Re: [PATCH 1/many] PROC macro to annotate functions in assembly files

["Alexander van Heukelum" <heukelum@fastmail.fm>]

On Thu, 18 Dec 2008 10:07:47 +0000, "Russell King"
<rmk+lkml@arm.linux.org.uk> said:
> On Thu, Dec 18, 2008 at 10:51:58AM +0100, Alexander van Heukelum wrote:
> > Agreed. I vote to complement the existing ENDPROC annotation with
> > the proposed PROC annotation. Let's call that an extension, not
> > something new ;). As it stands it is not impossible to go with
> > ENTRY/ENDPROC for code and ENTRY/END for data. However, ENTRY
> > implies alignment and the prefered alignment for code and data
> > might differ.
> 
> Have you looked at the number of ENTRY uses for code vs for data?
> If all you're after is separating the two uses, then it might be a
> smaller patch to change the ENTRY use for data rather than changing
> all the ENTRY uses for code.
> 
> There are 589 uses of ENTRY in arch/arm/*/*.S.  Of those about 50
> aren't called code.

Hi,

Things are similar for x86, but I didn't consider it a problem.

The alternative I see is to is to introduce DATAENTRY and DATAEND
for use with data objects in generic code, equal to ENTRY/END. Then
deprecate the use of ENDPROC, so we can try to get rid of it in the
long run.

Minor nit is that all archs need to override ENTRY and/or END to
include an assembly directive that indicates that the symbol is
a function. Changing this in generic code is not possible as long
as there are ARCHs which have not been converted.

ENDPROC might stick for a very long time, though.

Greetings,
    Alexander

> -- 
> Russell King
>  Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
>  maintainer of:
-- 
  Alexander van Heukelum
  heukelum@fastmail.fm

-- 
http://www.fastmail.fm - A fast, anti-spam email service.