On 07/10/2024 3:03 pm, Roger Pau Monne wrote:
> Some error paths in the MSR state loading logic don't contain error messages,
> which makes debugging them quite hard without adding extra patches to print
> the
> information.
>
> Add two new log messages to the MSR state load path that print information
> about the entry that failed to load.
>
> No functional change intended.
>
> Signed-off-by: Roger Pau Monné <[email protected]>
> ---
> xen/arch/x86/hvm/hvm.c | 9 +++++++++
Can we fix the PV side at the same time too?
> 1 file changed, 9 insertions(+)
>
> diff --git a/xen/arch/x86/hvm/hvm.c b/xen/arch/x86/hvm/hvm.c
> index 69a25571db8d..c71087f636c4 100644
> --- a/xen/arch/x86/hvm/hvm.c
> +++ b/xen/arch/x86/hvm/hvm.c
> @@ -1598,10 +1598,19 @@ static int cf_check hvm_load_cpu_msrs(struct domain
> *d, hvm_domain_context_t *h)
> rc = guest_wrmsr(v, ctxt->msr[i].index, ctxt->msr[i].val);
>
> if ( rc != X86EMUL_OKAY )
> + {
> + printk(XENLOG_G_ERR
> + "HVM%d.%d load MSR %#x with value %#lx failed: %d\n",
> + d->domain_id, vcpuid, ctxt->msr[i].index,
> + ctxt->msr[i].val, rc);
Just %pv please. I don't want to propagate the (occasionally ambiguous)
HVM%d form.
Also, rc may not be great to render. It's an X86EMUL_*, not an errno.
And saying that, we have a discontinuity between PV and HVM. PV
translates !OKAY into -EINVAL, whereas HVM translates into -ENXIO. /sigh
~Andrew
