Re: [PATCH] x86/kvm, objtool: Avoid fastop ENDBR from being sealed

[Peter Zijlstra <peterz@infradead.org>]

On Wed, Aug 17, 2022 at 06:10:45PM -0700, Josh Poimboeuf wrote:
> On Wed, Aug 17, 2022 at 12:56:15PM +0200, Peter Zijlstra wrote:
> > Add (yet another) annotation to inhibit objtool from sealing a specific
> > ENDBR instance.
> 
> Nah, just add a throwaway reference to the function.  Then no objtool
> change is needed:

Nice!, should I 'borrow' this and respin the patch or you want to make a
real patch out of it?

> diff --git a/arch/x86/include/asm/ibt.h b/arch/x86/include/asm/ibt.h
> index 689880eca9ba..3ae795150374 100644
> --- a/arch/x86/include/asm/ibt.h
> +++ b/arch/x86/include/asm/ibt.h
> @@ -31,6 +31,16 @@
>  
>  #define __noendbr	__attribute__((nocf_check))
>  
> +/*
> + * Create a dummy function pointer reference to prevent objtool from marking
> + * the function as needing to be "sealed" (i.e. ENDBR converted to NOP by
> + * apply_ibt_endbr()).
> + */
> +#define IBT_NOSEAL(fname)				\
> +	".pushsection .discard.endbr.noseal\n\t"	\
> +	_ASM_PTR fname "\n\t"				\
> +	".popsection\n\t"
> +
>  static inline __attribute_const__ u32 gen_endbr(void)
>  {
>  	u32 endbr;
> diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c
> index f8382abe22ff..88623add0db4 100644
> --- a/arch/x86/kvm/emulate.c
> +++ b/arch/x86/kvm/emulate.c
> @@ -316,7 +316,8 @@ static int fastop(struct x86_emulate_ctxt *ctxt, fastop_t fop);
>  	".align " __stringify(FASTOP_SIZE) " \n\t" \
>  	".type " name ", @function \n\t" \
>  	name ":\n\t" \
> -	ASM_ENDBR
> +	ASM_ENDBR \
> +	IBT_NOSEAL(name)
>  
>  #define FOP_FUNC(name) \
>  	__FOP_FUNC(#name)