Re: [PATCH V4 04/50] x86/xen: Add xenpv_restore_regs_and_return_to_usermode()

[Borislav Petkov <bp@xxxxxxxxx>]

On Tue, Oct 26, 2021 at 10:13:34PM +0800, Lai Jiangshan wrote:
> From: Lai Jiangshan <laijs@xxxxxxxxxxxxxxxxx>
> 
> While in the native case, PER_CPU_VAR(cpu_tss_rw + TSS_sp0) is the
> trampoline stack.  But XEN pv doesn't use trampoline stack, so
> PER_CPU_VAR(cpu_tss_rw + TSS_sp0) is also the kernel stack.  Hence source
> and destination stacks are identical in that case, which means reusing
> swapgs_restore_regs_and_return_to_usermode() in XEN pv would cause %rsp
> to move up to the top of the kernel stack and leave the IRET frame below
> %rsp, which is dangerous to be corrupted if #NMI / #MC hit as either of
> these events occurring in the middle of the stack pushing would clobber
> data on the (original) stack.
> 
> And swapgs_restore_regs_and_return_to_usermode() pushing the IRET frame
> on to the original address is useless and error-prone when there is any
> future attempt to modify the code.
> 
> Fixes: 7f2590a110b8 ("x86/entry/64: Use a per-CPU trampoline stack for IDT 
> entries")
> Cc: Jan Beulich <jbeulich@xxxxxxxx>
> Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
> Cc: Juergen Gross <jgross@xxxxxxxx>
> Cc: Peter Anvin <hpa@xxxxxxxxx>
> Cc: xen-devel@xxxxxxxxxxxxxxxxxxxx>
> Reviewed-by: Boris Ostrovsky <boris.ostrovsky@xxxxxxxxxx>
> Signed-off-by: Lai Jiangshan <laijs@xxxxxxxxxxxxxxxxx>
> ---
>  arch/x86/entry/entry_64.S        |  9 ++++++---
>  arch/x86/entry/entry_64_compat.S |  7 ++++---
>  arch/x86/xen/xen-asm.S           | 27 +++++++++++++++++++++++++++
>  3 files changed, 37 insertions(+), 6 deletions(-)
> 
> diff --git a/arch/x86/entry/entry_64.S b/arch/x86/entry/entry_64.S
> index 9d468c8877e2..0dde5a253dda 100644
> --- a/arch/x86/entry/entry_64.S
> +++ b/arch/x86/entry/entry_64.S
> @@ -119,7 +119,7 @@ SYM_INNER_LABEL(entry_SYSCALL_64_after_hwframe, 
> SYM_L_GLOBAL)
>        * In the Xen PV case we must use iret anyway.
>        */
>  
> -     ALTERNATIVE "", "jmp    swapgs_restore_regs_and_return_to_usermode", \
> +     ALTERNATIVE "", "jmp xenpv_restore_regs_and_return_to_usermode", \

Instead of sprinkling all those ALTERNATIVE calls everywhere,
why don't you simply jump to the xenpv-one at the
swapgs_restore_regs_and_return_to_usermode label itself and have a
single ALTERNATIVE there?

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette