Re: [PATCH v3 04/16] x86/virt/tdx: Allocate page bitmap for Dynamic PAMT

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

On Fri, 2025-09-26 at 16:41 +0800, Xiaoyao Li wrote:
> It looks like the update is not produced by the script[1], but
> manually, right?
> 
> Looking at the history
> 
>    git log arch/x86/virt/vmx/tdx/tdx_global_metadata.c
> 
> It looks the expectation is always adding new fields by updating the 
> script[1] and running it with the latest 'tdx_global_metadata.c'.

Yes, it's not symmetrical. It was probably added manually. I don't see
why we actually need the the part that makes it unsymmetrical
(tdx_supports_dynamic_pamt() check). In fact I think it's bad because
it depends on specific call order of the metadata reading to be valid.

Kirill, any reason we can't drop it?

As for a strict requirement to use the script, it wasn't my
understanding.

> 
> Dan Williams also expressed internally that we should have checked
> the script[1] into the kernel. I agree with him and it's never too
> lateto do it.
> 
> [1]
> https://lore.kernel.org/kvm/0853b155ec9aac09c594caa60914ed6ea4dc0a71.camel@intel.com
> /

Especially while the ABI docs file format is being revisited, it
doesn't seem like the right time to check it in.

For this patch we could have used it, however at this point it would
amount to verifying that it output the same few lines of code. Would
you like to do this test? (Assuming we can drop the
tdx_supports_dynamic_pamt() check) Probably a note that the code
generation matched our manual implementation would be a point of
reassurance to add to the log.