Bug 102510

Summary: defective cd's hang system
Product: [openSUSE] SUSE Linux 10.1 Reporter: Dirk Mueller <dmueller>
Component: KernelAssignee: Greg Kroah-Hartman <gregkh>
Status: RESOLVED WONTFIX QA Contact: E-mail List <qa-bugs>
Severity: Normal    
Priority: P5 - None CC: isaac.charles, kernel01
Version: Final   
Target Milestone: ---   
Hardware: Other   
OS: All   
Whiteboard:
Found By: Other Services Priority:
Business Priority: Blocker: ---
Marketing QA Status: --- IT Deployment: ---

Description Dirk Mueller 2005-08-06 10:44:54 UTC
while trying to read cd's with media defects using an ATAPI dvd-rom drive, the 
kernel keeps falling back to pio mode and doing atapi resets, even though that 
doesn't help anything with the media defects.  
 
The resulting system is almost unusable, it hangs for 5-10s while you can do 
nothing. its not even possible to clikc on the "stop" button or to ctrl-c the 
cp process because the system is entirely locked up.  
 
This was already fixed in the 2.4 kernel but it seems the fix for this was 
(again) lost.  
 
dmesg contains something like:  
ide: failed opcode was: unknown 
hdb: cdrom_decode_status: status=0x51 { DriveReady SeekComplete Error } 
hdb: cdrom_decode_status: error=0x40 { LastFailedSense=0x04 } 
ide: failed opcode was: unknown 
hdb: DMA disabled 
hdb: ide_intr: huh? expected NULL handler on exit 
hdb: ATAPI reset complete 
hdb: cdrom_decode_status: status=0x51 { DriveReady SeekComplete Error } 
hdb: cdrom_decode_status: error=0x40 { LastFailedSense=0x04 } 
ide: failed opcode was: unknown 
hdb: cdrom_decode_status: status=0x51 { DriveReady SeekComplete Error } 
hdb: cdrom_decode_status: error=0x40 { LastFailedSense=0x04 } 
ide: failed opcode was: unknown 
hdb: cdrom_decode_status: status=0x51 { DriveReady SeekComplete Error } 
hdb: cdrom_decode_status: error=0x40 { LastFailedSense=0x04 } 
ide: failed opcode was: unknown 
hdb: cdrom_decode_status: status=0x51 { DriveReady SeekComplete Error } 
hdb: cdrom_decode_status: error=0x40 { LastFailedSense=0x04 } 
ide: failed opcode was: unknown 
hdb: ide_intr: huh? expected NULL handler on exit 
hdb: ATAPI reset complete
Comment 1 Isaac Charles 2006-01-25 13:47:11 UTC
In my case, the defective medium is the Suse Linux 10.0 (32-bit) install DVD that I downloaded and burned.  Apparently my burn had problems.  It is not hanging the system, it is filling my logs with error messages.  Once per second, /var/log/messages and /var/log/warn receive these messages:

Jan 25 08:30:20 hot kernel: hdc: packet command error: status=0x51 { DriveReady SeekComplete Error }
Jan 25 08:30:20 hot kernel: hdc: packet command error: error=0x54 { AbortedCommand LastFailedSense=0x05 }
Jan 25 08:30:20 hot kernel: ide: failed opcode was: unknown

A little math will show that at 264 bytes per second, that's 21 megs per day, which seems really unreasonable.

It needs to limit its retries.

Comment 2 Jens Axboe 2006-09-12 12:51:43 UTC
Reopen for 10.1 if neccessary.
Comment 3 Dirk Mueller 2006-09-12 13:22:56 UTC
its not as if it would be anywhere near fixed..
Comment 4 Greg Kroah-Hartman 2006-09-12 13:29:32 UTC
This is not a problem, you have broken media, the kernel is reporting it.
So, there is not a bug here.
Comment 5 Dirk Mueller 2006-09-12 13:42:48 UTC
I agree. the problem is that I can't move the mouse while it tries to access the media and the only way to recover from the situation is to either forcefully remove the media or reboot
Comment 6 Greg Kroah-Hartman 2006-12-19 23:17:13 UTC
Not much we can do about broken hardware and media here.  If this still really bugs you, and is an issue for the 10.2 release, I suggest you open a bug at bugzilla.kernel.org and let the upstream kernel developers know about this issue, as they are the ones that would be fixing it.