Bug 184249

Summary: K3b detects USB-connected DVD-Writer as read-only
Product: [openSUSE] SUSE Linux 10.1 Reporter: Christian Mügge <linux>
Component: KernelAssignee: Jeff Mahoney <jeffm>
Status: RESOLVED WONTFIX QA Contact: E-mail List <qa-bugs>
Severity: Normal    
Priority: P5 - None CC: jeffm, suse-beta, trueg
Version: Final   
Target Milestone: ---   
Hardware: x86   
OS: SuSE Linux 10.1   
Whiteboard:
Found By: Customer Services Priority:
Business Priority: Blocker: ---
Marketing QA Status: --- IT Deployment: ---
Bug Depends on: 182768    
Bug Blocks:    
Attachments: cdrecord -scanbus
sginfo -l
Log from K3b
New test with k3b - Version 0.12.17
dmesg-output

Description Christian Mügge 2006-06-12 23:43:37 UTC
I can't burn ISO-Images with K3b because the burner is recognized as Removable CD-ROM-Device.
Comment 1 Christian Mügge 2006-06-12 23:50:45 UTC
It's a brand from LG/Model GSA-4167B connected with a USB 2.0 Case to a Notebook
Comment 2 Christian Mügge 2006-06-12 23:55:34 UTC
Created attachment 88802 [details]
cdrecord -scanbus
Comment 3 Christian Mügge 2006-06-12 23:57:19 UTC
Created attachment 88803 [details]
sginfo -l
Comment 4 Christian Mügge 2006-06-13 00:02:33 UTC
Created attachment 88804 [details]
Log from K3b

This is the last try to burn with the Version 0.12.15 from the Packman-repository. But it's the same Problem that comes with the served Version from SuSE 10.1
Comment 5 Christian Mügge 2006-06-13 00:17:37 UTC
I used the same DVD-Writer in SuSE 10.0 and it works great.
Comment 6 Dirk Mueller 2006-08-16 15:52:10 UTC
hmm, kernel issue ?! please assign back if not. 
Comment 8 Christian Mügge 2006-08-20 13:59:46 UTC
dmesg output says:

USB Mass Storage support registered.
  Vendor: HL-DT-ST  Model: DVDRAM GSA-4167B  Rev: DL13
  Type:   CD-ROM                             ANSI SCSI revision: 00
usb-storage: device scan complete
sr0: scsi3-mmc drive: 48x/48x writer dvd-ram cd/rw xa/form2 cdda tray
Uniform CD-ROM driver Revision: 3.20
sr 0:0:0:0: Attached scsi CD-ROM sr0
sr 0:0:0:0: Attached scsi generic sg0 type 5



Kernel is 2.6.16.21-0.13-default

In the meanwhile I get K3b Version 0.12.16 but it still not work.
Comment 9 Stephan Kulow 2006-09-13 14:40:03 UTC
k3b 0.12.17 is reported to work correctly (#182768). Can you try?
Comment 10 Stephan Binner 2006-09-13 14:47:35 UTC
> In the meanwhile I get K3b Version 0.12.16 but it still not work

That's strange as #182768 reports that 0.12.17 works and according to changelog it should be fixed in 0.12.15+.
Comment 11 Christian Mügge 2006-09-13 15:53:59 UTC
Created attachment 98609 [details]
New test with k3b - Version 0.12.17

This is the test from a newer k3b-Version.
But it's the same problem. I can't burn a CD-RW/DVD-RW medium.
Comment 12 Christian Mügge 2006-09-13 16:04:12 UTC
Created attachment 98613 [details]
dmesg-output

And here is the dmesg-output first with a cd-rw and last with a dvd-rw medium.
Comment 13 Hannes Reinecke 2007-01-08 14:20:05 UTC
I suspect this is the error:

sr0: scsi3-mmc drive: 48x/48x writer dvd-ram cd/rw xa/form2 cdda tray
Uniform CD-ROM driver Revision: 3.20
sr 0:0:0:0: Attached scsi CD-ROM sr0
sr 0:0:0:0: Attached scsi generic sg0 type 5
ISO 9660 Extensions: Microsoft Joliet Level 3
ISOFS: changing to secondary root 
ISO 9660 Extensions: Microsoft Joliet Level 3
ISO 9660 Extensions: RRIP_1991A
VFS: busy inodes on changed media.
VFS: busy inodes on changed media.
ISO 9660 Extensions: Microsoft Joliet Level 3
ISO 9660 Extensions: RRIP_1991A
sr 0:0:0:0: Device not ready.
end_request: I/O error, dev sr0, sector 120
VFS: busy inodes on changed media.

So it looks as if the 'changing to secondary root' also implies a media change. Somehow. Curious.
Comment 14 Hannes Reinecke 2007-01-26 13:18:37 UTC
Jeffm, could that be a VFS issue?
Comment 15 Sebastian Trueg 2007-02-16 10:05:07 UTC
Does this mean that the original issue is resolved? It sure seems so if you can start burning in K3b.
Comment 16 Jeff Mahoney 2007-02-20 20:01:36 UTC
The secondary root message doesn't imply a media change. The comment in the code indicates that the CD has both Joliet and Rock Ridge, and it's using Rock Ridge by default. The messages posted also back that up.

The "busy inodes on changed media" messages are cause for concern, but I don't believe they're related to the problem here.

This seems to be the problem:
/usr/bin/cdrecord: Cannot allocate memory. read buffer: scsi sendcmd: no error
CDB:  3C 00 00 00 00 00 00 FC 00 00
status: 0x0 (GOOD STATUS)
cmd finished after 0.000s timeout 40s
[...]
/usr/bin/cdrecord: Warning: controller returns zero sized CD write parameter page.
/usr/bin/cdrecord: Warning: controller returns wrong page 0 for CD write parameter page (5).
/usr/bin/cdrecord: Warning: controller returns zero sized CD write parameter page.
/usr/bin/cdrecord: Warning: controller returns wrong page 0 for CD write parameter page (5).
/usr/bin/cdrecord: Cannot init drive.

I don't know if it's a bug in cdrecord, a bug in the kernel's sg interface, or potentially faulty hardware. Either way, k3b is just calling out to cdrecord to do the heavy lifting, so it's not a bug there.

Christan, I'd like to gather some more information here.

Can you run "strace -o cdrecord.strace /usr/bin/cdrecord -v gracetime=2 dev=0,0,0 speed=10 -tao driveropts=burnfree -eject -overburn blank=fast -force" and send me both the output of cdrecord and the cdrecord.strace file? 

It looks as though it could be either trying to find the buffer capacity or the dma speed. The strace output will let me see if the -ENOMEM is coming from the kernel or from cdrecord itself. The output from cdrecord will also be more verbose than previous logs.
Comment 17 Jeff Mahoney 2007-03-21 17:27:03 UTC
Closing due to inactivity. If you'd care to provide the requested information, please REOPEN.