From mboxrd@z Thu Jan 1 00:00:00 1970 From: "John W. Linville" Subject: Re: linux-next: wireless tree build failure Date: Fri, 13 Feb 2009 13:24:56 -0500 Message-ID: <20090213182456.GA3581@tuxdriver.com> References: <20090212193517.3f393726.sfr@canb.auug.org.au> <20090212161639.GA4091@tuxdriver.com> <4995BADE.60700@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from charlotte.tuxdriver.com ([70.61.120.58]:55903 "EHLO smtp.tuxdriver.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752466AbZBMSay (ORCPT ); Fri, 13 Feb 2009 13:30:54 -0500 Content-Disposition: inline In-Reply-To: <4995BADE.60700@gmail.com> Sender: linux-next-owner@vger.kernel.org List-ID: To: Dave Cc: Stephen Rothwell , linux-next@vger.kernel.org On Fri, Feb 13, 2009 at 06:24:30PM +0000, Dave wrote: > John W. Linville wrote: > > On Thu, Feb 12, 2009 at 07:35:17PM +1100, Stephen Rothwell wrote: > >> Today's linux-next build (powerpc allyesconfig) failed like this: > >> > >> net/built-in.o: In function `michael_mic': > >> (.opd+0x3ba78): multiple definition of `michael_mic' > >> drivers/built-in.o:(.opd+0x552a8): first defined here > >> net/built-in.o: In function `michael_mic': > >> net/mac80211/michael.c:58: multiple definition of `.michael_mic' > >> drivers/built-in.o:drivers/net/wireless/orinoco/mic.c:49: first defined here > >> > >> Immediate cause is commit 84875201dd1150dc2c16780b944fe501d588ffba > >> ("orinoco: Move MIC helpers into new file"). > > > > I'll fix this up with a patch like below... > > > > John > > I guess it only worked for me because I'm building modules. > > Anyway, the fix needs to be slightly different: > > > diff --git a/drivers/net/wireless/orinoco/mic.c b/drivers/net/wireless/orinoco/mic.c > > index c39d9ea..8cdac75 100644 > > --- a/drivers/net/wireless/orinoco/mic.c > > +++ b/drivers/net/wireless/orinoco/mic.c > > @@ -16,18 +16,18 @@ > > /********************************************************************/ > > int orinoco_mic_init(struct orinoco_private *priv) > > { > > - priv->tx_tfm_mic = crypto_alloc_hash("michael_mic", 0, 0); > > + priv->tx_tfm_mic = crypto_alloc_hash("orinoco_mic", 0, 0); > > That wants to remain as michael_mic, so we get the right crypto alg. > > > if (IS_ERR(priv->tx_tfm_mic)) { > > printk(KERN_DEBUG "orinoco_mic_init: could not allocate " > > - "crypto API michael_mic\n"); > > + "crypto API orinoco_mic\n"); > > Since we're referring to the crypto alg, I suggest this stay the same. > > > priv->tx_tfm_mic = NULL; > > return -ENOMEM; > > } > > > > - priv->rx_tfm_mic = crypto_alloc_hash("michael_mic", 0, 0); > > + priv->rx_tfm_mic = crypto_alloc_hash("orinoco_mic", 0, 0); > > Ditto 1. > > > if (IS_ERR(priv->rx_tfm_mic)) { > > printk(KERN_DEBUG "orinoco_mic_init: could not allocate " > > - "crypto API michael_mic\n"); > > + "crypto API orinoco_mic\n"); > > Ditto 2. > > The rest is fine. Right, Johannes already scolded me. Sloppy search-n-replace... :-) John -- John W. Linville Someday the world will need a hero, and you linville@tuxdriver.com might be all we have. Be ready.