Re: [PATCH] x86/tdx: Remove the early #VE handler
From: Kiryl Shutsemau
Date: Thu Sep 24 2026 - 07:02:55 EST
On Tue, Sep 22, 2026 at 05:45:03PM +0000, Verma, Vishal L wrote:
> On Tue, 2026-09-22 at 11:19 +0100, Kiryl Shutsemau wrote:
> >
> > Could you just check what happens you panic early (like in
> > tdx_early_init()) with and without the patch?
>
> Tested this:
>
> ---
>
> arch/x86/coco/tdx/tdx.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/arch/x86/coco/tdx/tdx.c b/arch/x86/coco/tdx/tdx.c
> index f904a636d449..fcb2ecf055d1 100644
> --- a/arch/x86/coco/tdx/tdx.c
> +++ b/arch/x86/coco/tdx/tdx.c
> @@ -1192,4 +1192,7 @@ void __init tdx_early_init(void)
> x86_cpuinit.parallel_bringup = false;
>
> tdx_announce();
> +
> + panic("early #VE window panic test");
> }
>
> ---
>
> In both cases, with and without the removal patch, the guest just hangs
> without any actual text in the serial output. I didn't add any panic=
> cmdline, and CONFIG_PANIC_TIMEOUT was the default 0.
Okay, I run out of excuses. We can revert it later, if there's an actual
problem I cannot recall.
Reviewed-by: Kiryl Shutsemau (Meta) <kas@xxxxxxxxxx>
--
Kiryl Shutsemau / Kirill A. Shutemov