Re: [RFC PATCH] x86/virt/tdx: Disable KEXEC in the presence of TDX

[Nikolay Borisov]

On 19.01.24 г. 2:52 ч., Huang, Kai wrote:
> On Fri, 2024-01-19 at 02:49 +0200, Nikolay Borisov wrote:
> > [Adding Kirill as he has a series in flight about KEXEC]
> >
> > On 18.01.24 г. 23:31 ч., Huang, Kai wrote:
> > > On Thu, 2024-01-18 at 18:01 +0200, Nikolay Borisov wrote:
> > > > TDX doesn't currently support kexec so disable the latter when TDX
> > > > is detected at boot time. If kexec must absolutely be supported then
> > > > TDX can be disabled from the bios.
> > > >
> > > > Making this decision at run time rather than as a compile time option
> > > > allows distribution kernels to have both enabled and delegate to the
> > > > user whether they want to use TDX/Kexec.
> > > >
> > > > Signed-off-by: Nikolay Borisov <nik.borisov@xxxxxxxx>
> > >
> > > Hi Nikolay,
> > >
> > > Thanks for the patch.
> > >
> > > I am working in progress on patches to make kexec() work with TDX.  I plan to
> > > send them out soon (e.g., next week), but I think perhaps your patch is
> > > reasonable before that is merged.
> >
> > I thought Kirill's series rectifies this ?
>
> No that's for TDX guest.

Well in this case shouldn't INTEL_TDX_GUEST also be gated on KEXEC_CORE 
being inoperable either via Kconfig option or via a similar approach to 
this one (if it's accepted)?