Re: [PATCH] x86/math64: handle #DE in mul_u64_u64_div_u64()

[Oleg Nesterov <oleg@redhat.com>]

Finally. If we really want to optimize this function as much as possible,
we can add the CONFIG_CC_HAS_ASM_GOTO_OUTPUT version as Peter suggests.
I guess this should work:

	u64 test(u64 a, u64 mul, u64 div)
	{
		u64 q;

		asm goto ("mulq %2; 1: divq %3\n"
			_ASM_EXTABLE(1b, %l[fail])
			: "=a" (q)
			: "a" (a), "rm" (mul), "rm" (div)
			: "rdx"
			: fail);

		return q;
	fail:
		// BUG? WARN?
		return -1ul;
	}

I agree with everything ;)

Oleg.

On 07/24, Oleg Nesterov wrote:
>
> On 07/24, Oleg Nesterov wrote:
> > On 07/23, David Laight wrote:
> > >
> > > On Wed, 23 Jul 2025 11:38:25 +0200
> > > Oleg Nesterov <oleg@redhat.com> wrote:
> > > >
> > > > to remove the conditional branch and additional variable. Your version
> > > > is probably beterr... But this is without WARN/BUG.
> > >
> > > I wish there was a way of doing a WARN_ONCE from asm with a single instruction.
> > > Then you could put one after your 2:
> > > Otherwise is it a conditional and a load of inlined code.
> > >
> > > > So, which version do you prefer?
> > >
> > > I wish I knew :-)
> >
> > ;-)
> >
> > David, you understand this asm magic indefinitely better than me. Plus you are
> > working on the generic code. Can you send the patch which looks right to you?
> > I agree in advance with anything you do.
> >
> > I got lost. Now I don't even understand if we want to add BUG and/or WARN into
> > mul_u64_u64_div_u64().
>
> Forgot to mention... Not that I think this is a good idea, but if we don't
> use BUG/WARN, we can probably add EX_FLAG_ and do something like below.
>
> Oleg.
>
> --- a/arch/x86/mm/extable.c
> +++ b/arch/x86/mm/extable.c
> @@ -38,6 +38,9 @@ static bool ex_handler_default(const struct exception_table_entry *e,
>  	if (e->data & EX_FLAG_CLEAR_DX)
>  		regs->dx = 0;
>
> +	if (e->data & EX_FLAG_XXX_AX)
> +		regs->ax = -1ul;
> +
>  	regs->ip = ex_fixup_addr(e);
>  	return true;
>  }