v2] KVM: Optimize kvm_arch_vcpu_ioctl_run function

From: Tianjia Zhang
Date: Thu Apr 16 2020 - 05:09:49 EST




On 2020/4/16 16:50, Cornelia Huck wrote:
On Thu, 16 Apr 2020 16:45:33 +0800
Tianjia Zhang <tianjia.zhang@xxxxxxxxxxxxxxxxx> wrote:

On 2020/4/16 16:28, Marc Zyngier wrote:
On 2020-04-16 08:03, Vitaly Kuznetsov wrote:
Tianjia Zhang <tianjia.zhang@xxxxxxxxxxxxxxxxx> writes:
In earlier versions of kvm, 'kvm_run' is an independent structure
and is not included in the vcpu structure. At present, 'kvm_run'
is already included in the vcpu structure, so the parameter
'kvm_run' is redundant.

This patch simplify the function definition, removes the extra
'kvm_run' parameter, and extract it from the 'kvm_vcpu' structure
if necessary.

Signed-off-by: Tianjia Zhang <tianjia.zhang@xxxxxxxxxxxxxxxxx>
---

v2 change:
 remove 'kvm_run' parameter and extract it from 'kvm_vcpu'

Âarch/mips/kvm/mips.cÂÂÂÂÂÂ |Â 3 ++-
Âarch/powerpc/kvm/powerpc.c |Â 3 ++-
Âarch/s390/kvm/kvm-s390.cÂÂ |Â 3 ++-
Âarch/x86/kvm/x86.cÂÂÂÂÂÂÂÂ | 11 ++++++-----
Âinclude/linux/kvm_host.hÂÂ |Â 2 +-
Âvirt/kvm/arm/arm.cÂÂÂÂÂÂÂÂ |Â 6 +++---
Âvirt/kvm/kvm_main.cÂÂÂÂÂÂÂ |Â 2 +-
Â7 files changed, 17 insertions(+), 13 deletions(-)

Overall, there is a large set of cleanups to be done when both the vcpu
and the run
structures are passed as parameters at the same time. Just grepping the
tree for
kvm_run is pretty instructive.

ÂÂÂÂÂÂÂ M.

Sorry, it's my mistake, I only compiled the x86 platform, I will submit
patch again.

I think it's completely fine (and even preferable) to do cleanups like
that on top.

[FWIW, I compiled s390 here.]


Very good, I will do a comprehensive cleanup of this type of code.

Thanks,
Tianjia