Description of problem: Having built an iso fs with mkisofs to provide both rockridge, joliet and hfs support, a loop back mount using hfs works fine. However burning to cdrom and mounting -t hfs /dev/cdrom causes kernel BUG report. CD works fine in iso9660 mode. Version-Release number of selected component (if applicable): Linux jenjen3.localdomain 2.4.20-20.9 #1 Mon Aug 18 11:27:43 EDT 2003 i686 athlon i386 GNU/Linux How reproducible: Always Steps to Reproduce: 1. Build isofs with: mkisofs -r -J -l -D -L -V volid -hfs --macbin -o /home/tmp/dual-hfs-01.img -graft-points /=/home/cmiles/dual 2. Check with loop back mount -t hfs /home/tmp/dual-hfs-01.img /mnt/wrk -o loop,ro 3. Burn CD-R using xcdroast Wed Sep 24 13:58:36 2003 XCDR 0.98alpha13: Executing: /usr/lib/xcdroast-0.98/bin/xcdrwrap CDRECORD dev=1,0,0 gracetime=2 fs=8192k driveropts=burnfree -v -useinfo speed=40 -dao -eject -pad -data "/home/tmp/dual-hfs-01.img" 4. Mount CD defaults to iso9660 - no problems 5. Mount CD as type hfs gives log entry: Sep 24 15:47:04 jenjen3 kernel: kernel BUG at buffer.c:2618! Sep 24 15:47:04 jenjen3 kernel: invalid operand: 0000 Actual results: kernel bug causes all subsequent mounts to hang; shutdown also hangs on unmount; manual restart has to recover disks. Expected results: Access to mounted files from the CD as were available from the image looped back. Additional info: The CD has a large number of files, quite a number with long file names.
Created attachment 94684 [details] system log for kernel BUG at buffer.c:2618!
I have rebuilt the iso image with '-part' but to the same effect. However I have found (by mistake) that 'mount -t hfs /dev/cdrom /mnt/wrk -o loop' works. Strange?
The CD reads fine in a Mac OS 9.1 machine (only the without -part was tested).
Repeatable without all the proprietory binary modules loaded ?
*** This bug has been marked as a duplicate of 78616 ***
Created attachment 94830 [details] untainted dmesg output My apologies for submitting a tainted report. I have tested without the tainted modules and the kernel BUG still occurs. I find that the fault still occurs in single user mode. The dmesg output is attached The fault does not occur when the kernel boot parameter 'hdc=ide-scsi' is removed. Full X11 and running the application off the HFS mounted CD rom is then available.
This is repeatable without all the proprietory binary modules loaded.
Please provide a copy of your /var/log/messages from a full boot, as well as the full output from "lsmod > lsmod.txt". If you're using X on this machine, I have an idea also, please attach your /var/log/XFree86.0.log as I'd like to compare it with another bug report received.
Created attachment 94839 [details] Zip of messages, lsmod and XFree log The zip contains the requested files from a full X11 boot and hfs mount segfault. I find that the fault still occurs in single user mode. The fault does not occur if the kernel boot parameter 'hdc=ide-scsi' is removed.
Created attachment 95095 [details] lsmod after clean boot and CD mount -t hfs
Created attachment 95096 [details] messages from clean boot and mount -t hfs
Created attachment 95097 [details] XFree86.0.log from clean boot and mount -t hfs Same files as in the previous Zip attachment - easier to read in browser.
Please let me know if I can provide any more information or investigate further.
Thanks for the bug report. However, Red Hat no longer maintains this version of the product. Please upgrade to the latest version and open a new bug if the problem persists. The Fedora Legacy project (http://fedoralegacy.org/) maintains some older releases, and if you believe this bug is interesting to them, please report the problem in the bug tracker at: http://bugzilla.fedora.us/