From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Brownell Subject: Re: [Bulk] Re: [PATCH 1/2] [PCI] Check that MWI bit really did get set Date: Sun, 15 Oct 2006 17:21:53 -0700 Message-ID: <200610151721.54873.david-b@pacbell.net> References: <1160161519800-git-send-email-matthew@wil.cx> <20061015104544.5de31608.akpm@osdl.org> <17714.52121.962807.781244@cargo.ozlabs.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Cc: Andrew Morton , Matthew Wilcox , Alan Cox , val_henson@linux.intel.com, netdev@vger.kernel.org, linux-pci@atrey.karlin.mff.cuni.cz, linux-kernel@vger.kernel.org, gregkh@suse.de Return-path: Received: from smtp105.sbc.mail.mud.yahoo.com ([68.142.198.204]:54443 "HELO smtp105.sbc.mail.mud.yahoo.com") by vger.kernel.org with SMTP id S1161184AbWJPAV6 (ORCPT ); Sun, 15 Oct 2006 20:21:58 -0400 To: Paul Mackerras In-Reply-To: <17714.52121.962807.781244@cargo.ozlabs.ibm.com> Content-Disposition: inline Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org > > If the drivers doesn't care and if it makes no difference to performance > > then just delete the call to pci_set_mwi(). > > > > But if MWI _does_ make a difference to performance then we should tell > > someone that it isn't working rather than silently misbehaving? To repeat: it's not "misbehaving" since support for MWI cycles is completely optional. It'd be more interesting to get messages in the cases that it can be enabled, since typically it can't be. > That sounds like we need a printk inside pci_set_mwi then, rather than > adding a printk to every single caller. Maybe wrapped in #ifdef CONFIG_SPAM_MY_KERNEL_LOG_MESSAGES ... :)