From mboxrd@z Thu Jan 1 00:00:00 1970 From: Matthew Garrett Subject: Re: intel_scu_ipc: Fix to overcome late initialization of scu ipc driver Date: Sun, 30 May 2010 18:43:52 +0100 Message-ID: <20100530174352.GA2354@srcf.ucam.org> References: <0AE3E14D83C76F4994657326177D1FF9042CE9F774@bgsmsx501.gar.corp.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from cavan.codon.org.uk ([93.93.128.6]:38260 "EHLO cavan.codon.org.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751740Ab0E3RoA (ORCPT ); Sun, 30 May 2010 13:44:00 -0400 Content-Disposition: inline In-Reply-To: <0AE3E14D83C76F4994657326177D1FF9042CE9F774@bgsmsx501.gar.corp.intel.com> Sender: platform-driver-x86-owner@vger.kernel.org List-ID: To: "Ds, Sreedhara" Cc: "platform-driver-x86@vger.kernel.org" , "x86@kernel.org" , "Cox, Alan" On Sun, May 30, 2010 at 11:10:03PM +0530, Ds, Sreedhara wrote: > -static void __exit intel_scu_ipc_exit(void) > -{ > - pci_unregister_driver(&ipc_driver); > -} > - > MODULE_AUTHOR("Sreedhara DS "); > MODULE_DESCRIPTION("Intel SCU IPC driver"); > MODULE_LICENSE("GPL"); > > -module_init(intel_scu_ipc_init); > -module_exit(intel_scu_ipc_exit); > +fs_initcall(intel_scu_ipc_init); This is wrong. If it can't be a module then delete all the other module code from the driver and change it to a bool in Kconfig. And don't use fs_initcall - subsys or arch would seem more reasonable, but you're not describing the problem terribly clearly. Which drivers are getting initialised before this? -- Matthew Garrett | mjg59@srcf.ucam.org