From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.5) Gecko/20041111 Firefox/1.0 Description of problem: If the kernel is loaded with the options ide-scsi (to use the atapi dvd/cd-writer as scsi device) AND swiotlb (for the nvidia driver), the dvd/cd-drive is not accessible, eg. for mount, eject, cdrecord. Removing either the ide-scsi option or the swiotlb option, the dvd/cdrom-drive works as expected. Version-Release number of selected component (if applicable): 2.4.21-20.EL How reproducible: Always Steps to Reproduce: 1. boot the kernel with eg. hdc=ide-scsi swiotlb=16384 2. mount /dev/cdrom /mnt/cdrom; eject; cdrecord -scanbus 3. remove one of the options ide-scsi or swiotlb and boot again 4. mount /dev/cdrom /mnt/cdrom; eject; Actual Results: the dvd/cdrom-drive is not a valid block device for the mount command, it cannot accessed and used by cdrecord, it cannot be opend oder closed using the command eject Expected Results: opening and closing using the command eject mounting of cds and dvds writing cds using cdrecord Additional info: the option swiotlb is neccessary for the hardware accelerated graphic driver from nvidia.
Believe it or not, if you reverse the order of the kernel parameters, that is, change "hdc=ide-scsi swiotlb=16384" to "swiotlb=16384 hdc=ide-scsi" I believe you will find the problem does not occur. Please confirm. We are working on a proper fix. Thanks. hda=ide-scsi parameter gets mangled before ide_setup gets ahold of it
> Believe it or not, if you reverse the order of the kernel parameters, that is, > > change "hdc=ide-scsi swiotlb=16384" > to "swiotlb=16384 hdc=ide-scsi" I want to believe. ;-) > I believe you will find the problem does not occur. > Please confirm. We are working on a proper fix. Unfortunately, I'm on holidays now, I will check the revers order of the kernel parameters in january 2005.
The change to increase the default swiotlb to 16384 was committed into the RHEL3-U5 pool. Larry Woodman
This problem has been resolved by making swiotlb=16384 the default setting for ia32e (EM64T) platforms. Larry Woodman's patch was committed to the RHEL3 U5 patch pool on 23-Dec-2004 (in kernel version 2.4.21-27.5.EL).
Yep, reversing the order of the kernel parameters works like a charm.
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on the solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHSA-2005-294.html