Re: [PATCHv4] kvmppc: Implement H_LOGICAL_CI_{LOAD,STORE} in KVM

[Thomas Huth <thuth@redhat.com>]

Am Tue, 21 Apr 2015 10:41:51 +1000
schrieb David Gibson <david@gibson.dropbear.id.au>:

> On POWER, storage caching is usually configured via the MMU - attributes
> such as cache-inhibited are stored in the TLB and the hashed page table.
> 
> This makes correctly performing cache inhibited IO accesses awkward when
> the MMU is turned off (real mode).  Some CPU models provide special
> registers to control the cache attributes of real mode load and stores but
> this is not at all consistent.  This is a problem in particular for SLOF,
> the firmware used on KVM guests, which runs entirely in real mode, but
> which needs to do IO to load the kernel.
> 
> To simplify this qemu implements two special hypercalls, H_LOGICAL_CI_LOAD
> and H_LOGICAL_CI_STORE which simulate a cache-inhibited load or store to
> a logical address (aka guest physical address).  SLOF uses these for IO.
> 
> However, because these are implemented within qemu, not the host kernel,
> these bypass any IO devices emulated within KVM itself.  The simplest way
> to see this problem is to attempt to boot a KVM guest from a virtio-blk
> device with iothread / dataplane enabled.  The iothread code relies on an
> in kernel implementation of the virtio queue notification, which is not
> triggered by the IO hcalls, and so the guest will stall in SLOF unable to
> load the guest OS.
> 
> This patch addresses this by providing in-kernel implementations of the
> 2 hypercalls, which correctly scan the KVM IO bus.  Any access to an
> address not handled by the KVM IO bus will cause a VM exit, hitting the
> qemu implementation as before.
> 
> Note that a userspace change is also required, in order to enable these
> new hcall implementations with KVM_CAP_PPC_ENABLE_HCALL.
> 
> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
> ---
>  arch/powerpc/include/asm/kvm_book3s.h |  3 ++
>  arch/powerpc/kvm/book3s.c             | 76 +++++++++++++++++++++++++++++++++++
>  arch/powerpc/kvm/book3s_hv.c          | 12 ++++++
>  arch/powerpc/kvm/book3s_pr_papr.c     | 28 +++++++++++++
>  4 files changed, 119 insertions(+)
...
> diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
> index cfbcdc6..453a8a4 100644
> --- a/arch/powerpc/kvm/book3s.c
> +++ b/arch/powerpc/kvm/book3s.c
> @@ -821,6 +821,82 @@ void kvmppc_core_destroy_vm(struct kvm *kvm)
>  #endif
>  }
>  
> +int kvmppc_h_logical_ci_load(struct kvm_vcpu *vcpu)
> +{
> +	unsigned long size = kvmppc_get_gpr(vcpu, 4);
> +	unsigned long addr = kvmppc_get_gpr(vcpu, 5);
> +	u64 buf;
> +	int ret;
> +
> +	if (!is_power_of_2(size) || (size > sizeof(buf)))
> +		return H_TOO_HARD;
> +
> +	ret = kvm_io_bus_read(vcpu, KVM_MMIO_BUS, addr, size, &buf);
> +	if (ret != 0)
> +		return H_TOO_HARD;
> +
> +	switch (size) {
> +	case 1:
> +		kvmppc_set_gpr(vcpu, 4, *(u8 *)&buf);
> +		break;
> +

Most of the code in book3s.c seems not to use a empty line after a
"break;", so may I suggest to remove these empty lines here, too, to
keep the coding style a little bit more consistent?

> +	case 2:
> +		kvmppc_set_gpr(vcpu, 4, be16_to_cpu(*(__be16 *)&buf));
> +		break;
> +
> +	case 4:
> +		kvmppc_set_gpr(vcpu, 4, be32_to_cpu(*(__be32 *)&buf));
> +		break;
> +
> +	case 8:
> +		kvmppc_set_gpr(vcpu, 4, be64_to_cpu(*(__be64 *)&buf));
> +		break;
> +
> +	default:
> +		BUG();

If I got the code right, a malicious guest could easily trigger this
BUG() statement, couldn't it? ... so a BUG() is maybe not the right
thing to do here. Would it be appropriate to return an error value to
the guest instead?

> +	}
> +
> +	return H_SUCCESS;
> +}
> +EXPORT_SYMBOL_GPL(kvmppc_h_logical_ci_load);

 Thomas