Re: [PATCH v8 12/21] x86/virt/seamldr: Install a new TDX module

["Edgecombe, Rick P" <rick.p.edgecombe@intel.com>]

On Thu, 2026-04-30 at 12:00 -0700, Dave Hansen wrote:
> On 4/27/26 08:28, Chao Gao wrote:
> > +			case MODULE_UPDATE_CPU_INSTALL:
> > +				ret = seamldr_install(seamldr_params);
> > +				break;
> 
> I really despise this naming. Could you please clarify with comments?
> 
> This reads like it is installing a seamldr, not telling a seamldr to
> perform an install.

Yea it does read that way. We kind of have a convention around the seamcall
wrappers matching the tdx docs name. tdh_foo_bar() is pretty clear.
Unfortunately the seamldr calls are defined like SEAMLDR.INSTALL.

We don't need to make matching the wrappers name match the tdx docs be a rule. I
just considered trying to make a seamldr name schema to make it clear that these
are seamldr call names. But it is probably better to just special case this one
with a clearer name we give it, like seamldr_tdx_module_install(). Which is
slightly long, but both clear on the purpose and that it is in the family of
seamldr_() functions.