linux-mips
[Top] [All Lists]

Re: [PATCH 2.4] inlines in au1000_eth.c

To: Kevin Turner <kevin.m.turner@pobox.com>
Subject: Re: [PATCH 2.4] inlines in au1000_eth.c
From: "Maciej W. Rozycki" <macro@linux-mips.org>
Date: Mon, 13 Jun 2005 20:47:35 +0100 (BST)
Cc: linux-mips <linux-mips@linux-mips.org>
In-reply-to: <1118438524.1513.28.camel@troglodyte.asianpear>
Original-recipient: rfc822;linux-mips@linux-mips.org
References: <1118438524.1513.28.camel@troglodyte.asianpear>
Sender: linux-mips-bounce@linux-mips.org
On Fri, 10 Jun 2005, Kevin Turner wrote:

> The following patch against linux_2_4 fixes this sort of compilation
> error with gcc 3.4:
> 
> au1000_eth.c: In function `au1000_init_module':
> au1000_eth.c:97: sorry, unimplemented: inlining failed in call to 
> 'str2eaddr': function body not available
> au1000_eth.c:1219: sorry, unimplemented: called from here
> 
> I'm not wholly sure it's the Right Thing, as I define str2eaddr in
> au1000_eth.c, which makes for more code duplication.

 Note that by somebody's "brillant" idea "inline" is a macro that expands 
to "inline __attribute__((always_inline))".  That said, "inline" in a 
function prototype doesn't make much sense anyway.  In this case simply 
removing the qualifier will let code use the implementation from 
arch/mips/au1000/common/prom.c.

 As a side note it's quite amazing how duplicate copies of a function get 
scattered throughout the tree...

  Maciej

<Prev in Thread] Current Thread [Next in Thread>