Re: [RFC PATCH 08/27] x86/virt/tdx: Add tdx_enable_ext() to enable of TDX Module Extensions

[Xu Yilun <yilun.xu@linux.intel.com>]

On Thu, Oct 30, 2025 at 10:55:51AM +0000, Jonathan Cameron wrote:
> On Fri, 19 Sep 2025 07:22:17 -0700
> Dan Williams <dan.j.williams@intel.com> wrote:
> 
> > From: Xu Yilun <yilun.xu@linux.intel.com>
> > 
> > tdx_enable() implements a simple state machine with @tdx_module_status to
> > determine if TDX is already enabled, or failed to enable. Add another state
> > to that enum (TDX_MODULE_INITIALIZED_EXT) to track if extensions have been enabled.
> > 
> > The extension initialization uses the new TDH.EXT.MEM.ADD and TDX.EXT.INIT
> > seamcalls.
> > 
> > Note that this extension initialization does not impact existing in-flight
> > SEAMCALLs that are not implemented by the extension. So only the first user
> > of an extension-seamcall needs invoke this helper.
> > 
> > Co-developed-by: Zhenzhong Duan <zhenzhong.duan@intel.com>
> > Signed-off-by: Zhenzhong Duan <zhenzhong.duan@intel.com>
> > Signed-off-by: Xu Yilun <yilun.xu@linux.intel.com>
> > Signed-off-by: Dan Williams <dan.j.williams@intel.com>
> 
> A few more trivial comments.
> 
> > diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
> > index d47b2612c816..9d4cebace054 100644
> > --- a/arch/x86/virt/vmx/tdx/tdx.c
> > +++ b/arch/x86/virt/vmx/tdx/tdx.c
> 
> > +
> > +DEFINE_FREE(tdx_ext_mempool_free, struct tdx_page_array *, if (!IS_ERR_OR_NULL(_T)) tdx_ext_mempool_free(_T))
> Very long line. Add a break somewhere!

Yes, will wrap.

> 
> 
> > +/**
> > + * tdx_enable_ext - Enable TDX module extensions.
> > + *
> > + * This function assumes the caller has done VMXON.
> > + *
> > + * This function can be called in parallel by multiple callers.
> > + *
> > + * Return 0 if TDX module extension is enabled successfully, otherwise error.
> > + */
> > +int tdx_enable_ext(void)
> > +{
> > +	int ret;
> > +
> > +	mutex_lock(&tdx_module_lock);
> guard() perhaps which would make early returns an option if nothing esle
> gets added after the switch.

Will do. New version will be based on Sean's VMXON changes [1] so has
some differences, but guard() is still a good idea.

[1] https://lore.kernel.org/linux-coco/20251010220403.987927-4-seanjc@google.com/

Thanks,
Yilun