cdrecord woes... anybody want a coaster?

[Randy Kelsoe]

Gabriel M. Beddingfield wrote:

>Friends,
>
>[NOTE:  Please CC me on replies.  Thanks.]
>
>I'm having trouble with cdrecord when recording audio CD's.  During 
>the write process, a buffer underrun will occur and cdrecord will 
>stop and report the error.  This typically happens around the 5th 
>track.  When I try again, it happens almost immediately.  Further, my 
>hard drive continuously cranks away and they system becomes hardly 
>useable.  A 'halt -p' takes about 30 minutes to completely shut down 
>the computer... and that's the only way I've found to stop it.  Upon 
>review, after the first failure the system becomes sluggish.  After 
>the second failure it becomes unusable.
>
>This doesn't appear to happen when recording data disks.  This 
>happened with several brand new CD-R disks (which are now 
>coasters).  This happens at 1x, 4x, and 8x speed.  Happens in -DAO 
>and -TAO.  This happens even in -dummy mode.  Happens when su'ed to 
>root... happens when logged in as root.  I first noticed it when 
>running K3B... but I've since isolated it as a cdrecord problem.
>
>Questions:  Is this happening to anyone else?  STFW yields 
>nothing.  Is it me, or is it a bug?  If a bug in cdrecord -- how do I 
>report it when the Fedora distribution has "altered" it to add "DVD" 
>and so the program spits out the disclaimer "do not notify the author 
>of problems with this version?"
>
>  
>

I just tried to burn an audio cd with k3b for the first time since 
upgrading cdrecord and k3b. After burning the cd, I went to listen to 
it, and it was just noise. I tried again, and had the same symptoms you 
had: It wrote for a while, then hung. My system was very slow, and the 
disk access light was on constantly. I was able to still run programs, 
but they were very, very slow. Restarting X did not help. Rebooting did. 
I have this is my /var/log/messages:

Nov 11 22:43:32 quicksilver kernel: oom-killer: gfp_mask=0x1d2
Nov 11 22:43:52 quicksilver kernel: DMA per-cpu:
Nov 11 22:43:53 quicksilver kernel: cpu 0 hot: low 2, high 6, batch 1
Nov 11 22:43:53 quicksilver kernel: cpu 0 cold: low 0, high 2, batch 1
Nov 11 22:43:53 quicksilver kernel: Normal per-cpu:
Nov 11 22:43:53 quicksilver kernel: cpu 0 hot: low 32, high 96, batch 16
Nov 11 22:43:53 quicksilver kernel: cpu 0 cold: low 0, high 32, batch 16
Nov 11 22:43:53 quicksilver kernel: HighMem per-cpu: empty
Nov 11 22:43:53 quicksilver kernel:
Nov 11 22:43:53 quicksilver kernel: Free pages:        4156kB (0kB HighMem)
Nov 11 22:43:53 quicksilver kernel: Active:1688 inactive:12572 dirty:1 
writeback:10501 unstable:0 free:1039 slab:3495 mapped:2057 pagetables:1005
Nov 11 22:43:53 quicksilver kernel: DMA free:2028kB min:12kB low:24kB 
high:36kB active:756kB inactive:168kB present:16384kB
Nov 11 22:44:30 quicksilver kernel: protections[]: 6 508 508
Nov 11 22:45:37 quicksilver kernel: Normal free:2128kB min:1004kB 
low:2008kB high:3012kB active:5996kB inactive:50120kB present:1032128kB
Nov 11 22:47:07 quicksilver kernel: protections[]: 0 502 502
Nov 11 22:47:08 quicksilver kernel: HighMem free:0kB min:128kB low:256kB 
high:384kB active:0kB inactive:0kB present:0kB
Nov 11 22:47:08 quicksilver kernel: protections[]: 0 0 0
Nov 11 22:47:09 quicksilver kernel: DMA: 1*4kB 7*8kB 29*16kB 25*32kB 
9*64kB 1*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 2028kB
Nov 11 22:47:10 quicksilver kernel: Normal: 32*4kB 0*8kB 3*16kB 1*32kB 
0*64kB 1*128kB 1*256kB 1*512kB 1*1024kB 0*2048kB 0*4096kB = 2128kB
Nov 11 22:47:10 quicksilver kernel: HighMem: empty
Nov 11 22:47:11 quicksilver kernel: Swap cache: add 1250133, delete 
1238940, find 183044/383177, race 0+2
Nov 11 22:47:11 quicksilver kernel: Out of Memory: Killed process 8208 
(evolution).
Nov 11 22:47:12 quicksilver kernel: oom-killer: gfp_mask=0xd2

which goes on and on and keeps killing processes (killing everything 
that was running).

I just tried to burn an iso image with cdrecord, and it worked fine.