Re: cdrtools-2.01a37 ready
On Thu, Aug 19, 2004 at 08:47:50AM +0200, Lourens Veen wrote: > On Thu 19 August 2004 07:51, Jacob Meuser wrote: > > > > I doubt any government regards the GPL as an entity, so the GPL > > has no rights. the GPL is a contract; following the GPL gives > > entities rights. SuSE can release a really buggy version of > > anything covered by the GPL, as long as they are complying with > > the terms of the GPL. what I don't get, is that if Joerg is so > > sure SuSE is in violation, why doesn't he just take them to > > court. that's why the GPL exists; so people can be made to > > comply or pay the consequences. it's not about freedom, it's > > about control. > > That's not the point. By printing out that notice, SuSE does comply > with the spirit of the GPL (that preamble clause about protecting > the author) and by not clearly marking the source code files they > may be violating the letter of the GPL. > > However, none of this is the issue. What Jörg is getting worked up > about is that he gets email from people complaining about bugs in a > version of his software that he has nothing to do with (neither > that version of cdrecord nor the bugs). That's the real issue. then why all the griping about GPL violations? -- <[EMAIL PROTECTED]> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: cdrtools-2.01a37 ready
> However, none of this is the issue. What Jörg is getting worked up > about is that he gets email from people complaining about bugs in a > version of his software that he has nothing to do with (neither > that version of cdrecord nor the bugs). That's the real issue. I can totally sympathise with Jörg on this one. However, wielding the legal hammer is only going to get everyone worked up, Jörg included. Surely this problem is not uniqe to cdrecord. Wouldn't putting up a (user-friendly) web page explaining the problem and returning emails not about Jörg's version of cdrecord with a pointer to this page be a lot more productive? > Personally, I wish everyone would just use growisofs to burn DVDs. > It's FSF-free, Debian-free, and open source, and it's completely > unrelated to cdrecord. And it's well-supported and it works. Couldn't agree more!! 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, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: cdrtools-2.01a37 ready
> come up with suggestions how to force SuSE to follow the GPL again. > Jörg How about providing operational binaries for SuSE users at the cdrtools download site ? That would help the little SuSE users because they do not have to cope with the compile time problems. That would consolidate the cdrecord fork bush because the SuSE binaries are not always operational and many little SuSE users would have reason to use your official binaries. At least SuSE would have to take the blame that the original author (and some of his users) had to take action to substitute their unnecessary hacks. (I agree with Lourens : anybody who wants to burn DVD and who does not like cdrecord's DVD license terms may use growisofs. SuSE 9.0's growisofs 5.12 binary does work.) Have a nice day :) Thomas PS: Joerg, the message about "Operation not permitted. WARNING: Cannot set RR-scheduler" is probably not due to source changes. I had it with an original cdrtools-2.01a19 which i compiled on SuSE 7.2. On SuSE 7.2 it ran without complains and on SuSE 9.0 it issued said message (setuid root). No real problems when burning, though. A cdrecord 1.6 binary compiled static on Debian in 1999 still works on SuSE 9.0 without any problems. Nowadays it is nearly impossible to produce binaries that are portable from Intel Linux to Intel Linux. DLLs. Pffft. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: cdrtools-2.01a37 ready
> then why all the griping about GPL violations? If this question is in my direction: because I got fed up with constant and annoying gripes about GPL violations by someone who can't even point to a reasonable actual violation[1] and then can't comply himself[2]. Besides, I'd like to know the license status of cdrecord, and I mightn't be the only one. Volker [1] Users don't look at the source. Programmers who do, find all of SuSE's changes cleanly separated in patch files. Frankly I'm not going to waste any time on whether source files should be marked too (it's nitpicky at best), especially as Jörg hasn't even requested it. And I sure asked! [2] No disputes from anyone yet -- Volker Kuhlmann is possibly list0570 with the domain in header http://volker.dnsalias.net/ Please do not CC list postings to me. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Testers for OpenBSD & NetBSD needed
Hi all, as it turned out far too late that cdda2wav did not work when installed suid root on FreeBSD, I did include a dirty hack into cdda2wav/setuid.c in order to honor the fact that FreeBSD now did try to implement POSIX uid handling on top of the old BSD ID swapping method. In order to make sure that cdda2wav is not hitting the same problems on OpenBSD & NetBSD, I would need people who like to test a suid cdda2wav on OpenBSD & NetBSD systems. In case it turns out that there are problems on OpenBSD & NetBSD too, try to edit cdda2wav/setuid.c and replace the lines that contain #if defined(__FreeBSD__) by lines that contain: #if defined(__FreeBSD__) || defined(__NetBSD__) || defined(__OpenBSD__) and try again. Jörg -- EMail:[EMAIL PROTECTED] (home) Jörg Schilling D-13353 Berlin [EMAIL PROTECTED](uni) If you don't have iso-8859-1 [EMAIL PROTECTED](work) chars I am J"org Schilling URL: http://www.fokus.fraunhofer.de/usr/schilling ftp://ftp.berlios.de/pub/schily -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: cdrtools-2.01a37 ready
> How about providing operational binaries for > SuSE users at the cdrtools download site ? I would have provided cdrecord packages, alas I never had problems with the SuSE-supplied ones, therefore no point spending time on it. 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, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: cdrtools-2.01a37 ready
>From: Volker Kuhlmann <[EMAIL PROTECTED]> >Obviously you're on the back foot and ran out of arguments, otherwise >you wouldn't spend a whole email on discussing a minor point (so I >didn't check my PPS carefully enough) and nothing else, instead of >sticking to the 2 points in question. And so that even you can >understand it: introducing bugs is not a violation of the GPL. It's a >right of the GPL. You still did not understand the problem :-( A program is an artwork (this is what the European Union did write into laws). Even though you may get the permission to modify an artwork, you will not get the permission to create bad carricatures and call them just "modified versions". The GPL requires you not to impact the original authors' reputations, but this is what they are doing by publishing defective variants. Jörg -- EMail:[EMAIL PROTECTED] (home) Jörg Schilling D-13353 Berlin [EMAIL PROTECTED](uni) If you don't have iso-8859-1 [EMAIL PROTECTED](work) chars I am J"org Schilling URL: http://www.fokus.fraunhofer.de/usr/schilling ftp://ftp.berlios.de/pub/schily -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: cdrtools-2.01a37 ready
>From: [EMAIL PROTECTED] >> come up with suggestions how to force SuSE to follow the GPL again. >> Jörg >How about providing operational binaries for >SuSE users at the cdrtools download site ? If you would look a bit to the left and to the right before posting, you would know that this is true since August 2001, just check: ftp://ftp.berlios.de/pub/cdrecord/ProDVD/ Jörg -- EMail:[EMAIL PROTECTED] (home) Jörg Schilling D-13353 Berlin [EMAIL PROTECTED](uni) If you don't have iso-8859-1 [EMAIL PROTECTED](work) chars I am J"org Schilling URL: http://www.fokus.fraunhofer.de/usr/schilling ftp://ftp.berlios.de/pub/schily -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: cdrtools-2.01a37 ready
On Thu 19 August 2004 14:35, Joerg Schilling wrote: > From: Volker Kuhlmann <[EMAIL PROTECTED]> > > >Obviously you're on the back foot and ran out of arguments, > > otherwise you wouldn't spend a whole email on discussing a > > minor point (so I didn't check my PPS carefully enough) and > > nothing else, instead of sticking to the 2 points in question. > > And so that even you can understand it: introducing bugs is not > > a violation of the GPL. It's a right of the GPL. > > You still did not understand the problem :-( > > A program is an artwork (this is what the European Union did > write into laws). Since both you and SuSE are in Germany, I would think that German law would apply. There are of course the Berne treaty and the EU Copyright Directive, but neither are laws. In the case of the EUCD, it is up to the member states to make local laws that implement the EUCD. Incidentally, which law are you referring to? I couldn't really find anything in the text of the EUCD. > Even though you may get the permission to modify an artwork, you > will not get the permission to create bad carricatures and call > them just "modified versions". > > The GPL requires you not to impact the original authors' > reputations, but this is what they are doing by publishing > defective variants. Which is why they stick a big notice on it that literally says that they changed it and that they may have introduced new bugs. The problem is that people don't read the notice, and then start bothering you. But can you really blame SuSE for that? What more can they do about it? Send over an employee to each buyer to explain the situation? Lourens PS: congratulations on the new mail client. It's much better. -- GPG public key: http://home.student.utwente.nl/l.e.veen/lourens.key
Re: cdrtools-2.01a37 ready
>>> come up with suggestions how to force SuSE to follow the GPL again. >>> Jörg >>How about providing operational binaries for >>SuSE users at the cdrtools download site ? > > ... just check: > >ftp://ftp.berlios.de/pub/cdrecord/ProDVD/ > I meant vanilla cdrecord for CD. As produced by compiling the official cdrtools source tar balls on an official SuSE system (with a little help by the compiling SuSE user). 100 % GPL by refering to the source archive. Documenting any changes which became inavoidable and have been authorized by you, of course. No DVD, no special license, just a service for people who have difficulties compiling but want to work with the original thing and not a hack. Volker > > I would have provided cdrecord packages, alas I never had problems with the SuSE-supplied ones, therefore no point spending time on it. The binary (with DVD patch, disclaimer and all) which i found after system installation did not work setuid root. Since that method is advised by the man who must know, i will not advise my users to do it different. Thus, the SuSE 9.0 "cdrecord" has to be replaced after system installation or the users need to become superuser to backup their data. Really not my kind of usage model. I would be glad to point to a link on the cdrtools site where an original cdrecord binary could be obtained. I will not publish my own binary without Joerg's permission, because it is not from 100% original sources and i will not start another surplus fork of cdrecord (even if it's only about one #define HZ 100 to make it compile). Have a nice day :) Thomas -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Testers for OpenBSD & NetBSD needed
Hi Joerg, Joerg Schilling wrote on Thu, Aug 19 2004: >In order to make sure that cdda2wav is not hitting the same problems >on OpenBSD & NetBSD, I would need people who like to test a suid >cdda2wav on OpenBSD & NetBSD systems. /tmp$ su Password: /tmp$ whoami root /tmp$ uname -a OpenBSD thebetteros.oche.de 3.5 GENERIC#34 i386 /tmp$ ls -la `which cdda2wav` -rwxr-xr-x 1 root bin 267279 Aug 19 18:18 /usr/local/bin/cdda2wav /tmp$ cdda2wav -version cdda2wav version 2.01a37_openbsd_3.5_i386_i386 /tmp$ cdda2wav -V -vall dev=/dev/rcd0c -N -B >OUT.TXT 2>&1 Aug 19 18:34:35 thebetteros /bsd: pciide0:1: bogus intr Aug 19 18:39:45 thebetteros /bsd: cd0(pciide0:1:0): timeout Aug 19 18:39:45 thebetteros /bsd: type: atapi Aug 19 18:39:45 thebetteros /bsd: c_bcount: 1181 Aug 19 18:39:45 thebetteros /bsd: c_skip: 0 Aug 19 18:39:45 thebetteros /bsd: cd0: transfer error, downgrading to DMA mode 2 Aug 19 18:39:45 thebetteros /bsd: cd0(pciide0:1:0): using PIO mode 4, DMA mode 2 Aug 19 18:39:46 thebetteros /bsd: pciide0:1: bogus intr I attach OUT.TXT gzipped. HTH Frederick OUT.TXT.gz Description: application/gunzip
Re: Testers for OpenBSD & NetBSD needed
On Thu, Aug 19, 2004 at 11:46:37AM +0200, Joerg Schilling wrote: > Hi all, > > as it turned out far too late that cdda2wav did not > work when installed suid root on FreeBSD, I did include a dirty hack > into cdda2wav/setuid.c in order to honor the fact that FreeBSD now > did try to implement POSIX uid handling on top of the old > BSD ID swapping method. > > In order to make sure that cdda2wav is not hitting the same problems > on OpenBSD & NetBSD, I would need people who like to test a suid > cdda2wav on OpenBSD & NetBSD systems. Works fine for me on OpenBSD 3.6-beta. -- <[EMAIL PROTECTED]> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Testers for OpenBSD & NetBSD needed
On Thu, Aug 19, 2004 at 06:44:08PM +0200, Frederick Page wrote: > Hi Joerg, > > Joerg Schilling wrote on Thu, Aug 19 2004: > > >In order to make sure that cdda2wav is not hitting the same problems > >on OpenBSD & NetBSD, I would need people who like to test a suid > >cdda2wav on OpenBSD & NetBSD systems. > > /tmp$ su > Password: > /tmp$ whoami > root > /tmp$ uname -a > OpenBSD thebetteros.oche.de 3.5 GENERIC#34 i386 > /tmp$ ls -la `which cdda2wav` > -rwxr-xr-x 1 root bin 267279 Aug 19 18:18 /usr/local/bin/cdda2wav > /tmp$ cdda2wav -version > cdda2wav version 2.01a37_openbsd_3.5_i386_i386 > /tmp$ cdda2wav -V -vall dev=/dev/rcd0c -N -B >OUT.TXT 2>&1 > > > Aug 19 18:34:35 thebetteros /bsd: pciide0:1: bogus intr > Aug 19 18:39:45 thebetteros /bsd: cd0(pciide0:1:0): timeout > Aug 19 18:39:45 thebetteros /bsd: type: atapi > Aug 19 18:39:45 thebetteros /bsd: c_bcount: 1181 > Aug 19 18:39:45 thebetteros /bsd: c_skip: 0 > Aug 19 18:39:45 thebetteros /bsd: cd0: transfer error, downgrading to > DMA mode 2 > Aug 19 18:39:45 thebetteros /bsd: cd0(pciide0:1:0): using PIO mode 4, > DMA mode 2 > Aug 19 18:39:46 thebetteros /bsd: pciide0:1: bogus intr check your cables. out of curiousity, what is your IDE controller? downgrading dma modes and bogus interrupts are highly unlikely the fault of cdda2wav, and I'm pretty sure this has nothing to do with whether it was installed suid or not. this looks like a bad interaction with your IDE controller and your DVD burner. -- <[EMAIL PROTECTED]> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: cdrtools-2.01a37 ready
> Volker > > > I would have provided cdrecord packages, alas I never had problems with > the SuSE-supplied ones, therefore no point spending time on it. > > The binary (with DVD patch, disclaimer and all) which i > found after system installation did not work setuid root. > Since that method is advised by the man who must know, > i will not advise my users to do it different. That is a matter of opinion, of course. I dislike suid programs, and have only Jörg's word that it'll be ok. On the other hand I have a binary which is modified to not require suid, which seems the better concept to me in any case. If Jörg wants me to believe he's better than the SuSE security team (who have a bigger reputation to lose), he will have to supply better information than the exceptionally uninformative "it's defective", "you shouldn't use it", "they don't know what they're talking about" when asked about it. Answers like those don't give me the impression I want to trust that person in security matters thanks all the same. Let's not forget that there are 2 separate changes SuSE makes: the DVD addons (daft, and I've told them so), and the removal of the suid requirement. Jörg may have a different opinion about the latter, but he'll have to learn to accept that others don't share his. Dito for other technical matters. > because it is not from 100% original sources and i will not > start another surplus fork of cdrecord (even if it's only > about one #define HZ 100 to make it compile). Compiles OOTB on 9.1, though I never used it to actually burn anything. 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, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Testers for OpenBSD & NetBSD needed
>Frederick Page <[EMAIL PROTECTED]> wrote: >> >In order to make sure that cdda2wav is not hitting the same problems >> >on OpenBSD & NetBSD, I would need people who like to test a suid >> >cdda2wav on OpenBSD & NetBSD systems. >> >> /tmp$ su >> Password: >> /tmp$ whoami >> root >> /tmp$ uname -a >> OpenBSD thebetteros.oche.de 3.5 GENERIC#34 i386 >> /tmp$ ls -la `which cdda2wav` >> -rwxr-xr-x 1 root bin 267279 Aug 19 18:18 /usr/local/bin/cdda2wav >> /tmp$ cdda2wav -version >> cdda2wav version 2.01a37_openbsd_3.5_i386_i386 >> /tmp$ cdda2wav -V -vall dev=/dev/rcd0c -N -B >OUT.TXT 2>&1 >> >> >> Aug 19 18:34:35 thebetteros /bsd: pciide0:1: bogus intr >> Aug 19 18:39:45 thebetteros /bsd: cd0(pciide0:1:0): timeout >> Aug 19 18:39:45 thebetteros /bsd: type: atapi >> Aug 19 18:39:45 thebetteros /bsd: c_bcount: 1181 >> Aug 19 18:39:45 thebetteros /bsd: c_skip: 0 >> Aug 19 18:39:45 thebetteros /bsd: cd0: transfer error, downgrading to >> DMA mode 2 >> Aug 19 18:39:45 thebetteros /bsd: cd0(pciide0:1:0): using PIO mode 4, >> DMA mode 2 >> Aug 19 18:39:46 thebetteros /bsd: pciide0:1: bogus intr >> >> I attach OUT.TXT gzipped. Your problem is caused by a OpenBSD kernel bug. The kernel does not handle odd DMA counts for ATA correctly. But could you try to answer my question please: If you install cdda2wav suid root and call it as non root user, does it work? In order to avoid the Kernel bug from preventing you to do the test, check the number of traks on the CD. It it was a CD with an even number of tracks, use one with an odd number of tracks or vice versa. Jörg -- EMail:[EMAIL PROTECTED] (home) Jörg Schilling D-13353 Berlin [EMAIL PROTECTED](uni) If you don't have iso-8859-1 [EMAIL PROTECTED](work) chars I am J"org Schilling URL: http://www.fokus.fraunhofer.de/usr/schilling ftp://ftp.berlios.de/pub/schily -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Testers for OpenBSD & NetBSD needed
Jacob Meuser <[EMAIL PROTECTED]> wrote: > On Thu, Aug 19, 2004 at 11:46:37AM +0200, Joerg Schilling wrote: > > Hi all, > > > > as it turned out far too late that cdda2wav did not > > work when installed suid root on FreeBSD, I did include a dirty hack > > into cdda2wav/setuid.c in order to honor the fact that FreeBSD now > > did try to implement POSIX uid handling on top of the old > > BSD ID swapping method. > > > > In order to make sure that cdda2wav is not hitting the same problems > > on OpenBSD & NetBSD, I would need people who like to test a suid > > cdda2wav on OpenBSD & NetBSD systems. > > Works fine for me on OpenBSD 3.6-beta. Did you install cdda2wav suid root and call it as non root? Jörg -- EMail:[EMAIL PROTECTED] (home) Jörg Schilling D-13353 Berlin [EMAIL PROTECTED](uni) If you don't have iso-8859-1 [EMAIL PROTECTED](work) chars I am J"org Schilling URL: http://www.fokus.fraunhofer.de/usr/schilling ftp://ftp.berlios.de/pub/schily -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Testers for OpenBSD & NetBSD needed
Jacob Meuser <[EMAIL PROTECTED]> wrote: > > Aug 19 18:34:35 thebetteros /bsd: pciide0:1: bogus intr > > Aug 19 18:39:45 thebetteros /bsd: cd0(pciide0:1:0): timeout > > Aug 19 18:39:45 thebetteros /bsd: type: atapi > > Aug 19 18:39:45 thebetteros /bsd: c_bcount: 1181 > > Aug 19 18:39:45 thebetteros /bsd: c_skip: 0 > > Aug 19 18:39:45 thebetteros /bsd: cd0: transfer error, downgrading to > > DMA mode 2 > > Aug 19 18:39:45 thebetteros /bsd: cd0(pciide0:1:0): using PIO mode 4, > > DMA mode 2 > > Aug 19 18:39:46 thebetteros /bsd: pciide0:1: bogus intr > > check your cables. out of curiousity, what is your IDE controller? > downgrading dma modes and bogus interrupts are highly unlikely the > fault of cdda2wav, and I'm pretty sure this has nothing to do with > whether it was installed suid or not. this looks like a bad > interaction with your IDE controller and your DVD burner. No, it is not caused by cabling but by a defective ATA (DMA) driver in the kernel. The driver is unable to handle odd transfer counts correctly. Jörg -- EMail:[EMAIL PROTECTED] (home) Jörg Schilling D-13353 Berlin [EMAIL PROTECTED](uni) If you don't have iso-8859-1 [EMAIL PROTECTED](work) chars I am J"org Schilling URL: http://www.fokus.fraunhofer.de/usr/schilling ftp://ftp.berlios.de/pub/schily -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Testers for OpenBSD & NetBSD needed
On Fri, Aug 20, 2004 at 12:12:56AM +0200, Joerg Schilling wrote: > Jacob Meuser <[EMAIL PROTECTED]> wrote: > > > On Thu, Aug 19, 2004 at 11:46:37AM +0200, Joerg Schilling wrote: > > > Hi all, > > > > > > as it turned out far too late that cdda2wav did not > > > work when installed suid root on FreeBSD, I did include a dirty hack > > > into cdda2wav/setuid.c in order to honor the fact that FreeBSD now > > > did try to implement POSIX uid handling on top of the old > > > BSD ID swapping method. > > > > > > In order to make sure that cdda2wav is not hitting the same problems > > > on OpenBSD & NetBSD, I would need people who like to test a suid > > > cdda2wav on OpenBSD & NetBSD systems. > > > > Works fine for me on OpenBSD 3.6-beta. > > Did you install cdda2wav suid root and call it as non root? Of course, and only root had any privileges on /dev/rcd0c. Ripped a track without any issue. -J worked fine also. -- <[EMAIL PROTECTED]> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Testers for OpenBSD & NetBSD needed
On Fri, Aug 20, 2004 at 12:11:31AM +0200, Joerg Schilling wrote: > >Frederick Page <[EMAIL PROTECTED]> wrote: > >> > >> Aug 19 18:34:35 thebetteros /bsd: pciide0:1: bogus intr > >> Aug 19 18:39:45 thebetteros /bsd: cd0(pciide0:1:0): timeout > >> Aug 19 18:39:45 thebetteros /bsd: type: atapi > >> Aug 19 18:39:45 thebetteros /bsd: c_bcount: 1181 > >> Aug 19 18:39:45 thebetteros /bsd: c_skip: 0 > >> Aug 19 18:39:45 thebetteros /bsd: cd0: transfer error, downgrading to > >> DMA mode 2 > >> Aug 19 18:39:45 thebetteros /bsd: cd0(pciide0:1:0): using PIO mode 4, > >> DMA mode 2 > >> Aug 19 18:39:46 thebetteros /bsd: pciide0:1: bogus intr > >> > >> I attach OUT.TXT gzipped. > > Your problem is caused by a OpenBSD kernel bug. The kernel does not handle > odd DMA counts for ATA correctly. really? you have made a PR about this? I don't get anything matching 'schilling' or 'schilly' or 'schily'. but I do see http://www.openbsd.org/faq/faq14.html#pciideErr -- <[EMAIL PROTECTED]> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Testers for OpenBSD & NetBSD needed
On Fri, Aug 20, 2004 at 12:47:29AM +0200, Joerg Schilling wrote: > Jacob Meuser <[EMAIL PROTECTED]> wrote: > > > > Your problem is caused by a OpenBSD kernel bug. The kernel does not handle > > > odd DMA counts for ATA correctly. > > > > really? you have made a PR about this? I don't get anything matching > > 'schilling' or 'schilly' or 'schily'. > > > > but I do see http://www.openbsd.org/faq/faq14.html#pciideErr > > What is a 'PR'? Probelm Report -> http://www.openbsd.org/report.html > I just found this bug a few minutes ago when looking at Frederick's > log files. strange I have never run into that -- <[EMAIL PROTECTED]> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Testers for OpenBSD & NetBSD needed
Jacob Meuser <[EMAIL PROTECTED]> wrote: > > Your problem is caused by a OpenBSD kernel bug. The kernel does not handle > > odd DMA counts for ATA correctly. > > really? you have made a PR about this? I don't get anything matching > 'schilling' or 'schilly' or 'schily'. > > but I do see http://www.openbsd.org/faq/faq14.html#pciideErr What is a 'PR'? I just found this bug a few minutes ago when looking at Frederick's log files. Jörg -- EMail:[EMAIL PROTECTED] (home) Jörg Schilling D-13353 Berlin [EMAIL PROTECTED](uni) If you don't have iso-8859-1 [EMAIL PROTECTED](work) chars I am J"org Schilling URL: http://www.fokus.fraunhofer.de/usr/schilling ftp://ftp.berlios.de/pub/schily -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Testers for OpenBSD & NetBSD needed
Hi Joerg, Joerg Schilling wrote on Fri, Aug 20 2004: >>> Aug 19 18:39:45 thebetteros /bsd: cd0: transfer error, downgrading to >>> DMA mode 2 >Your problem is caused by a OpenBSD kernel bug. The kernel does not >handle odd DMA counts for ATA correctly. Are you sure it's a kernel-bug? (Please see below). >If you install cdda2wav suid root and call it as non root user, does >it work? To my total astonishment: YES. After suid it also works as root! -r-sr-xr-x 1 root bin261K Aug 19 18:18 /usr/local/bin/cdda2wav >In order to avoid the Kernel bug from preventing you to do the test, >check the number of tracks on the CD. It it was a CD with an even >number of tracks, use one with an odd number of tracks or vice versa. Did that a couple of times, but the number of tracks seems to be irrelevant. After suid'ing cdda2wav it works with each and every CD now, as root and as fpage. I'm totally confused about what might be going on here :-( Thanks for the hints, now I got a working cdda2wav on OpenBSD ;-) Kind regards Frederick -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Testers for OpenBSD & NetBSD needed
On Fri, Aug 20, 2004 at 03:18:52AM +0200, Frederick Page wrote: > Hi Joerg, > > Joerg Schilling wrote on Fri, Aug 20 2004: > > >>> Aug 19 18:39:45 thebetteros /bsd: cd0: transfer error, downgrading to > >>> DMA mode 2 > > >Your problem is caused by a OpenBSD kernel bug. The kernel does not > >handle odd DMA counts for ATA correctly. > > Are you sure it's a kernel-bug? (Please see below). > > >If you install cdda2wav suid root and call it as non root user, does > >it work? > > To my total astonishment: YES. After suid it also works as root! > > -r-sr-xr-x 1 root bin261K Aug 19 18:18 /usr/local/bin/cdda2wav > > >In order to avoid the Kernel bug from preventing you to do the test, > >check the number of tracks on the CD. It it was a CD with an even > >number of tracks, use one with an odd number of tracks or vice versa. > > Did that a couple of times, but the number of tracks seems to be > irrelevant. After suid'ing cdda2wav it works with each and every CD > now, as root and as fpage. > > I'm totally confused about what might be going on here :-( > > Thanks for the hints, now I got a working cdda2wav on OpenBSD ;-) Well, I have been using cdda2wav on OpenBSD for years ... and I have been the cdrtools port maintainer for a good while now. I've only seen such problems in two situations: 1) bad hardware: drive, cable, or IDE device (95%) 2) cd and hdd with heavy i/o on same IDE channel (5%) Of course, you shouldn't be able to use /dev/rcd?c at all unless either you're part of the operator group, or the programs that access the drive are suid root or sgid operator, or you change permissions on the device. I usually add my user to the operator group and then chmod g+w /dev/rcd0c so I can also write to it. -- <[EMAIL PROTECTED]> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Testers for OpenBSD & NetBSD needed
Frederick Page <[EMAIL PROTECTED]> wrote: > >>> Aug 19 18:39:45 thebetteros /bsd: cd0: transfer error, downgrading to > >>> DMA mode 2 > > >Your problem is caused by a OpenBSD kernel bug. The kernel does not > >handle odd DMA counts for ATA correctly. > > Are you sure it's a kernel-bug? (Please see below). I am very sure. ATA is a 16 bit system, so transfering an odd number of bytes is a bit tricky. In your case, READ FULL TOC did try to transfer an odd number of bytes and you observe a timeout. This is exactly what I would expect for a broken implementation. > >If you install cdda2wav suid root and call it as non root user, does > >it work? > > To my total astonishment: YES. After suid it also works as root! What you write is the complete contrary to what I like to know. I like to know whether it works ad _non_ root after it has been installed suid root. > -r-sr-xr-x 1 root bin261K Aug 19 18:18 /usr/local/bin/cdda2wav > > >In order to avoid the Kernel bug from preventing you to do the test, > >check the number of tracks on the CD. It it was a CD with an even > >number of tracks, use one with an odd number of tracks or vice versa. > > Did that a couple of times, but the number of tracks seems to be > irrelevant. After suid'ing cdda2wav it works with each and every CD > now, as root and as fpage. OK, now you write what I like to know > I'm totally confused about what might be going on here :-( Did you do a cold start of the machine to make sure that there is DMA again? It may be thye the PIO implementation does odd byte transfers correctly Jörg -- EMail:[EMAIL PROTECTED] (home) Jörg Schilling D-13353 Berlin [EMAIL PROTECTED](uni) If you don't have iso-8859-1 [EMAIL PROTECTED](work) chars I am J"org Schilling URL: http://www.fokus.fraunhofer.de/usr/schilling ftp://ftp.berlios.de/pub/schily -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: cdrtools-2.01a37 ready
Volker Kuhlmann> > Let's not forget that there are 2 separate changes SuSE makes: the DVD > addons (daft, and I've told them so), About that i don't care much. growisofs covers all my DVD needs. I understand, though, that these DVD "addons" are an unfriendly gesture towards the only person who provides CD burning on Linux and several other OSs. I mean, if one wants to show Joerg the middle finger then please not by constantly forking his alpha versions but by bringing a decent Unix style write device for data CD into the standard kernel - just for an example. > the SuSE security team (who have a bigger reputation to lose) There is a security team with SuSE ? (chuckle) Well, at least somebody must be in charge to write advisories like this one : Announcement-ID:SuSE-SA:2004:011 Date: Thursday, May 6th 2004 22:30 MEST Affected products: SUSE LINUX 9.1 Personal Edition Live CD Vulnerability Type: remote root access Severity (1-10):8 SuSE default package: yes http://www.linuxsecurity.com/advisories/suse_advisory-4305.html In that light and if i have to choose ... ... i'll bet on Joerg. :)) > and the removal of the suid requirement. Exactly with that, SuSE 9.0 failed. I would not call it "removal of the suid requirement" if it does not work for normal users with or without suid. You take an old version from SuSE 7.2 and all is well. My software is a parasite on cdrtools and growisofs. I am strongly interested in a low percentage of non-working installations of both. The cdrtools hacks produce a substantial amount of those, i can assure you. > Compiles OOTB on 9.1 Compiler problems come and go with versions of cdrecord and Linux. For a while i could safely point towards the distribution binaries. But meanwhile i can only wish good luck. Have a nice day :) Thomas -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]