From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeremy Fitzhardinge Subject: Re: [PATCH] parisc: fix module loading failure of large kernel modules (take 4) Date: Thu, 01 Jan 2009 18:12:41 +1100 Message-ID: <495C6CE9.3080402@goop.org> References: <4959346E.7060600@gmx.de> <200812310915.41693.rusty@rustcorp.com.au> <495B5806.3080505@gmx.de> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Cc: Helge Deller , Rusty Russell , linux-parisc , Linux Kernel Development , Kyle McMartin , Randolph Chung , Andrew Morton , Sam Ravnborg , John David Anglin To: Linus Torvalds Return-path: In-Reply-To: List-ID: List-Id: linux-parisc.vger.kernel.org Linus Torvalds wrote: > Some gcc versions will inline weak functions if they are in scope - even > if there is a non-weak function somewhere else. So you MUST NOT have the > weak definition in the same file (or indirectly called through some inline > functions in a header file) as the call. Because if you do, then any user > with the wrong version of gcc will get the weak function semantics, even > if it was meant to be overridden by something else. > Yes. I think this behaviour is considered to be desperately broken by the gcc developers and has been fixed in all recent gccs. There are a couple of broken versions, and there have been patches floating around to just refuse to use them; otherwise __weak is effectively unusable. (Ah, I see Adrian has posted it again and everyone agrees with it.) On the other hand, I have seen a couple of instances of "inline __weak" which is insane, but I don't know if gcc does anything crazy with it, or if its common enough to bother warning about. J