From mboxrd@z Thu Jan 1 00:00:00 1970 From: Kyle McMartin Subject: [PATCH] r8169: initialize rtl8169_stats seqlock Date: Mon, 24 Feb 2014 20:12:28 -0500 Message-ID: <20140225011228.GC14970@merlin.infradead.org> References: <20140224231748.GA18598@pd.tnic> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Realtek linux nic maintainers , Francois Romieu , netdev@vger.kernel.org, linux-kernel@vger.kernel.org To: Borislav Petkov Return-path: Received: from merlin.infradead.org ([205.233.59.134]:55856 "EHLO merlin.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752161AbaBYBMg (ORCPT ); Mon, 24 Feb 2014 20:12:36 -0500 Content-Disposition: inline In-Reply-To: <20140224231748.GA18598@pd.tnic> Sender: netdev-owner@vger.kernel.org List-ID: From: Kyle McMartin Boris reports he's seeing: > [ 9.195943] INFO: trying to register non-static key. > [ 9.196031] the code is fine but needs lockdep annotation. > [ 9.196031] turning off the locking correctness validator. > [ 9.196031] CPU: 1 PID: 933 Comm: modprobe Not tainted 3.14.0-rc4+ #1 with the r8169 driver. These are occuring because the seqcount embedded in u64_stats_sync on 32-bit SMP is uninitialized which is making lockdep unhappy. Signed-off-by: Kyle McMartin --- Patch is only compile tested by me, as these initializers will disappear on 64-bit. --- a/drivers/net/ethernet/realtek/r8169.c +++ b/drivers/net/ethernet/realtek/r8169.c @@ -7118,6 +7118,8 @@ rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) } mutex_init(&tp->wk.mutex); + u64_stats_init(&tp->rx_stats.syncp); + u64_stats_init(&tp->tx_stats.syncp); /* Get MAC address */ for (i = 0; i < ETH_ALEN; i++)