Jens Axboe wrote:
Hi,
This project has been sleeping for a while, so I thought it was about time something happened. And now it has - I've put up version 0.0.2c of the CD-RW packet writing module, aka the "happy birthday grandma" release 8)
A summary of some of the changes:
- inc usage count of buffer heads - add internal buffer pool to avoid deadlock on oom - gather data for as many buffers as we have, before initiating write. this allows the laser to stay on longer, giving better performance. - fix always busy when tray can't be locked - remove request duplication nastiness, inject directly into the target - adapted to devfs and elevator changes - added proc interface
Not a whole lot, this is just to get the ball rolling again. However, the addition of the internal buffer pool was something that should have been added a long time ago. So this release should be significantly more stable than 0.0.2b -- at least it is for me, your mileage may vary.
It's still blowing away the whole packet on write for me, at least on my IDE drive (I havn't tried the SCSI yet). After mounting/unmounting (so only 1 block gets written) trying to read back any block in the packet gives: hdc: cdrom_decode_status: status=0x51 { DriveReady SeekComplete Error } hdc: cdrom_decode_status: error=0x30 hdc: cdrom_decode_status: status=0x51 { DriveReady SeekComplete Error } hdc: cdrom_decode_status: error=0x30 hdc: cdrom_decode_status: status=0x51 { DriveReady SeekComplete Error } hdc: cdrom_decode_status: error=0x30 hdc: cdrom_decode_status: status=0x51 { DriveReady SeekComplete Error } hdc: cdrom_decode_status: error=0x30 hdc: DMA disabled hdc: ATAPI reset complete ATAPI device hdc: Error: Unit attention -- (Sense key=0x06) Power on, reset or hardware reset occurred -- (asc=0x29, ascq=0x00) Using cdrwtool, I can rewrite the packet, at which point I can re-read from it. Ben -- Linux UDF - http://linux-udf.sourceforge.net Latest Is - udf-0.9.2.1 (http://www.csc.calpoly.edu/~bfennema/udf.html)