A recent posting to the lkml noted a kernel panic in the most recent 2.4 kernel caused when reading a copy-protected audio CD. The IDE maintainer, Andre Hedrick was quick to note that while a panic should not happen under normal circumstances, certain devices can detect an attempt to copy a copy-protected disk, and will attempt to stop the action, even if it means crashing the system.

Andre added, "Asking me to make it so you or anyone else can bypass copy-content-protection is out of the question. If you do not ask the device to do bad things, then it will not do bad things back at you." He went on to add, "I agree this is silly but the DMCA/RIAA/MPAA and company are out to hurt people and their rights. [...] These people are sneaky and in the dark." The problematic drive in question was manufactured by Sony, a company notorious for its "aggressive anti-piracy stance". Andre went so far as to suggest, "Make a note, DON'T BUY SONY CDRW Products."

From: Christian Borntraeger
To: Andre Hedrick, linux-kernel mailing list
Subject: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sat, 19 Oct 2002 20:41:45 +0100

PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11

Trying to read a copy-protected audio CD (I think Cactus Data Shield 200) with
readcd I get a reproduceable kernel panic in 2.4.19 and 2.4.20-pre11.
As I can reproduce the bug feel free to ask me any further question, if possible
cc me. (Otherwise it will take a little longer)
The information I gathered is below.
some more files (system.map etc) are on 
http://www.cborntraeger.de/linux/panic.tgz

cheers

Christian
 
-------------------------------------------------
I send the panic through ksymoops:
-------------------------------------------------

ksymoops 2.4.5 on i686 2.4.20-pre11.  Options used
     -V (default)
     -k ksyms (specified)
     -l modules (specified)
     -o /lib/modules/2.4.20-pre11/ (default)
     -m System.map (specified)

Warning (compare_maps): ksyms_base symbol __io_virt_debug_R__ver___io_virt_debug not found in System.map.  
Ignoring ksyms_base entry
Unable to handle Kernel Null pointer dereference at virtual address 00000018
f08968f8
CPU 0
EFLAGS: 00010286
eax: ee88e1c4 ebx: c0322f24 ecx: ebedcb64 edx: 0000005a
esi: 00000000 edi: 00000040 ebp: c02a9e74 esp: c02a9e50
00000000 c01da5dc 000003e8 00000046 00000082 ee88e1c4 00000000 c0322f24
00000040 c02a9eac c01da999 c0322f24 00000000 00000000 00000000 000001f4
c02a9ec0 00000000 c0322da4 00000001 c0322f24 c18f218c c0322da4 c02a9ed0
Call Trace: [] [] [] [] [] 
            [] [] [] [] []
            [] [] [] [] []
            []
Code: 8b 56 18 c7 45 ec 00 00 00 00 89 70 04 8b 7e 0c 8b 46 1c c7
Using defaults from ksymoops -t elf32-i386 -a i386

>>eax; ee88e1c4 
>>ebx; c0322f24 
>>ecx; ebedcb64 
>>ebp; c02a9e74 
>>esp; c02a9e50 

Trace; c01da5dc 
Trace; c01dab78 
Trace; c01dae54 
Trace; c01dad60 
Trace; c0124ba3 
Trace; c012432d 
Trace; c0120c0f 
Trace; c0120b26 
Trace; c0120957 
Trace; c010a74d 
Trace; c010cc18 
Trace; c01070c3 
Trace; c01145e6 
Trace; c0114530 
Trace; c0107132 
Trace; c0105000 

Code;  00000000 Before first symbol
00000000 :
Code;  00000000 Before first symbol
   0:   8b 56 18                  mov    0x18(%esi),%edx
Code;  00000003 Before first symbol
   3:   c7 45 ec 00 00 00 00      movl   $0x0,0xffffffec(%ebp)
Code;  0000000a Before first symbol
   a:   89 70 04                  mov    %esi,0x4(%eax)
Code;  0000000d Before first symbol
   d:   8b 7e 0c                  mov    0xc(%esi),%edi
Code;  00000010 Before first symbol
  10:   8b 46 1c                  mov    0x1c(%esi),%eax
Code;  00000013 Before first symbol
  13:   c7 00 00 00 00 00         movl   $0x0,(%eax)

Kernel Panic, Aiee..

1 warning issued.  Results may not be reliable.
---------------------------------------------------
/proc/modules:
---------------------------------------------------

usbmouse                2264   0 (unused)
keybdev                 2080   0 (unused)
mousedev                4244   1
hid                    19300   0 (unused)
input                   3584   0 [usbmouse keybdev mousedev hid]
ide-scsi                8816   0
scsi_mod               56276   1 [ide-scsi]
ide-cd                 30148   0
cdrom                  28608   0 [ide-cd]
usb-uhci               23020   0 (unused)
usbcore                60960   1 [usbmouse hid usb-uhci]
rtc                     6940   0 (autoclean)

-----------------------------------------------
ver-linux
----------------------------------------------- 
Linux cubus.mynet 2.4.20-pre11 #3 Sat Oct 19 18:50:52 BST 2002 i686 unknown unknown GNU/Linux
 
Gnu C                  3.2
Gnu make               3.79.1
util-linux             2.11u
mount                  2.11u
modutils               2.4.19
e2fsprogs              1.27ea
reiserfsprogs          3.6.3
PPP                    2.4.1
Linux C Library        2.2.5
Dynamic linker (ldd)   2.2.5
Procps                 2.0.7
Net-tools              1.60
Console-tools          0.2.3
Sh-utils               2.0.15
Modules Loaded         usbmouse keybdev mousedev hid input ide-scsi scsi_mod ide-cd cdrom 
                       usb-uhci usbcore rtc

-------------------------------------------------
/proc/scsi/scsi
-------------------------------------------------
Attached devices: 
Host: scsi0 Channel: 00 Id: 00 Lun: 00
  Vendor: SONY     Model: CD-RW  CRX140E   Rev: 1.0p
  Type:   CD-ROM                           ANSI SCSI revision: 02



From: Andre Hedrick
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sat, 19 Oct 2002 15:02:46 -0700 (PDT)

So could you ask the question a little more blunt?

"Gee, I am trying to break a US Law on content protection, would you be my
enabler?  Don't worry, it only effects the US, and we are in a public
forum.  Also, do you prefer gray or black in your future pin stripped
suit?"

Now if it works in 2.4.18, use it.

If there is a technical issue like loading both ide-cd and ide-scsi on the
same device, that is a problem.  There have been no changes to ide-scsi in
2.4.19 by me directly.  Currently all updates to "stable" are released
through Alan Cox, come to think of it so are any releases to "unstable".

Have a little more sense about asking about "copy-protected" media in a
public forum, and DON'T !!  Regardless if I could answer the question, you
have placed me in a position where I can not now.

I should have ignored this and issued a reply.

Andre Hedrick
LAD Storage Consulting Group



From: Andreas Steinmetz
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sun, 20 Oct 2002 00:32:07 +0200

Andre Hedrick wrote:
> So could you ask the question a little more blunt?
> 
> "Gee, I am trying to break a US Law on content protection, would you be my
> enabler?  Don't worry, it only effects the US, and we are in a public
> forum.  Also, do you prefer gray or black in your future pin stripped
> suit?"
> 
I'm not taking any side in this, but:

US law != World law (and hopefully this will stay so for a long time)

Looking at the originators address and name he's from germany. FYI: 
There's no provision in the law here that denies you personal copies of 
copy protected contents assuming you own the original.
In fact you even pay for this use when buying emtpy media, regardless if 
for data backup or copying.
So what may be illegal in the US is legal in other countries and 
royalties are already taken care of (by law).
As I do assume that in case of this posting the actual target was a 
personal record compilation for mobile use you should not accuse people 
lightly of breaking the law when you seemingly don't know what law applies.
This doesn't mean that you have to take any action as it seems you live 
in a country where any private software development already tends to be 
illegal.
To state it again: I'm strictly against illegal copying but what you 
state here is nonsense for a lot of countries.
-- 
Andreas Steinmetz



From: Alan Cox
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	21 Oct 2002 13:22:38 +0100

On Sat, 2002-10-19 at 23:32, Andreas Steinmetz wrote:
> Looking at the originators address and name he's from germany. FYI: 
> There's no provision in the law here that denies you personal copies of 
> copy protected contents assuming you own the original.

Germany is obliged by the european union to pass such a law very soon.
The German government may not be full of corrupt corporate lobbyists but
the lobbyists own their masters and since they sold their souls to
europe on this matter they have no option any more



From: Andre Hedrick
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sat, 19 Oct 2002 15:49:37 -0700 (PDT)

I am in California, USA where the Law exists.
Please note the question in quotes was directed back at me.
I assumed nothing about the country of origin from the original poster.
Please respect the point of not putting "me" in a compromised position.

Andre Hedrick
LAD Storage Consulting Group



From: Brian Gerst
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sat, 19 Oct 2002 18:39:48 -0400

Andre Hedrick wrote:
> So could you ask the question a little more blunt?
> 
> "Gee, I am trying to break a US Law on content protection, would you be my
> enabler?  Don't worry, it only effects the US, and we are in a public
> forum.  Also, do you prefer gray or black in your future pin stripped
> suit?"

Attempting to read a "defective" disc should never, ever, cause a kernel 
oops.  Whether it succeeds or not is irrelevant.

--
				Brian Gerst



From: Andre Hedrick
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sat, 19 Oct 2002 15:54:46 -0700 (PDT)

On Sat, 19 Oct 2002, Brian Gerst wrote:

> Andre Hedrick wrote:
> > So could you ask the question a little more blunt?
> > 
> > "Gee, I am trying to break a US Law on content protection, would you be my
> > enabler?  Don't worry, it only effects the US, and we are in a public
> > forum.  Also, do you prefer gray or black in your future pin stripped
> > suit?"
> 
> Attempting to read a "defective" disc should never, ever, cause a kernel 
> oops.  Whether it succeeds or not is irrelevant.

Please point out where in the original post, the referrence to "defective"
media.  If this would have been the case, your point it valid.  If I
missed something, thus am wrong, I will admit to being wrong.

Warning (compare_maps): ksyms_base symbol
__io_virt_debug_R__ver___io_virt_debug not found in System.map.
  Ignoring ksyms_base entry
Unable to handle Kernel Null pointer dereference at virtual address 00000018
f08968f8

First how can the trace be meaningful if ksymoops detects a miscompare on
the system map?

Cheers,

Andre Hedrick
LAD Storage Consulting Group



From: Brian Gerst
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sat, 19 Oct 2002 19:19:18 -0400

Andre Hedrick wrote:
 > On Sat, 19 Oct 2002, Brian Gerst wrote:
 >
 >
 >>Andre Hedrick wrote:
 >>
 >>>So could you ask the question a little more blunt?
 >>>
 >>>"Gee, I am trying to break a US Law on content protection, would you 
be my
 >>>enabler?  Don't worry, it only effects the US, and we are in a public
 >>>forum.  Also, do you prefer gray or black in your future pin stripped
 >>>suit?"
 >>
 >>Attempting to read a "defective" disc should never, ever, cause a kernel
 >>oops.  Whether it succeeds or not is irrelevant.
 >
 >
 > Please point out where in the original post, the referrence to 
"defective"
 > media.  If this would have been the case, your point it valid.  If I
 > missed something, thus am wrong, I will admit to being wrong.

Copy-protected discs abuse the CD standards to the point where CDROM
drives consider them defective and can't/won't read them, while less
intelligent devices can.  Trying to read one of these discs should only 
cause the kernel to return an error, never an oops.

--
				Brian Gerst



From: Andre Hedrick
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sat, 19 Oct 2002 16:44:32 -0700 (PDT)

Did you consider the attempt to copy may invoke the device to attempt to
jam and crash the transport?  Did you consider the SCSI layer may not be
capable of handling the task management command? 

They are not broken if you have a device which is "Copy-protected" aware.
It will attempt to thwart your operation request and that can include
crashing a system.

> Copy-protected discs abuse the CD standards to the point where CDROM
> drives consider them defective and can't/won't read them, while less
> intelligent devices can.  Trying to read one of these discs should only
> cause the kernel to return an error, never an oops.

You admit that older dumber devices just work.
So much for new and improved, go find the old and lousy that works.

Asking me to make it so you or anyone else can bypass
copy-content-protection is out of the question.  If you do not ask the
device to do bad things, then it will not do bad things back at you.

If your memory is short, recall I was the only person to stand up and take
issue about having CPRM stuffed into your harddrives by default.

Make a note, DON"T BUY SONY CDRW Products. 

Now if you are serious about want to fix the issue and not rant about
issues that have no meaning because they are your opinion on how the world
should work as it relates to hardware, then we can move on.

Regards,

Andre Hedrick
LAD Storage Consulting Group



From: Brian Gerst
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sat, 19 Oct 2002 20:14:22 -0400

Andre Hedrick wrote:
>>Copy-protected discs abuse the CD standards to the point where CDROM
>>drives consider them defective and can't/won't read them, while less
>>intelligent devices can.  Trying to read one of these discs should only
>>cause the kernel to return an error, never an oops.
> 
> 
> You admit that older dumber devices just work.
> So much for new and improved, go find the old and lousy that works.

Audio-only CD players are cheap and dumb.  The standard audio CD format 
is not complex, and certain parts of the disc that are needed for data 
CDs are ignored by audio players.  This is where the copy-protected 
discs use false data to confuse CDROM drives.

> Asking me to make it so you or anyone else can bypass
> copy-content-protection is out of the question.  If you do not ask the
> device to do bad things, then it will not do bad things back at you.

Nobody asked you to bypass the protection, only to sanely error out when 
it is found.  Refusing to read the disk is ok, but allowing the system 
to crash is not.

--
				Brian Gerst



From: Andre Hedrick
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sat, 19 Oct 2002 17:41:03 -0700 (PDT)

On Sat, 19 Oct 2002, Brian Gerst wrote:

> > Asking me to make it so you or anyone else can bypass
> > copy-content-protection is out of the question.  If you do not ask the
> > device to do bad things, then it will not do bad things back at you.
> 
> Nobody asked you to bypass the protection, only to sanely error out when 
> it is found.  Refusing to read the disk is ok, but allowing the system 
> to crash is not.

I thought I specified what was need to decode the issue, maybe since there
are two multiple threads now I have lost track of which one I am
responding.  Thus I will repeat in this thread.

True, however since I suspect the device was attempting to thwart and
crash the system, until a trace of the sense data returns from the device
and the re-action of the kernel to those target responses, not much can be
done to prevent such a crash.

Cheers,

Andre Hedrick
LAD Storage Consulting Group



From: Henning P. Schmiedehausen
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sun, 20 Oct 2002 14:59:09 +0000 (UTC)

Andre Hedrick writes:

>They are not broken if you have a device which is "Copy-protected" aware.
>It will attempt to thwart your operation request and that can include
>crashing a system.

Which in my country would result in a direct lawsuit ($303b of the
german law code: "Anschlaege auf die Datenverarbeitung durch
Veraenderung oder Vernichtung von Computerdaten, Datentraegern oder
Anlagen") against the vendor of the drive, the computer and the CD.

Thank goodness, I'm living in a country where "law enforcement at gun
point" is not usual.

	Regards
		Henning

-- 
Dipl.-Inf. (Univ.) Henning P. Schmiedehausen       -- Geschaeftsfuehrer
INTERMETA - Gesellschaft fuer Mehrwertdienste mbH     [email blocked]

Am Schwabachgrund 22  Fon.: 09131 / 50654-0   [email blocked]
D-91054 Buckenhof     Fax.: 09131 / 50654-20   



From: Benjamin Herrenschmidt
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Mon, 21 Oct 2002 00:54:30 +0200

>Asking me to make it so you or anyone else can bypass
>copy-content-protection is out of the question.  If you do not ask the
>device to do bad things, then it will not do bad things back at you.

Andre, your argument is pointless, the right answer is indeed
what you wrote below, that is don't buy those ;) Though still,
I agree we shouldn't oops.

Also one problem here is that machines like all new macs can only
_play_ CDs by reading their audio datas via ATAPI and sending those
to the sound chip, they have no analog output on the drive and
if they had, Apple didn't wire it to the sound chip.
So on these configs, there is nothing wrong even in the US wanting
to _play_ a CD you have legally purchased...

Trying to do that shouldn't result into a kernel oops :)

So the point is that even if the drive fucks up, it should
be possible to not crash, detect the fuckup, error the read
request (at least), and eventually reset the drive if it
needs that to be put back in a sane state.

I don't know what exactly is going on at the transport level
with those so-called "copy protected" CDs though, I'm afraid
only you knows exactly what's up there ;) I could investigate,
but I have no intend spending money on a copy-protected CD :)

>If your memory is short, recall I was the only person to stand up and take
>issue about having CPRM stuffed into your harddrives by default.
>
>Make a note, DON"T BUY SONY CDRW Products. 
>
>Now if you are serious about want to fix the issue and not rant about
>issues that have no meaning because they are your opinion on how the world
>should work as it relates to hardware, then we can move on.



From: Toon van der Pas
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sun, 20 Oct 2002 01:20:57 +0200

On Sat, Oct 19, 2002 at 03:54:46PM -0700, Andre Hedrick wrote:
> On Sat, 19 Oct 2002, Brian Gerst wrote:
> > 
> > Attempting to read a "defective" disc should never, ever, cause a
> > kernel oops.  Whether it succeeds or not is irrelevant.
> 
> Please point out where in the original post, the referrence to
> "defective" media.  If this would have been the case, your point it
> valid.  If I missed something, thus am wrong, I will admit to being
> wrong.

AFAICS you missed something indeed:
Attempts to read copy-protected media should also never result in a kernel oops.

Regards,
Toon.
-- 
 /"                             |
  /     ASCII RIBBON CAMPAIGN   |  "Who is this General Failure, and
  X        AGAINST HTML MAIL     |   what is he doing on my harddisk?"
 / 



From: Andre Hedrick
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sat, 19 Oct 2002 17:01:13 -0700 (PDT)

On Sun, 20 Oct 2002, Toon van der Pas wrote:

> On Sat, Oct 19, 2002 at 03:54:46PM -0700, Andre Hedrick wrote:
> > On Sat, 19 Oct 2002, Brian Gerst wrote:
> > > 
> > > Attempting to read a "defective" disc should never, ever, cause a
> > > kernel oops.  Whether it succeeds or not is irrelevant.
> > 
> > Please point out where in the original post, the referrence to
> > "defective" media.  If this would have been the case, your point it
> > valid.  If I missed something, thus am wrong, I will admit to being
> > wrong.
> 
> AFAICS you missed something indeed:
> Attempts to read copy-protected media should also never result in a kernel oops.

True, however since I suspect the device was attempting to thwart and
crash the system, until a trace of the sense data returns from the device
and the re-action of the kernel to those target responses, not much can be
done to prevent such a crash.

Again, the world of copy-protection has no boundaries.

There is a product out there called a tether (sp).

By design it will white screen and then blue screen the MicroSoft
environment.  If you continue to attempt to violate the ACL of the
device/media combination it will become a tattle-tail and send email to a
specified server and include all the evidence to charge you with a crime.

I agree this is silly but the DMCA/RIAA/MPAA and company are out to hurt
people and their rights.

When was the last time you called the MPAA's general council and tell the
lead attorney and clearly stated you would oppose any new changes into the
a specification regardless?  I did it about a year ago.

Now it is only going to get worse.

Recall SCSI via MMC is tainted.
Recall SAS is being pushed into SATA as SATA II.
Now since SATA II will be effectively under SAS which is under SAM
overseen by T10 and STA, this all translate to SATA/SATAPI devices will
totally polluted like SCSI is today.

Now is there a way to win at this game, I am working on it.

These people are sneaky and in the dark.

Cheers,

Andre Hedrick
LAD Storage Consulting Group



From: Henning P. Schmiedehausen
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sun, 20 Oct 2002 15:01:00 +0000 (UTC)

Andre Hedrick writes:

>There is a product out there called a tether (sp).

>By design it will white screen and then blue screen the MicroSoft
>environment.  If you continue to attempt to violate the ACL of the
>device/media combination it will become a tattle-tail and send email to a
>specified server and include all the evidence to charge you with a crime.

These are two crimes, each up to three years in prison in
germany. ($202a, $303b). Vendors might not want to do this...

	Regards
		Henning



From: Christian Borntraeger
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sun, 20 Oct 2002 11:14:33 +0100

On Sat, 19 Oct 2002 15:02:46 -0700 (PDT), Andre Hedrick wrote:
>Now if it works in 2.4.18, use it.

I havent tried it yet with 2.4.18. So it might be a general problem.

>If there is a technical issue like loading both ide-cd and ide-scsi

no there are 2 drives. hdc is with ide-cd and hdd with ide-scsi

>Have a little more sense about asking about "copy-protected" media
>in a
>public forum, and DON'T !! _Regardless if I could answer the

I just do NOT care about not being able to read this CD. 
But I care a lot about, when a non-root-user is able to crash the kernel. 

If this happens with a copy protected CD it is probable that a well s
cratched CD has the same effect.

I have absolutely no problem with being not able to read this CD if changing
 your code will break US law.

and YES I understand your position knowing - now knowing you live in california
 - that you dont want to change the code.
I cc'ed you because you are the IDE maintainer of IDE in 2.4. 

If you dont care - fine. Whether somebody else takes care or I have to live 
with this situation.

cheers

Christian



From: Bill Davidsen
Subject: Re: PROBLEM: ide-related kernel panic in 2.4.19 and 2.4.20-pre11
Date: 	Sun, 20 Oct 2002 18:49:56 -0400 (EDT)

On Sat, 19 Oct 2002, Andre Hedrick wrote:

> 
> So could you ask the question a little more blunt?
> 
> "Gee, I am trying to break a US Law on content protection, would you be my
> enabler?  Don't worry, it only effects the US, and we are in a public
> forum.  Also, do you prefer gray or black in your future pin stripped
> suit?"

Unless the rules have changed VERY recently, making a copy of legally
owned music for personal use, such as in the car, MP3 player, etc, is
called "fair use" and is totally legal.

I have no problem with copy protected CDs, as long as they are labeled
clearly so you know they aren't in CDDA *book format.

-- 
bill davidsen
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.

Related links:

• Google archive of above thread.