Takuya Yoshikawa wrote:
Fix the x86 emulation of in:kvm_emulate_pio() and complete_pio() both read out the RAX register value and copy it to a place into which the value read out from the port will be copied later. This patch removes this redundancy. /*** snippet from arch/x86/kvm/x86.c ***/ int complete_pio(struct kvm_vcpu *vcpu) { ... if (!io->string) { if (io->in) { val = kvm_register_read(vcpu, VCPU_REGS_RAX); memcpy(&val, vcpu->arch.pio_data, io->size); kvm_register_write(vcpu, VCPU_REGS_RAX, val); } ...
Self explanation: reading RAX before copying pio data is for preserving higher bits of the register. If we are sure that pio emulation ends by complete_pio() and complete_pio() does this higher bits preservation, we no longer need to mind preserving the higer bits of arch.pio_data in each device's read handler: may reduce some switches and memcpies, if we want. Anyway, if no device's read handler depends on the memcpied 4 bytes of RAX, I hope so, this patch seems to work and reduces extra memcpy.
Signed-off-by: Takuya Yoshikawa <yoshikawa.tak...@oss.ntt.co.jp> --- arch/x86/kvm/x86.c | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index d47ceda..fcbe3a7 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -3550,8 +3550,10 @@ int kvm_emulate_pio(struct kvm_vcpu *vcpu, int in, int size, unsigned port) trace_kvm_pio(vcpu->run->io.direction == KVM_EXIT_IO_OUT, port, size, 1);- val = kvm_register_read(vcpu, VCPU_REGS_RAX);- memcpy(vcpu->arch.pio_data, &val, 4); + if (!vcpu->arch.pio.in) { + val = kvm_register_read(vcpu, VCPU_REGS_RAX); + memcpy(vcpu->arch.pio_data, &val, 4); + }if (!kernel_pio(vcpu, vcpu->arch.pio_data)) {complete_pio(vcpu);
-- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
