September 2003 - Packet Writing - openSUSE Mailing Lists

SUSE 8.2 Packet-writing problems
by Daniel Pörsch 30 Sep '03

30 Sep '03

Hi friends, I recently upgraded to SuSE 8.2 and was happy to find udf support. Getting happier and happier as i read the manpages. Immediatly started a test. cdrwtool -d /dev/sr0 -q -t 4 finished without an error, but also without an confirmation ... So i did the next step pktsetup /dev/pktcdvd0 /dev/sr0 ... and got the message open packet device: No such device or address So I searched the internet and found mknod /dev/pktcdvd0 b 97 0 but still got the error, some more googling pointed that there should be a module pktvdcd, but it isn't available in my installation (searched trough ARCHIVES.gz) a .cz site (so i only understood the commands), it was said that modules.conf does need another line: alias block-major-97 pktcdvd I did this and a depmod afterwards, but without success. modprobe pktcdvd, tells ... modprobe: Can't locate module pktcdvd Thanx for any help!!! Greetings!!! Daniel -- Daniel Pörsch <daniel.poersch(a)web.de>

3 2

Re: Re: DVD packet writing
by Dyson 18 Sep '03

18 Sep '03

-I understand that it is dumb to write a buch of small files to a huge media where tar or bru would be the better option. I am just interested in where or why the problem with this is. write file read file write next file Where it would seem to make more sense to do this to pktcdvd in a batch like mode. I'm not sure if it might be just the way that UDF works with Linux. Does it read data on the disk before positioning the file? Does it read directory data before adding file "inodes" to the directory data? You are very correct that I could loop mount and burn quicker. growisofs works great for functions like this also. It would be convenient to have a program like mkisofs for udf that would lay files down in a stream rather than one at a time. You said you can't grow UDF. Could the packet stream be preordered and written in a stream to the media? Thanks, Dyson Volker Kuhlmann wrote: > > I tried cp-sort to the dvd. The built Linux source tree 176MB took about > > 45 minutes > > Useless. DVDs are less than 5GB. Just get a small fast disk if you > don't have a big one already, and image the DVD contents on it. I made > cp-sort to copy into udf filesystems created with mkudffs in a file > which is loop-mounted. It's a good idea to use it for ext2 as well. The > smaller your average file size is, the more you're forced to image on > disk. If you want to add, you'd have to read the dvd image to disk, > enlarge (oops, not with udf, so you need to copy to a new one, this is > when you decide to use ext2 instead), add files, run cp-sort, unmount > and burn back. With 176MB you can still do that in less than 45min. > > Packet writing is not useful until it is as reliable and at least as > fast as the stupid way of dvd->harddisk->addfiles->dvd. If you only have > to add very large files, it's probably ok. > > I am unsure whether your observation that it works better with cp-sort > than with cp is due to cp-sort. > > Volker > > -- > Volker Kuhlmann is possibly list0570 with the domain in header > http://volker.dnsalias.net/ Please do not CC list postings to me. > > -- > To unsubscribe, e-mail: packet-writing-unsubscribe(a)suse.com > For additional commands, e-mail: packet-writing-help(a)suse.com >

2 1

[Fwd: Re: DVD packet writing]
by Dyson 18 Sep '03

18 Sep '03

-------- Original Message -------- From: - Tue Sep 16 05:57:56 2003 X-Mozilla-Status: 0001 X-Mozilla-Status2: 00800000 Message-ID: <3F66DE9D.7000200(a)Adelphia.net> Date: Tue, 16 Sep 2003 05:57:49 -0400 From: Dyson <Dyson(a)Adelphia.net> User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030624 Netscape/7.1 X-Accept-Language: en-us, en MIME-Version: 1.0 To: sezero(a)superonline.com Subject: Re: DVD packet writing References: <1243132.1063693511594.JavaMail.root@qfep03> In-Reply-To: <1243132.1063693511594.JavaMail.root@qfep03> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 8bit Yes I did unmount and remount it before the diff. I think it is actually the cp command that is the culprit. It verifies every file after it is written. On the Philips DVDRW228 it has a blue light for reading and red for writing. It seemed to write then read over and over as I copied the linux source tree to it. I did another test with large files 500MB, 1.2GB and 2.9GB using tar then diff'ed them after restoring to another directory on the hd. I'm thinking of working on a script that will read the directory contents then pass that to dd or sdd to do each file in a batch mode. I could call it cpdir, cpudf, cpcdr or something. Thanks, Dyson sezero(a)superonline.com wrote: >Hi Dyson: > > > the directories were missing. However the files it > > did copy were the same. > >I don't have a dvd writer, so I can't say anything >about that. But about the "files being the same": >Can you please: >- umount the newly written fs and remount it r/o >and then compare ( to ensure nothing is cached); >- write some files to the fs, delete some others, >write some others again. (overwriting a file >doesn't count). Then maka the comparison again >after re-mounting the fs r/o. > >On cdrw media, I occasionally hit some corrupted >files using the above test. Most times (not always) >the corruption occurs when delete happens at about >the same time the write. > >Regards, >Özkan Sezer > > > >

2 5

Re: Re: [Fwd: Re: DVD packet writing]
by Dyson 17 Sep '03

17 Sep '03

Thanks! Volker Kuhlmann wrote: > > I think it is actually the cp command that is the culprit. It verifies > > every file after it is written. On the Philips DVDRW228 it has a blue > > > I'm thinking of working on a script that will read the directory > > contents then pass that to dd or sdd to do each file in a batch mode. I > > could call it cpdir, cpudf, cpcdr or something. > > I made cp-sort, which uses cpio (and rsync afterwards) for copying > files in alphabetic order (as kind of a defrag). It might do want you > want. It's in the scriptutils package on my website. > > Volker > > -- > Volker Kuhlmann is possibly list0570 with the domain in header > http://volker.dnsalias.net/ Please do not CC list postings to me. > > -- > To unsubscribe, e-mail: packet-writing-unsubscribe(a)suse.com > For additional commands, e-mail: packet-writing-help(a)suse.com >

1 0

Re: DVD packet writing
by Dyson 16 Sep '03

16 Sep '03

Hi- I am running SuSE 8.0 with custom kernels. The patch would most likely have to be modified to patch against 2.4.20-SuSE since it is brewed from the 2.4.22-packet patch. Check out http://w1.894.telia.com/~u89404340/patches/packet/2.4/ <http://w1.894.telia.com/%7Eu89404340/patches/packet/2.4/> for other patches for different kernel versions. I suspect you might have to run a custom non-SuSE kernel yourself. -Dyson Luescher, Carl wrote: >Hi all, > >The issue I am having is, as of this moment, not related to DVD. Although I >am more than willing to to test these patches to see if it doesn't resolve >my cdrwtool problem with SCSI emulation. My question first is - I am >running SuSE 8.2 with a 2.4.20 kernel compiled with packet writing, do you >think your patch would cause any problems on this release? > >Any advice appreciated, and thanks! > >Carl > >-----Original Message----- >From: Dyson [mailto:Dyson@Adelphia.net] >Sent: Sunday, September 14, 2003 5:29 AM >To: John Williams >Cc: packet-writing(a)suse.com >Subject: Re: DVD packet writing > > >Hi- > >I put another patch at http://dyson.kristenscatbeds.com that is written >for Linux-2.4.22. It joins the three patches together. > >This includes John Williams' patch, the packet-2.4.22 patch by Jens >Axboe and the DVD+RW patch by Andy Polyakov. > >I used a Philips DVDRW228 and it worked fine as long as I used cdrwtool >with /dev/pktcdvd0 instead of /dev/scd0. > >I used the udf format instructions from Andy Polyakov's DVD+RW for Linux >site. > >I also have a USB/Firewire HP DVD 300e I will try it on later. > >The computer does stop responding if I use cp -R (some huge directory >with many small files) /cdrom/ -but this was true before as well. > >Thanks, > >Dyson > > >John Williams wrote: > > > >>Hello, >> >>I've got an NEC ND1300 DVD burner, and the DVD+RW patch from >> >> http://fy.chalmers.se/~appro/linux/DVD+RW >> >>doesn't work on it: evidently this a firmware bug. Also this patch >>conflicts with the CDRW packet writing patch and stops it from working. >> >>I've made a fix for this, which can be downloaded from >> >> http://www.pond-weed.com/dvd.html >> >>With the fix, the CDRW patch works, and the same CDRW-style packet >>writing can be used to write to a DVD+RW too. >> >>If your DVD writer doesn't have buggy firmware, my patch will still >>allow you to use both CDRW packet writing and the DVD+RW patch at the >>same time. >> >>John >> >> >> >> >> > > > >

1 0

UDF on a CDR : problems mounting
by Peter Van Biesen 16 Sep '03

16 Sep '03

Hi, I recently received a packet-written CDR from informix. I tried to read it on a windows ( NT4.0SP4 ) box and this works fine. I then tried to mount it on a linux box, but this fails. cdrwtools produces this output : -------------------------------------------- using device /dev/cdrecorder 1280KB internal buffer setting write speed to 12x DISC INFO: erasable : No border = 1 Disc status = 1 number of first track = 1 number of sessions = 3 number of tracks = 3 status of last track = 3 uru = 0 did_v = 1 dbc_v = 0 disc type = 0 disc_id = 130389 lead_in = 20:44:24 (93324) lead_out = 79:59:74 (359999) OPC entries = 0 TRACK INFO: Track 1 track_number = 1 session_number = 1 damage = 0 copy = 0 track_mode = 7 Rt = 1 blank = 0 packet = 1 fp = 0 data_mode = 1 lra_v = 0 nwa_v = 0 track_start = 0 next_writable = 78743 last_recorded = 0 free_blocks = 0 packet_size = 0 track_size = 78743 (157486KB) Track 2 track_number = 2 session_number = 2 damage = 0 copy = 0 track_mode = 7 Rt = 1 blank = 0 packet = 1 fp = 0 data_mode = 1 lra_v = 0 nwa_v = 0 track_start = 90145 next_writable = 90922 last_recorded = 0 free_blocks = 0 packet_size = 0 track_size = 777 (1554KB) Track 3 track_number = 3 session_number = 3 damage = 0 copy = 0 track_mode = 7 Rt = 0 blank = 0 packet = 1 fp = 0 data_mode = 1 lra_v = 0 nwa_v = 1 track_start = 97824 next_writable = 98352 last_recorded = 0 free_blocks = 261495 packet_size = 0 track_size = 262023 (524046KB) ------------------------------------------------ Mounting the drive : pcvlf936:~ # mount -t udf /dev/cdrecorder /media/cdrecorder mount: block device /dev/cdrecorder is write-protected, mounting read-only mount: wrong fs type, bad option, bad superblock on /dev/cdrecorder, or too many mounted file systems from /var/log/messages : ---------------------------------------------------- Sep 16 10:52:24 pcvlf936 kernel: UDF-fs DEBUG lowlevel.c:57:udf_get_last_session: XA disk: no, vol_desc_start=0 Sep 16 10:52:24 pcvlf936 kernel: UDF-fs DEBUG super.c:1421:udf_read_super: Multi-session=0 Sep 16 10:52:24 pcvlf936 kernel: UDF-fs DEBUG super.c:410:udf_vrs: Starting at sector 16 (2048 byte sectors) Sep 16 10:52:28 pcvlf936 kernel: UDF-fs DEBUG super.c:753:udf_load_pvoldesc: recording time 1062419651/0, 2003/09/01 14:34 (1078) Sep 16 10:52:28 pcvlf936 kernel: UDF-fs DEBUG super.c:763:udf_load_pvoldesc: volIdent[] = 'IDS940.FC2W2' Sep 16 10:52:28 pcvlf936 kernel: UDF-fs DEBUG super.c:770:udf_load_pvoldesc: volSetIdent[] = '0x3F5358VERITAS ' Sep 16 10:52:28 pcvlf936 kernel: UDF-fs DEBUG super.c:962:udf_load_logicalvol: Partition (0:0) type 1 on volume 1 Sep 16 10:52:28 pcvlf936 kernel: UDF-fs DEBUG super.c:962:udf_load_logicalvol: Partition (1:0) type 2 on volume 1 Sep 16 10:52:28 pcvlf936 kernel: UDF-fs DEBUG super.c:972:udf_load_logicalvol: FileSet found in LogicalVolDesc at block=0, partition=1 Sep 16 10:52:28 pcvlf936 kernel: UDF-fs DEBUG super.c:800:udf_load_partdesc: Searching map: (0 == 0) Sep 16 10:52:28 pcvlf936 kernel: UDF-fs DEBUG super.c:882:udf_load_partdesc: Partition (0:0 type 1511) starts at physical 513, block length 359182 Sep 16 10:52:28 pcvlf936 kernel: UDF-fs DEBUG super.c:1215:udf_load_partition: Using anchor in block 256 Sep 16 10:52:28 pcvlf936 kernel: attempt to access beyond end of device Sep 16 10:52:28 pcvlf936 kernel: 0b:00: rw=0, want=196690, limit=181844 Sep 16 10:52:28 pcvlf936 kernel: attempt to access beyond end of device Sep 16 10:52:28 pcvlf936 kernel: 0b:00: rw=0, want=196178, limit=181844 Sep 16 10:52:28 pcvlf936 kernel: attempt to access beyond end of device Sep 16 10:52:28 pcvlf936 kernel: 0b:00: rw=0, want=196066, limit=181844 Sep 16 10:52:28 pcvlf936 kernel: attempt to access beyond end of device Sep 16 10:52:28 pcvlf936 kernel: 0b:00: rw=0, want=196686, limit=181844 Sep 16 10:52:28 pcvlf936 kernel: attempt to access beyond end of device Sep 16 10:52:28 pcvlf936 kernel: 0b:00: rw=0, want=196174, limit=181844 Sep 16 10:52:28 pcvlf936 kernel: attempt to access beyond end of device Sep 16 10:52:28 pcvlf936 kernel: 0b:00: rw=0, want=196062, limit=181844 Sep 16 10:52:28 pcvlf936 kernel: attempt to access beyond end of device Sep 16 10:52:28 pcvlf936 kernel: 0b:00: rw=0, want=196390, limit=181844 Sep 16 10:52:28 pcvlf936 kernel: attempt to access beyond end of device Sep 16 10:52:28 pcvlf936 kernel: 0b:00: rw=0, want=195878, limit=181844 Sep 16 10:52:28 pcvlf936 kernel: attempt to access beyond end of device Sep 16 10:52:28 pcvlf936 kernel: 0b:00: rw=0, want=195766, limit=181844 Sep 16 10:52:28 pcvlf936 kernel: attempt to access beyond end of device Sep 16 10:52:28 pcvlf936 kernel: 0b:00: rw=0, want=196386, limit=181844 Sep 16 10:52:28 pcvlf936 kernel: attempt to access beyond end of device Sep 16 10:52:28 pcvlf936 kernel: 0b:00: rw=0, want=195874, limit=181844 Sep 16 10:52:28 pcvlf936 kernel: attempt to access beyond end of device Sep 16 10:52:28 pcvlf936 kernel: 0b:00: rw=0, want=195762, limit=181844 Sep 16 10:52:28 pcvlf936 kernel: I/O error: dev 0b:00, sector 322788 Sep 16 10:52:28 pcvlf936 kernel: I/O error: dev 0b:00, sector 321764 Sep 16 10:52:28 pcvlf936 kernel: I/O error: dev 0b:00, sector 321540 Sep 16 10:52:28 pcvlf936 kernel: I/O error: dev 0b:00, sector 322780 Sep 16 10:52:28 pcvlf936 kernel: I/O error: dev 0b:00, sector 321756 Sep 16 10:52:28 pcvlf936 kernel: I/O error: dev 0b:00, sector 321532 Sep 16 10:52:28 pcvlf936 kernel: I/O error: dev 0b:00, sector 322188 Sep 16 10:52:28 pcvlf936 kernel: I/O error: dev 0b:00, sector 321164 Sep 16 10:52:28 pcvlf936 kernel: I/O error: dev 0b:00, sector 320940 Sep 16 10:52:28 pcvlf936 kernel: I/O error: dev 0b:00, sector 322180 Sep 16 10:52:28 pcvlf936 kernel: I/O error: dev 0b:00, sector 321156 Sep 16 10:52:28 pcvlf936 kernel: I/O error: dev 0b:00, sector 320932 Sep 16 10:52:28 pcvlf936 kernel: UDF-fs DEBUG super.c:1234:udf_load_partition: Unable to determine Lastblock (For Virtual Partition) Sep 16 10:52:28 pcvlf936 kernel: UDF-fs: No partition found (1) ------------------------------------------------------- My kernel has NOT been patched for packet writing, I only want to mount the disk read-only. Does anybody have an idea how to do this ? Thanks in advance ! Peter. -- Peter Van Biesen Adj. Sysadmin V.F.S.I.P.H. tel: +32 (0) 2 225 85 70 fax: +32 (0) 2 225 85 88 e-mail: peter.vanbiesen(a)vlafo.be

1 0

Re: DVD packet writing
by O.Sezer 16 Sep '03

16 Sep '03

Hi Dyson: > the directories were missing. However the files it > did copy were the same. I don't have a dvd writer, so I can't say anything about that. But about the "files being the same": Can you please: - umount the newly written fs and remount it r/o and then compare ( to ensure nothing is cached); - write some files to the fs, delete some others, write some others again. (overwriting a file doesn't count). Then maka the comparison again after re-mounting the fs r/o. On cdrw media, I occasionally hit some corrupted files using the above test. Most times (not always) the corruption occurs when delete happens at about the same time the write. Regards, Özkan Sezer

1 0

DVD packet writing
by John Williams 14 Sep '03

14 Sep '03

Hello, I've got an NEC ND1300 DVD burner, and the DVD+RW patch from http://fy.chalmers.se/~appro/linux/DVD+RW doesn't work on it: evidently this a firmware bug. Also this patch conflicts with the CDRW packet writing patch and stops it from working. I've made a fix for this, which can be downloaded from http://www.pond-weed.com/dvd.html With the fix, the CDRW patch works, and the same CDRW-style packet writing can be used to write to a DVD+RW too. If your DVD writer doesn't have buggy firmware, my patch will still allow you to use both CDRW packet writing and the DVD+RW patch at the same time. John

3 4