help on ss extraction with plscsi

[Pandor]

As a linux user, i've started writing a backup script to make life easier.
Now i've come to the part where I have to extract the ss with plscsi.

Windows users seem to use dvdinfo pro for this by isueing the following cdb commands:

AD 00 FF 02 FD FF FE 00 08 00 01 C0
AD 00 FF 02 FD FF FE 00 08 00 03 C0
AD 00 FF 02 FD FF FE 00 08 00 05 C0
AD 00 FF 02 FD FF FE 00 08 00 07 C0

How does dvdinfo pro handle this?
It's seems to be fixed to a 2048b buffer, but contains the output of all 4 commands? Does it merge the ouput after eacht comand?

And now the real question, how could this be done with plscsi?
Do i need to extrac to 4 different files and then merge those?

any help would be much appreciated.

[Pandor]

Code:

./plscsi /dev/sr0 -x "AD 00 FF 02 FD FF FE 00 08 00 01 C0" -i 2048
./plscsi /dev/sr0 -x "AD 00 FF 02 FD FF FE 00 08 00 03 C0" -i 2048
./plscsi /dev/sr0 -x "AD 00 FF 02 FD FF FE 00 08 00 05 C0" -i 2048
./plscsi /dev/sr0 -x "AD 00 FF 02 FD FF FE 00 08 00 07 C0" -i 2048 -t ss.bin

Seems to do the trick.
Haven't tested to ss, but it looks okay in a hex compare.

[Pandor]

just in case some linux users might be interested:

Here are my findings for ripping games with a sammy SH-D162C or TS-H943:

Code:

rip wxripper style (SH-D162C - mode2):
sg_dd blk_sgio=1 bs=2048 count=3697696 coe=1 verbose=2 if=/dev/hda of=xbox360.iso

rip xtreme style (SH-D162C_KREON - mode1 & TS-H943):
sg_dd blk_sgio=1 bs=2048 count=3567872 coe=1 verbose=2 if=/dev/hda of=xbox360.iso

rip ss.bin:
./plscsi /dev/sr0 -x "AD 00 FF 02 FD FF FE 00 08 00 01 C0" -i 2048
./plscsi /dev/sr0 -x "AD 00 FF 02 FD FF FE 00 08 00 03 C0" -i 2048
./plscsi /dev/sr0 -x "AD 00 FF 02 FD FF FE 00 08 00 05 C0" -i 2048
./plscsi /dev/sr0 -x "AD 00 FF 02 FD FF FE 00 08 00 07 C0" -i 2048 -t ss.bin

rip PFI (Physical Format Information) Sector:
./plscsi /dev/sr0 -x "AD 00 00 00 00 00 00 00 08 00 00 00" -i 2048 -t pfi.bin
*./plscsi /dev/sr0 -x "AD 00 00 00 00 00 00 00 08 04 00 00" -i 2052  -t "" | dd ibs=1 obs=2048 skip=4 of=pfi.bin

rip DMI (Disk Manufacturing Information) Sector:
./plscsi /dev/sr0 -x "AD 00 00 00 00 00 00 04 08 00 00 00" -i 2048 -t dmi.bin
*./plscsi /dev/sr0 -x "AD 00 00 00 00 00 00 04 08 04 00 00" -i 2052 -t "" | dd ibs=1 obs=2048 skip=4 of=dmi.bin

inject SS (wxripper):
dd bs=2048 seek=129823 conv=notrunc if=ss.bin of=xbox360.iso

inject pfi/dmi (wxripper):
dd bs=2048 seek=129821 conv=notrunc if=pfi.bin of=xbox360.iso
dd bs=2048 seek=129822 conv=notrunc if=dmi.bin of=xbox360.iso


build xtreme iso:
cat video360.iso pfi.bin dmi.bin ss.bin > xbox360.iso


burn:
growisofs -use-the-force-luke=dao -use-the-force-luke=break:1913760  -dvd-compat -speed=2 -Z /dev/hdx=xbox360.iso



notes:
=========
* fix for SH-D162C

I might script this all toghether if people are interested.


If anyone could help me get some tech info on ss extraction with the Hitachi, it would be much appreciated.

[probutus]

Have seen this thread too late! shame on me ;-)

Great work!

[Pandor]

Just some minor corrections.

Don't use:

Code:

sg_dd blk_sgio=1 bs=2048 count=3697696 coe=1 verbose=2 if=/dev/hda of=xbox360.iso

it's better to use:

Code:

sg_dd blk_sgio=1 bs=2048 count=3697696 iflag=coe verbose=2 if=/dev/hda of=xbox360.iso

If coe is set globally, it will also continue on write errors, which, is not advisable. so use iflag=coe instead.
count flag is optional, as it will be calculated when omited. I included it for completeness.

Another thing i've noticed is that the xtreme build command is not complete,
It should have been:

Code:

cat video360.iso pfi.bin dmi.bin ss.bin game.iso> xbox360.iso

DD can also be used, in the same way as sg_dd to rip a game, but it's a tad slower.
iso's made with dd don't compare to iso's made with sg_dd (md5 don't match). I think they may have different ways of dealing with bad blocks.
But both produce working backups.

command to use for dd:

Code:

dd bs=2048 conv=noerror,sync if=/dev/sr0 of=xbox360.iso

noerror= continue on error, sync = on bad (input) blocks pad with zero till blocksize is met.



Now, if there's anyone who could supply some info on ss ripping with the hitachi, it would be grately appreciated..

[abuali]

Great work Pandor! Thanks a lot for your work. I tried and failed burning, but since that was some time back I don't have the error message anymore, and my DVDRW is currently in my second box running Windoze (because of the failed burn), but I think it's time to switch back. If I get any errors, can I post them here for feedback?

I might script this all toghether if people are interested.

Please do...

[User75]

Thank you very much dude ! Is there any compiled linux i386 version for mtkflash outthere That will save me for booting up to dos everytime I need to flash an x360 drive Thanx again

[abuali]

Got it working. Thanks!

However, burning some give me this:

Code:

$ growisofs -use-the-force-luke=dao -use-the-force-luke=break:1913760  -dvd-compat -speed=2 -Z /dev/hdb=foo.iso
Executing 'builtin_dd if=foo.iso of=/dev/hdb obs=32k seek=0'
/dev/hdb: splitting layers at 1913760 blocks
:-( unable to SEND DVD+R DOUBLE LAYER RECORDING INFORMATION: Input/output error
$

Not much info there, but has anyone experienced the same? Must be something with the iso, since most work...?

[abuali]

Stupid, stupid me. Of course it works.

I made a kernel update some time back and had forgotten to reboot (my uptime was something like 4 or 5 weeks). That's what caused it.

[Pandor]

Thank you very much dude ! Is there any compiled linux i386 version for mtkflash outthere That will save me for booting up to dos everytime I need to flash an x360 drive Thanx again

MTKFLASH is closed source, and quite frankly, it's a p.o.s.
Somebody should sniff the IDE bus, collect the commands and make a proper (open-source?) flasher... if only i had the knowledge/tools to do it

[uberfry]

Thank you very much dude ! Is there any compiled linux i386 version for mtkflash outthere That will save me for booting up to dos everytime I need to flash an x360 drive Thanx again

MTKFLASH is closed source, and quite frankly, it's a p.o.s.
Somebody should sniff the IDE bus, collect the commands and make a proper (open-source?) flasher... if only i had the knowledge/tools to do it

you could also just disassemble the executable...

[probutus]

maybe there is an easier way

I found a software usb sniffer for windows

http://benoit.papillault.free.fr/usbsnoop/

maybe we can have a look at the communication there (NO hardware needed)

[Pandor]

Sorry probutus, but I fail to see how a usb sniffer could help in this case...

[probutus]

If we connect the drive via sata->usb converter and use the tool to patch the SecSector we should see what data is transferred...

[xboxtilburg.tk]

MTKflash does not work with USB-SATA devices...

[probutus]

sh... I didnt know that (I have a hitachi and flashed it already via usb...)

[okkeg]

Hello Pandor!

This is quite good to hear, that it's also possible to backup games within linux since I'm passionated linux-user, but I've a question about that:

Code:

build xtreme iso:
cat video360.iso pfi.bin dmi.bin ss.bin game.iso > xbox360.iso

First of all, how do you make the files video360.iso and game.iso? And the ">" Operator means, that the previously created file xbox360.iso will be overriden so the command

Code:

sg_dd blk_sgio=1 bs=2048 count=3697696 iflag=coe verbose=2 if=/dev/hda of=xbox360.iso

would be obsolete or how should I understand it?

I would be very happy if I could make my backups on linux :-)