xcdroast - cdrecord.mmap: Input/output error. write_g1: scsi sendcmd: cmd timeout
 

Hello everyone,

I am running into trouble burning a large (approximately 630MB) binary iso. I am running Debian Testing, Kernel version 2.4.25. I have successfully burned using xcdroast on this machine (even other large iso files ~600MB), however, not sure what is going on here and have not been able to track anything down. Essentially the write process hangs/times-out with the following error messages:

cdrecord.mmap: faio_wait_on_buffer for writer timed out.
cdrecord.mmap: Input/output error. write_g1: scsi sendcmd: cmd timeout after 399.990 (200) s
CDB: 2A 00 00 00 28 EE 00 00 1F 00
cmd finished after 399.990s timeout 200s
cdrecord.mmap: A write error occured.
cdrecord.mmap: Please properly read the error message above.


Here is the compelete CD out from xcdroast:

Calling: /usr/lib/xcdroast/bin/xcdrwrap CDRECORD dev= "1,0,0" gracetime=2 fs=4096k -v -useinfo speed=6 -dao -dummy -eject -pad -immed -data "/var/cdimages/yarrow-i386-disc1.iso" ...

scsidev: '1,0,0'
scsibus: 1 target: 0 lun: 0
Linux sg driver version: 3.1.25
SCSI buffer size: 64512
Cdrecord-Clone 2.01a25 (i686-pc-linux-gnu) Copyright (C) 1995-2004 Jörg Schilling
NOTE: this version of cdrecord is an inofficial (modified) release of cdrecord
and thus may have bugs that are not present in the original version.
Please send bug reports and support requests to <cdrtools@packages.debian.org>.
The original author should not be bothered with problems of this version.

TOC Type: 1 = CD-ROM
Using libscg version 'schily-0.8'.
atapi: 0
Device type : Removable CD-ROM
Version : 2
Response Format: 2
Capabilities : SYNC
Vendor_info : 'YAMAHA '
Identifikation : 'CRW6416S '
Revision : '1.0d'
Device seems to be: Generic mmc CD-RW.
Using generic SCSI-3/mmc CD-R/CD-RW driver (mmc_cdr).
Driver flags : MMC SWABAUDIO
Supported modes: TAO PACKET SAO SAO/R96P SAO/R96R RAW/R96R
Drive buf size : 1566432 = 1529 KB
FIFO size : 4194304 = 4096 KB
pregap1: -1
Track 01: data 629 MB padsize: 30 KB
Total size: 723 MB (71:39.29) = 322447 sectors
Lout start: 723 MB (71:41/22) = 322447 sectors
Current Secsize: 2048
ATIP info from disk:
Indicated writing power: 4
Is not unrestricted
Is not erasable
Disk sub type: Medium Type B, low Beta category (B-) (4)
ATIP start of lead in: -11753 (97:25/22)
ATIP start of lead out: 335100 (74:30/00)
Disk type: Long strategy type (Cyanine, AZO or similar)
Manuf. index: 8
Manufacturer: Hitachi Maxell, Ltd.
Blocks total: 335100 Blocks current: 335100 Blocks remaining: 12653
Starting to write CD/DVD at speed 6 in dummy SAO mode for single session.
Waiting for reader process to fill input buffer ...
input buffer ready.
Sending CUE sheet...
Writing pregap for track 1 at -150
Starting new track at sector: 0
cdrecord.mmap: faio_wait_on_buffer for writer timed out.
cdrecord.mmap: Input/output error. write_g1: scsi sendcmd: cmd timeout after 399.990 (200) s
CDB: 2A 00 00 00 28 EE 00 00 1F 00
cmd finished after 399.990s timeout 200s
cdrecord.mmap: A write error occured.
cdrecord.mmap: Please properly read the error message above.
cdrecord.mmap: Caught interrupt.

cdrecord.mmap: Caught interrupt.
write track data: error after 21458944 bytes
cdrecord.mmap: Second interrupt. Doing hard abort.
Writing time: 686.679s
cdrecord.mmap: fifo had 402 puts and 339 gets.
Min drive buffer fill was 70%
cdrecord.mmap: fifo was 0 times empty and 331 times full, min fill was 96%.

Any ideas, direction, guidance would be much appreciated.

Thanks.

Dunno if it helps much - but I got a similar error when my DVD burner didn't like the media - switched the media brand and it's been fine so far.

Has it happened more than once - could be media.

Is it always happening at the same place - might not be media - bad media would tend to be more random?

Thanks for your reply.

It is definitely random. It hangs/timesout anywhere from 2% to 24% completion. After my third wasted CD, I started running/testing burns in simulation mode and even in simulation mode, when it isn't actually writring, it hangs/timesout. Would the simulation mode be affected by the media too?

Actually, one other note is that I have successfuly been able to burn on the same media brand/type before, however, not successful with large binary ISOs.

Thanks.

That sound similar to my experience - I/O error during the first 25% of a burn.

But in your case if the media worked before - perhaps the drive is on the way out.

Have you ugraded your kernel or anything else?

Have you tried the direct ATAPI option of cdrecord:
Apparently ide-scsi is on the way out.

Is the burner getting too hot. After problems burning one DVD-R, I immediately tried simulation burns, and then real burns - eventually the kernel reported a servo motor failure and the burner stopped working totally - wouldn't even eject. I opened up the case and found the burner to be quite warm. After letting it cool down it was OK again. I've since moved it so that there is an empty bay above and below it. If you suspect temperature, do a burn with extra cooling, for example: remove the side panel and point a desk-fan at the interior.

Do smaller burns still work?

Do slower burns work? Force the speed down with a command line option.

I think it still might be worth trying a different brand of media - be careful that it isn't the same media rebranded - if the color of the recording surface looks different, then it must be made by a different plant. I'm not sure what a reliable brand is - reitek? verbatim? maxel?

Do you have one of those other OS's that you could try a burn on? Or temporarily move the burner to another machine.

Googling for "faio_wait_on_buffer for writer timed out" yields many hits - perhaps one of these might be relevant.

Are you using 650 Mg blanks or 700Mg blanks? I'm no expert on analyzing xcdroast errors, but if a ~600Mg disk burns OK but a ~630 one doesn't you might just be running out of physical space. Note also that about a third of the way down, the "total size" shows as 723Mg -- who knows how accurate/precise that figure may be, but obviously you'd have a far better chance of fitting it on a 700Mg blank than a 650Mg blank.

I'll also echo michael@actrix's comments: other likely causes are that you aren't using quality media, or that your burn speed is higher than what your CD can handle, or that the CD itself is flaking out and is on its last legs. Personally, I have been using TDK 700 blanks with very good results, and my burn speed is set to 8X even though my CD supposedly can handle 24X. After making a nice set of coasters for myself at 24X and 16X, things finally got happy at 8X. Burning is a one time cost so to speak, so at least for me, whether the burn process takes 4 minutes or 10 minutes is pretty much irrelevant. -- J.W.

Thanks for your reply/feedback guys.

I just upgraded my kernel to 2.4.25. Prior to the upgrade I had not tried burning on my Linux box. This box used to be running one of those other OS's and it worked fine. The series of tests I have performed so far include: attempting at slower burn rates (4X instead of 6X), and trying different size media (both 650MB and 700MB). The only burns I have succeeded with are around 600MB (on the same 650MB media) in size. Nothing smaller than that yet. Having said that, in this instance, my attempts with this large iso are even failing in simulation mode. It is not even getting to the actual burn process. The media type is not even coming into play yet. Part of the reason why I am at a total loss.

I will try your other suggestions tonight. Hopefully it will lead down to some other ideas to try.

Thanks guys. Your feedback/ideas is much appreciated.

you may want to try recompiling cdrecord (or if it isn't available standalone, compile cdrtools)

k3b is also a good frontend for burning if you have kde installed; it uses different commands and may help with burning

Under 2.4.x, I believe you need scsi emulation to get the CD-RW to work properly, and perhaps on your PC it has not yet been set up. (At least in Slack you do, dunno about Debian but I'd bet it would be the same). Two quick questions:

1. Assuming you use lilo, is this line present: append="hdc=ide-scsi"
2. Is your CD listed when you run this command (it should be): cat /proc/scsi/scsi

I can tell you what to do to enable scsi emulation under Slack, but as a non-Debian user I'm a little reluctant to offer advice on what steps are needed, because it might be wrong. However, presumably there are other threads that do describe it for Debian, and I'd recommend taking advantage of the Search function here at LQ. Let us know how it goes and good luck -- J.W.

Hello everyone,

Thank you to all for your feedback. I am amazed by the generosity of feedback and ideas on this forum.

Anyway, I thought I got close tonight, but unfortunately, no cigar. Here are the results from tonight's tests:

1. Tried a different media type (Maxell Black) and size (700MB) - failed
2. Tried K3B - Great application. It actually got my hopes up as it managed to get up to 80% complete before it froze. - failed
3. Ran cat /proc/scsi/scsi - my CD drive does show up
4. Checked google - there are a few people with IDE-SCSI related issues. However, none that match my symptoms exactly.

Next, I am going to try cdrecord in command line. Fingers crossed.

Thanks.

Just spotted a note in the SCSI howto -

ide-scsi problems in 2.4

Hi michael@actrix,

Thanks for your latest guidance. The information helped me understand what was going on. However, I am still having the same problem. Now the percentage complete is getting to 60%-80% instead of 20%. I will keep at it.... I am actually thinking of trying my luck with Fedora. I will let you guys know how it goes.

Regards,

masgari

Hello everyone,

Final post on this one...

Just wanted to let you guys know I was never successful in solving this problem under Debian. For reasons not necessarily related to this issue, I switched over to Fedora Core one and after some adjustments/configuration, all is good. I have successfully burned very large ISOs using K3B.

Thanks to all who advised and invoked ideas.

masgari