Re: [PATCH 1/2] x86/fred: Fix build with clang

[H. Peter Anvin]

On 1/27/24 11:46, Li, Xin3 wrote:
> > > diff --git a/arch/x86/entry/entry_64_fred.S
> > > b/arch/x86/entry/entry_64_fred.S index eedf98de7538..5427e0da190d
> > > 100644
> > > --- a/arch/x86/entry/entry_64_fred.S
> > > +++ b/arch/x86/entry/entry_64_fred.S
> > > @@ -43,13 +43,12 @@ SYM_INNER_LABEL(asm_fred_exit_user,
> > SYM_L_GLOBAL)
> > > 	_ASM_EXTABLE_TYPE(1b, asm_fred_entrypoint_user, EX_TYPE_ERETU)
> > > SYM_CODE_END(asm_fred_entrypoint_user)
> > >
> > > -.fill asm_fred_entrypoint_kernel - ., 1, 0xcc
> > > -
> > > /*
> > >   * The new RIP value that FRED event delivery establishes is
> > >   * (IA32_FRED_CONFIG & ~FFFH) + 256 for events that occur in
> > >   * ring 0, i.e., asm_fred_entrypoint_user + 256.
> > >   */
> > > +	.fill asm_fred_entrypoint_user + 256 - ., 1, 0xcc
> > > 	.org asm_fred_entrypoint_user + 256
> > > SYM_CODE_START_NOALIGN(asm_fred_entrypoint_kernel)
> > > 	FRED_ENTER
> >
> > .fill and .org here are redundant; in fact, there two directives mean exactly the
> > same thing except that .org implicitly subtracts the current offset.
>
> Ah, right, .fill already does the job!
>
> I will remove .org.

Incidentally, was there a problem with .org ..., 0xcc?

Not a criticism, I just want to know to better understand current 
binutils limitations.

	-hpa