July 2001 - Packet Writing - openSUSE Mailing Lists

packet-0.0.2m fails with 2.4.7 kernel
by Graham Vincent 19 Aug '01

19 Aug '01

Hello. The patch fails on one hunk ... patching file drivers/block/ll_rw_blk.c Hunk #1 succeeded at 439 (offset 1 line). Hunk #3 succeeded at 483 (offset 1 line). Hunk #5 succeeded at 863 (offset 1 line). Hunk #7 succeeded at 1005 (offset 1 line). Hunk #9 succeeded at 1102 (offset 1 line). Hunk #10 FAILED at 1133. 1 out of 10 hunks FAILED -- saving rejects to file drivers/block/ll_rw_blk.c.rej The .rej file ... *************** *** 1143,1148 **** { if (req->sem != NULL) up(req->sem); blkdev_release_request(req); } --- 1133,1140 ---- { if (req->sem != NULL) up(req->sem); + if (req->end_io) + req->end_io(req); blkdev_release_request(req); } I look forward to the next version. Keep up the good work guys! Regards, Graham

4 5

bugs in latest patch
by Matthew Gabeler-Lee 19 Aug '01

19 Aug '01

I managed to get the latest patch to apply to 2.4.6 (had to do a couple small bits manually), and when I booted I noticed something weird: The /dev/pktcdvd that devfs made magically appear was major 0, minor 97! Rummaging around in the code, I found that the line in drivers/block/pktcdvd.c which calls devfs_register has its parameters mixed about. It has as 3 of its parameters "0, DEVFS_FL_DEFAULT, PACKET_MAJOR" (sorry about minor errors in const names, don't have code to look at here). The parameters to devfs_register for those 3 are supposed to be in the order flags, major, minor, so the 0 needs to be moved to be after PACKET_MAJOR. Making that change made the devfs node appear correctly. Also, when built as a module, it seems to leak the /proc/drivers/pktcdvd directory node when unloaded, causing multiple of those nodes to appear if it gets loaded and unloaded repeatedly. -- -Matt Using words to describe magic is like using a screwdriver to cut roast beef. -- Tom Robbins

6 6

trouble with hp 9510i
by Mark Vytlacil 31 Jul '01

31 Jul '01

I have been using an hp 8100 atapi with good results. I just upgraded to a hp 9510i atapi and can't get it to make the udf: lab:~# cdrwtool -d /dev/hdc -q using device /dev/hdc 5064KB internal buffer setting write speed to 12x Settings for /dev/hdc: Fixed packets, size 32 Mode-2 disc I'm going to do a quick setup of /dev/hdc. The disc is going to be blanked and formatted with one big track. All data on the device will be lost!! Press CTRL-C to cancel now. Initiating quick disc blank Disc capacity is 295264 blocks (590528KB/576MB) Formatting track Writing UDF structures to disc -14400 (EDT)/240 wait_cmd: Input/output error Command failed: 2a 00 00 00 00 00 00 00 20 00 00 00 - sense 05.24.00 Error writing packet 0 (ffffffff) In the logs: Jul 29 19:03:44 lab kernel: hdc: ATAPI 32X CD-ROM CD-R/RW drive, 8192kB Cache Jul 29 19:03:44 lab kernel: Uniform CD-ROM driver Revision: 3.12 Jul 29 19:03:44 lab kernel: hdd: ATAPI 50X CD-ROM drive, 128kB Cache Jul 29 19:03:44 lab kernel: cdrom: This disc doesn't have any tracks I recognize! Jul 29 19:18:35 lab -- MARK -- Jul 29 19:26:05 lab kernel: hdc: packet command error: status=0x51 { DriveReady SeekComplete Error } Jul 29 19:26:05 lab kernel: hdc: packet command error: error=0x50 Jul 29 19:26:05 lab kernel: ATAPI device hdc: Jul 29 19:26:05 lab kernel: Error: Illegal request -- (Sense key=0x05) Jul 29 19:26:05 lab kernel: Invalid field in command packet -- (asc=0x24, ascq=0x00) Jul 29 19:26:05 lab kernel: The failed "Write 10" packet command was: Jul 29 19:26:05 lab kernel: "2a 00 00 00 00 00 00 00 20 00 00 00 " Jul 29 19:26:05 lab kernel: sense key 05.24.00 I tried getting the cvs udf tools and using both 2x and 4x speeds with no change. I had been using 2.4.5 and packet-0.0.2l-as, so I got 2.4.7 and packet-0.0.2o-pre1 with no change. I have been using ide-cd all along. I hope this isn't a problem with some sort of proprietary version of the command set. Any suggestions? Mark Vytlacil

3 4

Ricoh MP6200A + Cache Memory
by Cameron Maxwell 31 Jul '01

31 Jul '01

Hello. I recently got given a second hand Ricoh MP6200A and tried the packet-0.0.2m with 2.4.7-pre7. I managed to get it working by turning on the immediate write option drivers/ide/ide-cd.c According to the Mt. Fuji specs this is the Force Unit Acess (FUA) bit, and that it disables write caching. Is this an artifact of my drive that it doesn't support write caching at all or is it more subtle? and is it a common feature of many drives? Sorry if this has been asked before. Thanks, Cameron Maxwell -- Not drinking, chasing women, or doing drugs won't make you live longer -- it just seems that way.

3 4

packet writing/UDF problems on 2.4.7
by Graham Vincent 28 Jul '01

28 Jul '01

Hello. I'm having a few problems getting packet writing to work satisfactorily with the 2.4.7 kernel. I have a PIII/Via chipset PC with an HP 8200+ cd writer. The system works fine with the 2.4.5 packet writing patch and UDF 0.9.4 The new kernel is compiled with UDF RW and packet writing enabled. I couldn't get /dev/pktsetup to work until I deleted /dev/pktcdvd0 and remade all the UDF 0.9.4.1 tools. Now it lets me mount the device and read/write to it. On inserting a formatted CDRW the kernel automounter tries to recognise the format to no avail even though it notes that the udf filesystem is registered on bootup (UDF problem perhaps?) I mounted the cdrw and can read and write files but I note an error message in the logs. When I umount the device the kernel goes back to unsuccessful attempts to identify it. A chunk of /var/log/messages for info... ul 28 16:11:32 hal kernel: Unable to identify CD-ROM format. Jul 28 16:12:03 hal last message repeated 21 times snip - several lines of this Jul 28 16:18:04 hal last message repeated 38 times Jul 28 16:18:04 hal kernel: pktcdvd: writer sr0 sucessfully registered Jul 28 16:18:06 hal kernel: Unable to identify CD-ROM format. Jul 28 16:18:38 hal last message repeated 28 times Jul 28 16:19:39 hal last message repeated 54 times Jul 28 16:19:39 hal kernel: pktcdvd: inserted media is CD-RW Jul 28 16:19:39 hal kernel: pktcdvd: Fixed packets, 32 blocks, Mode-2 disc Jul 28 16:19:40 hal kernel: pktcdvd: speed (R/W) 6/4 Jul 28 16:19:40 hal kernel: pktcdvd: 551488kB available on disc Jul 28 16:19:41 hal kernel: UDF-fs INFO UDF 0.9.4.1-rw (2001/06/13) Mounting volume 'gpv_linux2', timestamp 2001/06/18 18:50 (1d30) Jul 28 16:19:41 hal kernel: Unable to identify CD-ROM format. Jul 28 16:20:12 hal last message repeated 26 times snip again - multiple lines Jul 28 16:23:31 hal last message repeated 14 times Jul 28 16:23:35 hal kernel: cdrom: This disc doesn't have any tracks I recognize! Jul 28 16:23:37 hal kernel: Unable to identify CD-ROM format. Jul 28 16:24:09 hal last message repeated 28 times Jul 28 16:24:45 hal last message repeated 32 times Jul 28 16:24:57 hal kernel: sr0: CDROM (ioctl) reports ILLEGAL REQUEST. Jul 28 16:24:57 hal kernel: cdrom: This disc doesn't have any tracks I recognize! Jul 28 16:24:58 hal kernel: Unable to identify CD-ROM format. Jul 28 16:25:29 hal last message repeated 27 times Jul 28 16:26:30 hal last message repeated 50 times Jul 28 16:26:43 hal last message repeated 11 times Jul 28 16:26:44 hal kernel: pktcdvd: speed (R/W) 6/4 Jul 28 16:26:44 hal kernel: pktcdvd: writer sr0 unregistered Jul 28 16:26:45 hal kernel: Unable to identify CD-ROM format. Jul 28 16:27:17 hal last message repeated 22 times Jul 28 16:28:18 hal last message repeated 43 times Can anyone help sort this problem out for me? Thanks, Graham

1 0

ac kernel branch
by Alexander Trotsai 25 Jul '01

25 Jul '01

Hello all Is there patch for latest AC kernel branch? -- UaNic: MAGE1-UANIC, RIPE: MAGE-RIPE, InterNic: AT2963, ICQ: 18035130 PGP: ftp://blackhole.adamant.net/pgp/mykey.pgp.asc Trouble of the day: SCSI's too wide.

2 1

ext3 + packet patches: compatibility ?
by Frédéric L. W. Meunier 24 Jul '01

24 Jul '01

Hi. I'm using ext3 0.9.0 for 2.4.6 and noticed that both ext3 and packet patch the following files: patching file Documentation/Configure.help patching file drivers/block/ll_rw_blk.c patching file fs/buffer.c patching file include/linux/fs.h The only problems I can see would be ll_rw_blk.c and buffer.c . I'm still waiting the ext3 patch for 2.4.7, but would like to know if I can get a broken Kernel by using both ext3 and packet. I only know that sometimes both patches not generating rejects doesn't mean that everything will work. -- 0(a)pervalidus.{net, {dyndns.}org} Tel: 55-21-2717-2399 (Niterói-RJ BR)

2 1

Crash induced by buggy firmware
by Peter Osterlund 23 Jul '01

23 Jul '01

Hello! I am trying to use an Acer 4406EU USB CDRW drive for packet writing. This drive has a firmware bug that causes SCSI commands with a reply length of 2 to fail. (I don't know exactly what happens, but by reading the changelog for cdrecord, this bug seems to cause DMA problems when using a regular SCSI drive, which will trigger a SCSI bus reset. I have no idea what happens in the USB case.) When using plain 2.4.7 kernel and mounting a UDF disk, the kernel says: Jul 23 23:29:05 pengo kernel: sr1: CDROM (ioctl) error, command: 0x51 00 00 00 00 00 00 00 02 00 Jul 23 23:29:05 pengo kernel: Current sr00:00: sns = 70 0 Jul 23 23:29:05 pengo kernel: Raw sense data:0x70 0x00 0x00 0x00 0x00 0x00 0x00 0x0a 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 Jul 23 23:29:06 pengo kernel: I/O error: dev 0b:01, sector 1334868 Jul 23 23:29:13 pengo kernel: I/O error: dev 0b:01, sector 1333844 Jul 23 23:29:21 pengo kernel: I/O error: dev 0b:01, sector 1333620 Jul 23 23:29:21 pengo kernel: I/O error: dev 0b:01, sector 1334860 Jul 23 23:29:29 pengo kernel: I/O error: dev 0b:01, sector 1333836 Jul 23 23:29:38 pengo kernel: I/O error: dev 0b:01, sector 1333612 Jul 23 23:29:46 pengo kernel: I/O error: dev 0b:01, sector 1334268 Jul 23 23:29:53 pengo kernel: I/O error: dev 0b:01, sector 1333244 Jul 23 23:30:01 pengo kernel: I/O error: dev 0b:01, sector 1333020 Jul 23 23:30:07 pengo kernel: I/O error: dev 0b:01, sector 1334260 Jul 23 23:30:15 pengo kernel: I/O error: dev 0b:01, sector 1333236 Jul 23 23:30:23 pengo kernel: I/O error: dev 0b:01, sector 1333012 Jul 23 23:30:23 pengo kernel: I/O error: dev 0b:01, sector 1095300 Jul 23 23:30:25 pengo kernel: UDF-fs INFO UDF 0.9.4.1-rw (2001/06/13) Mounting volume 'test01', timestamp 1999/10/09 22:29 (1078) I don't know if the I/O errors are related to the firmware bug, but anyway, with this 2.4.7 kernel, the mount eventually succeeds. When I try the same thing with 2.4.7+packet patch, (mounting read-only, the pktcdvd module is not involved), I get an error interrupt from the USB controller, followed by an oops and a hard hang. Manual decoding of the oops: Unable to handle kernel NULL pointer dereference at virtual address 0000000c. EIP: kfree+35 Call Trace: [bh_action] [handle_IRQ_event] [do_IRQ] [ret_from_intr] Kernel panic: Aiee, killing interrupt handler! -- Peter Österlund peter.osterlund(a)mailbox.swipnet.se Sköndalsvägen 35 http://home1.swipnet.se/~w-15919 S-128 66 Sköndal +46 8 942647 Sweden

1 0

write speed detection
by Ravi Swamy 22 Jul '01

22 Jul '01

I just installed the new version of the packet writing patch and I was wondering about the write speed detection. When I used cdrwtool to format my blank CD-RW disc it said "setting write speed to 12x" yet I have an HP9200 SCSI CD-RW drive that is *supposed* to be an 8X CD-R writer, and only 4X CD-RW writer. I used the -t option to cdrwtool to blank and format at 4X. Then when I mounted the disc I got: pktcdvd: inserted media is CD-RW pktcdvd: Fixed packets, 32 blocks, Mode-2 disc pktcdvd: speed (R/W) 12/8 pktcdvd: 553088kB available on disc I looked through the source code of drivers/block/pktcdvd.c and saw the comment about reads being 1.5X of write speed which matches the above line but my burner is still only officially a 4X rewriter. Can I force the rewrite speed through a mount option or additional code to pktsetup.c from the udf package? I copied 620 files totalling 41MB, unmounted the disc, remounted in a DVD-ROM drive, and checked all of the md5sums with the originals and everything was the same. Should I be worried about this? Thanks, Ravi

2 1

Mitumi Problems
by Henning Glawe 19 Jul '01

19 Jul '01

Moin, ich habe einen 4802TE von Mitsumi, und wie auch schon bei verwendung von cdrecord gibt er bei cdrwtool grundlos die wildesten fehlermeldungen zurueck. beim fixiaten unter cdrecord sind es 0x00-no error-meldungen, die joerg schilling vor einiger zeit einer fehlerhaften scsi-implementation der mitsumi-brenner zugeschrieben hatte. um auf den punkt zu kommen: cdrwtool -q bricht einfach ab, der brenner faengt nach dem abbruch zu schreiben an. einzelsteps: cdrwtool -bfast verhaelt sich genau so cdrwtool -m 20000 fliegt mit wait-cmd: no medium found command failed: 04 17 00 00 [...] - sense 02.04.04 format disc: no medium found raus... sollte mir wohl doch lieber einen vernuenftigen brenner kaufen ;( c u henning

3 2