From: Tejun Heo Subject: Re: long boot delay for SATA cdrom on current 2.6.17-git Date: Fri, 30 Jun 2006 14:26:33 +0900 Message-ID: <20060630052633.GF23844@htj.dyndns.org> References: <20060630025341.GA4675@kroah.com> <20060630044912.GA2592@htj.dyndns.org> <20060630052021.GA4722@kroah.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from nz-out-0102.google.com ([64.233.162.192]:16834 "EHLO nz-out-0102.google.com") by vger.kernel.org with ESMTP id S1751057AbWF3F0J (ORCPT ); Fri, 30 Jun 2006 01:26:09 -0400 Received: by nz-out-0102.google.com with SMTP id z31so128621nzd for ; Thu, 29 Jun 2006 22:26:08 -0700 (PDT) Content-Disposition: inline In-Reply-To: <20060630052021.GA4722@kroah.com> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Greg KH Cc: jgarzik@pobox.com, linux-ide@vger.kernel.org On Thu, Jun 29, 2006 at 10:20:22PM -0700, Greg KH wrote: > On Fri, Jun 30, 2006 at 01:49:12PM +0900, Tejun Heo wrote: > > On Thu, Jun 29, 2006 at 07:53:41PM -0700, Greg KH wrote: > > > This laptop does not have a SATA hard drive, only a SATA cdrom drive. > > > I've included the boot log below and trimmed it after the delay > > > happened, as I doubt anything after it would be relevant. > > > > > > Is there any patch I can test to help fix this? > > > > This seems to be the ghost device problem which occurs when the first > > device is ATAPI but the second slot is empty. > > > > > My .config is also attached, if that helps any. > > > > Please try the following patch. > > Yeah, that worked great, no timeout issues anymore! Do you want to see > the boot log again from this? Great. :-) I don't need the successful boot log but I want to see why it wasn't a problem in 2.6.17. Can you apply the following patch to 2.6.17 and post the result? Thanks. diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c index b046ffa..35f4721 100644 --- a/drivers/scsi/libata-core.c +++ b/drivers/scsi/libata-core.c @@ -2193,7 +2193,7 @@ int ata_std_softreset(struct ata_port *a unsigned int devmask = 0, err_mask; u8 err; - DPRINTK("ENTER\n"); + DPRINTK("ENTER ata%u\n", ap->id); if (ap->ops->scr_read && !sata_dev_present(ap)) { classes[0] = ATA_DEV_NONE; diff --git a/include/linux/libata.h b/include/linux/libata.h index b80d2e7..b6b21f6 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -38,7 +38,7 @@ #include * compile-time options: to be removed as soon as all the drivers are * converted to the new debugging mechanism */ -#undef ATA_DEBUG /* debugging output */ +#define ATA_DEBUG /* debugging output */ #undef ATA_VERBOSE_DEBUG /* yet more debugging output */ #undef ATA_IRQ_TRAP /* define to ack screaming irqs */ #undef ATA_NDEBUG /* define to disable quick runtime checks */