Re: [avr-libc-dev] [RFC] New eeprom.h

[Rick Altherr]

On Feb 29, 2008, at 12:32 PM, Shaun Jackman wrote:

> On Fri, Feb 29, 2008 at 11:46 AM, Rick Altherr <address@hidden>  
> wrote:
> > It was on the mailing list.  There are only one of each, but they are
> > inline functions defined in the header so they will be correct for
> > every device.  The address of that inline function is passed to the
> > library routines.  That will cause a single copy of the inline
> > function to be generated as a true function in the binary so the
> > address of it can be found.
>
> Not exactly. It will generate a single copy of the inline function per
> object file in which it's used -- unless very tricksy linker magic is
> used, which is possible, I believe, but not typical. Still, I think
> one copy of eeprom_read_byte and eeprom_write_byte per object file is
> a reasonable compromise considering the technical challenges.
>
> Cheers,
> Shaun


Many linkers will coalesce copies of the same function from multiple  
object files.  I'm not familiar with binutils enough to know if it  
does or not.

--
Rick Altherr
address@hidden

"He said he hadn't had a byte in three days. I had a short, so I split  
it with him."
 -- Slashdot signature

smime.p7s
Description: S/MIME cryptographic signature