Re: [PATCH] x86/tdx: Remove the early #VE handler

[Kiryl Shutsemau <kas@kernel.org>]

On Mon, Sep 21, 2026 at 09:20:18PM +0000, Verma, Vishal L wrote:
> On Fri, 2026-09-11 at 11:37 +0100, Kiryl Shutsemau wrote:
> > 
> > Rick, could you actually test it?
> > 
> > Are we sure there's no other port I/O in before idt_setup_early_traps()?
> > 
> > I don't see anything direct, but exception path is different story.
> > 
> > native_machine_emergency_restart() does port I/O for BOOT_KBD which
> > seems to be reachable and leads to #VE with the patch:
> > 
> > machine_emergency_restart()
> >   __machine_emergency_restart(1)
> >     machine_ops.emergency_restart()
> >       native_machine_emergency_restart()
> >         reboot_type == BOOT_ACPI (default)
> > 	acpi_reboot()
> > 	reboot_type = BOOT_KBD
> >   	  kb_wait()
> >             inb(0x64) --> #VE
> > 
> > Could you check if the patch changes panic() behaviour in the window
> > before idt_setup_early_traps()?
> 
> Hi Kiryl - not sure how I would test this - but looking at it a bit
> more - 
> 
> vpanic() only goes to the machine_emergency_restart() path if
> CONFIG_PANIC_TIMEOUT is something other than 0. Distros and Kconfig
> both default it to 0. This could be overridden by panic=N on the
> command line, but that is parsed after idt_setup_early_traps(), so that
> would behave normally.
> 
> So at least this path should, in practice, never cause a problem.

Could you just check what happens you panic early (like in
tdx_early_init()) with and without the patch?

-- 
  Kiryl Shutsemau / Kirill A. Shutemov