Mailinglist Archive: packet-writing (84 mails)

< Previous Next >
Re: New cdrwtool (trying to format too many blocks)
  • From: Peter Osterlund <petero2@xxxxxxxxx>
  • Date: Fri, 8 Feb 2002 20:15:26 +0000 (UTC)
  • Message-id: <Pine.LNX.4.44.0202082036170.3639-100000@xxxxxxxxxxxxxxxx>
On Fri, 8 Feb 2002, Sergiy Kudryk wrote:

> Hello Peter
>
> Unfortunally your hint doesn't help me to achieve stable writing
> (synchronously or in background) to CDRW disc.
>
> Copying any big portion of data (tested with mc help) failed after few
> second with message "no space left on device" on console and many i/o
> errors in /var/log/messages

Is the same disc working in DirectCD in windows? A worn out disc would
probably generate the same errors you are seeing.

Another thing you could try is to make the read speed equal to the write
speed. Your drive/disc combination seemed to be able to handle 8x, so you
could use this patch. (I have no idea if this will really help.)

--- pktcdvd.c.old Fri Feb 8 21:09:59 2002
+++ pktcdvd.c Fri Feb 8 21:09:17 2002
@@ -1670,7 +1670,7 @@
* and gather data. maybe 1/1 factor is faster, needs a bit of testing.
*/
speed = speed * 0xb0;
- read_speed = (speed * 3) >> 1;
+ read_speed = speed;
read_speed = min_t(unsigned, read_speed, 0xffff);

init_cdrom_command(&cgc, NULL, 0, CGC_DATA_NONE);
@@ -1784,7 +1784,7 @@

(void) pkt_write_caching(pd, USE_WCACHING);

- if ((ret = pkt_adjust_speed(pd, 16))) {
+ if ((ret = pkt_adjust_speed(pd, 8))) {
DPRINTK("pktcdvd: %s couldn't set write speed\n", pd->name);
return -EIO;
}

> Feb 7 15:31:34 my kernel: pktcdvd: Fixed packets, 32 blocks, Mode-2 disc
> Feb 7 15:31:34 my kernel: pktcdvd: enabled write caching on pktcdvd0
> Feb 7 15:31:34 my kernel: pktcdvd: speed (R/W) 12/8
> Feb 7 15:31:34 my kernel: pktcdvd: 590528kB available on disc
> Feb 7 15:31:36 my kernel: UDF-fs INFO UDF 0.9.5-rw (2001/10/10) Mounting volume 'LinuxUDF',
> timestamp 2002/02/07 13:15 (1078)
> Feb 7 15:32:13 my kernel: I/O error: dev 0b:00, sector 5648
> Feb 7 15:32:13 my kernel: Ignoring read error on sector:5648
> ...
>
> Feb 7 15:32:41 my kernel: Ignoring read error on sector:1400
> Feb 7 15:32:41 my kernel: Ignoring read error on sector:1404
> Feb 7 15:32:42 my kernel: I/O error: dev 0b:00, sector 124548
> Feb 7 15:32:44 my last message repeated 855 times
> Feb 7 15:32:54 my kernel: I/O error: dev 0b:00, sector 213988
> Feb 7 15:32:54 my kernel: I/O error: dev 0b:00, sector 213988
> Feb 7 15:32:54 my kernel: Unable to handle kernel NULL pointer dereference at virtual address
> 00000034
> Feb 7 15:32:54 my kernel: printing eip:
> Feb 7 15:32:54 my kernel: c0173c07
> Feb 7 15:32:54 my kernel: *pde = 00000000
> Feb 7 15:32:54 my kernel: Oops: 0000
> Feb 7 15:32:54 my kernel: CPU: 0
> Feb 7 15:32:54 my kernel: EIP: 0010:[udf_add_entry+2407/3044] Not tainted

...

> In any case after write fault i can't unmount disc (get error "device is
> busy") - device is locked.

It seems the udf filesystem can't handle the massive amount of media
errors you get. Instead it oopses and the device is left in "busy" state.

--
Peter Osterlund - petero2@xxxxxxxxx
http://w1.894.telia.com/~u89404340


< Previous Next >
References