Skip to Content.
Sympa Menu

sm-discuss - Re: [SM-Discuss] [sm-discuss] Any one familiar with always_inline attribute problems with new gcc 4.7.2?

sm-discuss AT lists.ibiblio.org

Subject: Public SourceMage Discussion List

List archive

Chronological Thread  
  • From: Remko van der Vossen <wich AT yuugen.jp>
  • To: Javier Vasquez <j.e.vasquez.v AT gmail.com>
  • Cc: "sm, discuss" <sm-discuss AT lists.ibiblio.org>
  • Subject: Re: [SM-Discuss] [sm-discuss] Any one familiar with always_inline attribute problems with new gcc 4.7.2?
  • Date: Mon, 11 Mar 2013 08:41:18 +0900

Hi Javier,

On 2013/03/11, at 2:34, Javier Vasquez <j.e.vasquez.v AT gmail.com> wrote:

> Any one familiar with this gcc error?

Nope, but I see what goes wrong anyway.

> I've read several links about always_inline, and they all tend to
> indicate the function itself needs to be inline, for example:
>
> http://stackoverflow.com/questions/13472341/inlining-failed-function-body-can-be-overwritten-at-link-time
>
> However the failing function is actually inline itself:

The problem is that the Makefile that adds the attribute removes the inline
keyword itself:

> -Dinline=__attribute__\ \(\(__always_inline__\)\)


> Any suggested way to overcome this, other than removing the attribute,
> which is supposed to provide some performance benefit?

Just change the makefile to say:

> -Dinline=inline\ __attribute__\ \(\(__always_inline__\)\)

Regards, Remko



Archive powered by MHonArc 2.6.24.

Top of Page