two card monte with the kernel? A new possibility for hacking the HDVR2 w/o prom mods

[pitt30]

well, i tried it one more time. chmod the hell out of every file, followed the instuctions to a "T" and guess what guys!!!!!!!




nothin. nadda zero.

fyi, i restored my backup 3.1 image and installed into tivo to be sure the image didn't get corrupt. worked fine.
restored u5 image and re-extracted the kern and FS.
remade the romfs file and CurrentKern file.

im at a loss for words. i have never been beaten by a computer yet, but it looks like at the end of this game, its TIVO 1 , ME 0.

as easy as cobelli's guide was to follow, i like the thought of not having to worry about the var partition being wiped and losing my hacks (already happened once which is why im trying monte).

[ronnythunder]

pitt30, man, i'm determined to get this thing going. if you're still willing, we'll get you through it.

ok, basics. you've confirmed the 3.1.0 image works. that's good. now, how about this: try restoring the u5 image (with no mods whatsoever) and make sure it works. when you're sure it works, move to the next step...

then, if you haven't already, concat d7o's two post faq together and print it out. have it beside you as you're doing the hacks. make a note on your printout for any differences in the instructions that you don't do. for example, my drive was hooked up such that it was hdb instead of hdc, so i had to remember to put hdb anywhere d7o put hdc. at *every* step, make a not of even the slightest keystroke that you do differently.

in other words, what I'm saying, is that i have followed every single keystroke of the procedure (except for the hdc versus hdb), and it worked for me. so, there are only a few possibilities:

(1) bad hardware (unlikely if not impossible in this case)
(2) bad u5 or 3.1.0 backup
(3) bad tool(s)
(4) bad scratch disk (usually your windows disk, used for temp storage)
(5) not following directions to the absolute, tee-total letter

i'll be up for at least another 4 or 5 hours, and i'll keep checking in if you want to give it a go.

ronny

[pitt30]

ronny,
i appreciate your persistance, unfortunatly i had enough for tonight. i will try again maybe tomorrow.
believe it or not d70's guide is set up exactly like my set up. meaning where his root is, my root is etc.
i will try however restoring the U5 to the drive and trying it out, i was a little rushed today due to football. (tivo wins, steelers lose, pitt30 pulls trigger on shotgun)
actually i was wondering if you have the u5kern and u5fs images you can share with me?

anyway, thanks again, and go enjoy your tivo. even unhacked, it still rocks.

[ClanMod]

Now that I've monte'd, do I have to replace my rc.sysinit or any other hacks at every reboot like I had to with hackinit. Also, should I load all my hacks from rc.sysinit or rc.sysinit.author? What is the difference?

TIA:D

[ronnythunder]

clanmod - no, you don't have to do anything at every boot. monte leaves your rc.sysinit (heck, *any* mods in the 3.1.0 root fs will stay) untouched.

as for rc.sysinit vs rc.sysinit.author, everyone has their favorite way, here's mine: the only change in rc.sysinit is to do bash (the call to rc.sysinit.author is already there). that way, you'll never end up hammering the startup process bad enough that you have to yank the drive again. i do all of my network startup and servers in the rc.sysinit.author.

ronny

[ClanMod]

One last question...if you've monte'd successfully, is this a sure fire way to prevent your hacks from being deleted?

[dogbreath]

My usb adapter quit working. Is there a new version of the OS for the 3.1 users? Anyone else had this problem?

dogbreath

[TomW]

I just checked mine and it is down as well???
The satellite feed was unplugged last night and I'm running U5.
A reboot brought it back online.

[ClanMod]

I finished monte'ing my HDVR2 yesterday and finally hooked it up today. At first, it booted normally. It worked about 5 times, then it brang up "A Severe Error". It says leave it plugged in for 3 hours to repair itself. So I did. It repaired itself and is now working fine. What happened? Will it happen again?

TIA

[pitt30]

OK, heres a question. using this for my runmonte, do i need the indentions, are they meaningfull, or can i just start each line from the left. if i do need them, is it just 1 tab over????


#!/bin/bash

# Derive monte settings from the bootparams stored on disk
set -a
newboot=/dev/hda$(/sbin/bootpage -a /dev/hda)
case "$root" in
*4) newroot=/dev/hda7 ;;
*7) newroot=/dev/hda4 ;;
*) handcraft=true
esac

# Unless "handcraft" is set, immediately load the next kernel via monte.
#
# Extra bootparams for that kernel can be set via a comma-separated list
# stuffed into the original "prom" bootparam (e.g. prom=FOO=bar,BAZ=bat)
#
if [ ! "$handcraft" ]; then
/sbin/insmod -f /mnt/kmonte.o
/mnt/monte $newboot root=$newroot dsscon=true console=2,115200 \
upgradesoftware=false ${prom//,/ }
fi

# If this point is reached, either monte failed or was never run.
# Just continually spawn off bash shells on the serial port for debugging.
export -n BASH_ENV
BASH_ENV=
while true; do /bin/bash -login</dev/ttyS2&>/dev/ttyS2; sleep 5; done



interesting, after i pasted this there are no indentions?? question still stands though. thanks in advance

[ronnythunder]

the indentions don't matter. i actually downloaded musclenerd's file from his post and copied it to the proper location - my editor never touched it.

ronny

[pitt30]

i got it to work!!!!!

i followed the directions exactly like before. the only thing i did different which may help other people is that i restored the u5 image again to the drive (seriously started over). once restored, i made sure it worked by putting it in the tivo. the u5 image came up and now verified as being good. continued with the instructions, as usual. this time i got the warning at step 25-26. at this point im feeling good cause i never saw that before.
continue on, put drive into tivo and got bash.

so to make a long story short, i think putting the u5 image in the drive and letting it boot in the tivo made the difference.

one thing i did notce is there is no rc.sysinit.author file. do i make this or should it be there. if i make it, where do i put the call to it. meaning in rc.sysinit (at the end after bash).

i hate to ask you to hold my hand here but spent too long trying to get this up.

[ronnythunder]

yay! congrats, pitt30! the rc.sysinit.author file is not there by default; you get to create it. it goes in /etc/rc.d and the rc.sysinit already has a call to it, so you don't have to add it. i would, though, put bash in the rc.sysinit (but that's me).

ronny

[pitt30]

i made a /hacks folder in hda7 under there i have various folder for eth, ftp etc. so now all i have to do is create the file and call the hacks.
can i call kmem like in the hackinit version through rc.sysinit.author file?

whats the easiest way to create the file without pulling the drive?

oh btw i manually insmoded the eth drivers and it came up, but how do i set the ip and all?

thanks for everyones help!!!!! i owe ya all a beer!!!


nevermind im a total ***** and im not thinking straight. i got it

[ronnythunder]

you can follow cobelli's u5/3.1.0 faq for the networking stuff; it's very well written.

ronny

[pitt30]

i am following the guide but for some reason it wont load my drivers from the author file.

insmod /hacks/usbcore.o
etc
etc
etc

do the hacks have to be under the var folder??

[ronnythunder]

when you say "won't load"... what does it do/say? you don't have to have the modules in any certain dir, but for some reason, i usually change to a dir other than where they're at to load them. i think i got a device busy error once when i was in the same dir trying to either insmod or rmmod...

ronny

[DeCSS]

Should I make my rc.sysinit.author look like my old hackinit? Here's how my rc.sysinit.author looks now and it does work, but I was just wondering if it should look different; also my hacks are loaded directly into the root (hda7), not into any /hack folder or anything:

------------------------------------------------------------------------------------

#!/bin/bash

echo "Loading Environmental Variables..." > /dev/ttyS2
PATH=/bin:/sbin:/bin:/tivobin:/tvbin:.
TIVO_ROOT=
MFS_DEVICE=/dev/hda10
IGNOREEOF=1000
export PATH TIVO_ROOT MFS_DEVICE IGNOREEOF

echo "Clearing Logs..." > /dev/ttyS2
rm /var/log/*

echo "Synchronizing Time..." > /dev/ttyS2
sleep 60

echo "Making Daily Call..." > /dev/ttyS2
/callfixsub.tcl &

------------------------------------------------------------------------------------

I already load BASH from rc.sysinit.

TIA

[geowar]

I try to keep my non-/hacks files & directories as pure (un-hacked) as possible. For this reason I don't copy rc.sysinit.author to /etc/rc.d, I create a symbolic link to it instead:

ln -s /mnt/hacks/rc.sysinit.author /etc/rc.d/rc.sysinit.author

This keeps all my hacked files in my hacks directory (where I can edit them) plus prevents them from getting deleted by an unexpected "clean-up" (the links will get deleted instead).

[pitt30]

this is my rc.init.author

#Load ethernet drivers
insmod /hacks/eth/usbcore.o
insmod /hacks/eth/hcd.o
insmod /hacks/eth/ehci_hcd.o
insmod /hacks/eth/ax8817x.o
sleep 10

ifconfig eth0 192.168.40.253 broadcast 192.168.40.255 netmask 255.255.255.0

#Telnet and FTP
tnlited 23 /bin/bash -login &
/hacks/tivoftpd

#!/bin/bash is actually at the end of rc.sysinit. also, the author file is in my /hacks directory and i used a symbolic link to it using:
ln -s /hacks/rc.sysinit.author /etc/rc.d/rc.sysinit.author

my questions are, is my file set up correctly and do in need the author file both in my hack direcort and also in /etc/rc.d to create the link.

im asking because no matter what i do, it never loads my drivers. and i dont see any error messages upon boot

[devnull]

Originally posted by pitt30
this is my rc.init.author

#Load ethernet drivers
insmod /hacks/eth/usbcore.o
insmod /hacks/eth/hcd.o
insmod /hacks/eth/ehci_hcd.o
insmod /hacks/eth/ax8817x.o
sleep 10

ifconfig eth0 192.168.40.253 broadcast 192.168.40.255 netmask 255.255.255.0

#Telnet and FTP
tnlited 23 /bin/bash -login &
/hacks/tivoftpd

#!/bin/bash is actually at the end of rc.sysinit. also, the author file is in my /hacks directory and i used a symbolic link to it using:
ln -s /hacks/rc.sysinit.author /etc/rc.d/rc.sysinit.author

my questions are, is my file set up correctly and do in need the author file both in my hack direcort and also in /etc/rc.d to create the link.

im asking because no matter what i do, it never loads my drivers. and i dont see any error messages upon boot

Check to see that you have execute permission set on the link and read permission on the file in /hacks.

For debug purposes, you might also edit rc.sysinit so that it tells you if it cannot execute rc.sysinit.author.

Last, you should also put print statements inside rc.sysinit.author that tell you when you enter and exit.

gary

[geowar]

> my questions are, [...] do in need the author file both in my hack direcort and also in /etc/rc.d to create the link.

The link command will create the symbolic link (file) in the /etc/rc.d directory so no you don't need the author there before you do the "ln -s …" command.

Make sure that both files have the execute protection set:

> chmod 755 /etc/rc.d/rc.sysinit.author

[pitt30]

ok i made the link and is see both files, i chmod 777 both of them. i rebooted and nothing, no usb. Man i hate this, i hate asking for so much hand holding but im missing something here.
actually i think im missing my /hacks directory in my path statement. am i on the right track here? if so how would i go about adding my path?

also, would anybody like to post thier author file, maybe this will help some of us and not have to post so many ?'s. i did search for one here, tivocom, and google. no sample is coming up

TIA

[ronnythunder]

here's mine. note that i only have one mod in my rc.sysinit: the addition of bash on the serial port. this is /etc/rc.d/rc.sysinit.author, and it's called by default from rc.sysinit:#!/bin/bash
if [ -x /hack/kmem ]; then
/hack/kmem 800b23b4 00001021
fi
insmod /hack/usbcore.o
insmod /hack/hcd.o
insmod /hack/ehci-hcd.o
insmod /hack/ax8817x.o
sleep 2
ifconfig eth0 192.168.1.63 broadcast 192.168.1.255 netmask 255.255.255.0
route add default gw 192.168.1.6 netmask 0.0.0.0 metric 1
sleep 2
/sbin/tnlited 23 /bin/bash -login &
/bin/tivoftpd
/hack/tivoweb-tcl/tivoweb
cd /hack
./tserver &
no links, no nuttin'. just the one file in the /etc/rc.d dir. it's executable (chmod +x), and, of course, the modules are in the /hack dir.

are you sure your ethernet adapter uses the ax8817x.o driver? maybe that's the problem?

ronny

[pitt30]

if i insmod manually the drivers load and i can telnet and ftp fine

that kmem statement, im assuming you didnt patch the kernel and are just calling kmem like in hackinit. am i right?

[devnull]

Here's what my stuff looks like:

First the /etc/rc.d directory, showing permissions and links:

btivo# cd /etc/rc.d
btivo# ls
finishInstall.tcl rc.arch rc.sysinit rc.sysinit.author
btivo# ls -l
total 33
-rwxr-xr-x 1 0 0 1776 Aug 14 2002 finishInstall.tcl
-rwxr-xr-x 1 0 0 7719 Aug 14 2002 rc.arch
-rwxr-xr-x 1 0 0 22030 Aug 6 05:58 rc.sysinit
lrwxrwxrwx 1 0 0 31 Aug 4 05:28 rc.sysinit.author -> /var/hack
/etc/rc.sysinit.author


This is how rc.sysinit.author gets invoked from within rc.sysinit. Note that this came from Tivo this way. I didn't add this line.


btivo# grep author rc.sysinit
if [ -f /etc/rc.d/rc.sysinit.author ] ; then
/etc/rc.d/rc.sysinit.author


Finally, here's the actual .author file that lives at the other end of the symbolic link. Note the print statements that show up in the /var/log/*.


btivo# cat /var/hack/etc/rc.sysinit.author
#!/bin/bash
#

echo "Starting /var/hack/rc.sysinit.author"
date
date>>/var/log/hackinit.log
PATH=/sbin:/bin:/tivobin:/tvbin:/var/hack/bin:.

TIVO_ROOT=
MFS_DEVICE=/dev/hda10
IGNOREEOF=1000
export PATH TIVO_ROOT MFS_DEVICE IGNOREEOF

/bin/bash</dev/ttyS2&>/dev/ttyS2&

insmod /lib/modules/usbcore.o
insmod /lib/modules/usb-ohci.o
insmod /lib/modules/pegasus.o

ifconfig eth0 192.168.123.251 broadcast 192.168.123.255 netmask 255.255.255.0

tnlited 23 /bin/bash -login &
tivoftpd

/var/hack/tivoweb-tcl/tivoweb

#TyTools
#kmem 800b23b4 00001021

mount -o remount,ro /

echo "/var/hack/rc.sysinit.author complete"

btivo#


gary

[ronnythunder]

Originally posted by pitt30
if i insmod manually the drivers load and i can telnet and ftp fine

that kmem statement, im assuming you didnt patch the kernel and are just calling kmem like in hackinit. am i right? did you try some sleeps around the insmods? also, there should be some messages in the logs when you insmod, especially if it fails, so it'd probably be useful to run it from the startup script, let it fail and see what's logged.

your kmem observation is correct. i'll get around to patching the kernel someday.

ronny

[pitt30]

date
date>>/var/log/hackinit.log
PATH=/sbin:/bin:/tivobin:/tvbin:/var/hack/bin:.

TIVO_ROOT=
MFS_DEVICE=/dev/hda10
IGNOREEOF=1000
export PATH TIVO_ROOT MFS_DEVICE IGNOREEOF


should my path be this:
PATH=/sbin:/bin:/tivobin:/tvbin:/hack:. (since my hack directory is in root and i have no bin directory under it)

also, that mfs device looks interesting can you explain?, and why 2 date statements. sorry if these are dumb questions but i like to understand what im typing so i dont feel like im not learning anything and just following a procedure.

heres my file attributes
bash-2.02# ls -l
total 34
-rwxr-xr-x 1 0 0 1776 Aug 14 2002 finishInstall.tcl
-rwxr-xr-x 1 0 0 7719 Aug 14 2002 rc.arch
-rwxrwxrwx 1 0 0 21947 Sep 15 19:12 rc.sysinit
-rwxr-xr-x 1 0 0 367 Sep 17 02:01 rc.sysinit.author

my rc.sysinit doesnt look like yours, ill check this out. chmod 777 or +x doesnt change it to what yours looks like
thanks

[pitt30]

and this is what i get at startup

No res
Scanning for phase3 repair scriptsed 10 times.
Starting Services.
Micro
Scanning for phase4 repair scripts
ApgManager Transition
/etc/rc.d/rc.sysinit: /etc/rc.d/rc.sysinit.author: No such file or directory
Adding networks of interest...ApgBootState:::DoBackgroundWork
rc.sysinit is complete
bash: no job control in this shellccess Card: Inserted Tuner[0
bash-2.02# MCP startup complete
Calling Run...
APGOpen calledt Locked
APGInitRead called00 Access Card:
APGOpen calledTuner[0]: Lock


and this is my sysinit statement to call author. (never touched the sysinit file)
[ ! -f /etc/rc.d/rc.sysinit.author ] || /etc/rc.d/rc.sysinit.author

[devnull]

Originally posted by pitt30
and this is what i get at startup
/etc/rc.d/rc.sysinit: /etc/rc.d/rc.sysinit.author: No such file or directory


There's your problem right there. This file doesn't exist. What do you get when you do:

ls -l /etc/rc.d/rc.sysinit.author

You might have successfully created a link, but linked it to a file that doesn't exist. The 'ln' command just does what you tell it without verifying that what you're trying to link to exists.


and this is my sysinit statement to call author. (never touched the sysinit file)
[ ! -f /etc/rc.d/rc.sysinit.author ] || /etc/rc.d/rc.sysinit.author [/B]

This line does the following:

if /etc/rc.d/rc.sysinit.author is a file, execute it.

The error message you got makes me think that it passed the "file" test but failed on execution. Check your link. I'll bet it doesn't point to a real file.

gary

[pitt30]

actually i got rid of the symbolic link and just put the author file in rc.d. i didnt want to use symbolic link as its just one more step to troubleshoot. once i get this working then ill test a sym link

[devnull]

Just saw the output from the previous page and it looks good. The only thing that's not clear is where you are when you do 'ls -l'

Can you do the 'ls -l' command on the full path to the file?

gary

btw - the two 'date' commands are because one goes to the messages (kernel?) file and one goes to the hackinit.log.

[pitt30]

bash-2.02# cd /
bash-2.02# ls -l /etc/rc.d/
total 34
-rwxr-xr-x 1 0 0 1776 Aug 14 2002 finishInstall.tcl
-rwxr-xr-x 1 0 0 7719 Aug 14 2002 rc.arch
-rwxr-xr-x 1 0 0 21947 Sep 15 19:12 rc.sysinit
-rwxr-xr-x 1 0 0 614 Sep 17 21:23 rc.sysinit.author

[devnull]

Originally posted by pitt30
bash-2.02# cd /
bash-2.02# ls -l /etc/rc.d/
total 34
-rwxr-xr-x 1 0 0 1776 Aug 14 2002 finishInstall.tcl
-rwxr-xr-x 1 0 0 7719 Aug 14 2002 rc.arch
-rwxr-xr-x 1 0 0 21947 Sep 15 19:12 rc.sysinit
-rwxr-xr-x 1 0 0 614 Sep 17 21:23 rc.sysinit.author

Looks good (unfortunately).

Can you send me the output of the following:

/etc/rc.d/rc.sysinit.author
cat /etc/rc.d/rc.sysinit.author

Maybe the #/bin/bash line at the top isn't right. This could also cause a problem.

gary

[pitt30]

bash-2.02# cat rc.sysinit.author

#!/bin/bash

echo "starting /etc/rc.d/rc.sysinit.author"

#Environmental variables
date
date>>/hacks/hackinit.log
PATH=/sbin:/bin:/tivobin:/tvbin:/hacks:.
TIVO_ROOT=
MFS_DEVICE=/dev/hda10
IGNOREEOF=1000
export PATH TIVO_ROOT MFS_DEVICE IGNOREEOF


#Load ethernet drivers
insmod /hacks/eth/usbcore.o
insmod /hacks/eth/hcd.o
insmod /hacks/eth/ehci_hcd.o
insmod /hacks/eth/ax8817x.o
sleep 10
ifconfig eth0 192.168.40.253 broadcast 192.168.40.255 netmask 255.255.255.0

#Telnet and FTP
/sbin/tnlited 23 /bin/bash -login &
/hacks/tivoftpd

#Tivoweb
/hacks/tivoweb-tcl/tivoweb
cd /hacks

[devnull]

Can you try to execute it by issuing:

/etc/rc.d/rc.sysinit.author

And then try to call that is made by rc.sysinit:

[ ! -f /etc/rc.d/rc.sysinit.author ] || /etc/rc.d/rc.sysinit.author

gary

[pitt30]

bash-2.02#
bash-2.02#
bash-2.02# /etc/rc.d/rc.sysinit.author
bash: /etc/rc.d/rc.sysinit.author: No such file or directory
bash-2.02#
bash-2.02#
bash-2.02#
</rc.sysinit.author ] || /etc/rc.d/rc.sysinit.author
bash: /etc/rc.d/rc.sysinit.author: No such file or directory
bash-2.02#

[devnull]

Originally posted by pitt30
bash-2.02#
bash-2.02#
bash-2.02# /etc/rc.d/rc.sysinit.author
bash: /etc/rc.d/rc.sysinit.author: No such file or directory
bash-2.02#
bash-2.02#
bash-2.02#
</rc.sysinit.author ] || /etc/rc.d/rc.sysinit.author
bash: /etc/rc.d/rc.sysinit.author: No such file or directory
bash-2.02#

So it looks like you can easily reproduce the failure by running rc.sysinit.author at the command line.

My first guess is that you have something wrong with the first line that tells the script what shell to run under. If it were one of the other lines, the display would tell you that.


There's nothing obvious that I can see. It looks exactly as it should. I HAVE seen something that looks like this once and it was because of the file was edited on a windows machine and then transfered over (^M and all).

Try to delete the entire first line and retype it IN LINUX. If that doesn't work, I'll leave it to you to figure out which line is bad by removing them and rerunning the script until it passes.

gary

[pitt30]

ok retyped the whole thing using edit pad chmod apropriatly and restarted. same thing.
i took out the first line and it found the file but other hacks failed.

im totally lost. maybe i should post in the forum

[ronnythunder]

you transferred the file in ascii mode at some point. it's got dos style carriage returns at the ends of the lines. this can be very tricky to diagnose. for example, consider this:[badn@carby badn]$ ls -lt try[1,2]
-rwxrwxr-x 1 badn badn 39 Sep 17 19:33 try1
-rwxrwxr-x 1 badn badn 35 Sep 17 19:32 try2
[badn@carby badn]$ cat try1
#!/bin/bash
echo hello
echo there

[badn@carby badn]$ cat try2
#!/bin/bash
echo hello
echo there

[badn@carby badn]$ ./try1
: No such file or directory
[badn@carby badn]$ ./try2
hello
theretwo files, two different results, yet they appear exactly the same when catted. the difference is that try1 has dos carriage returns at the end of the lines.

how are you creating the file?

ronny

[NutKase]

Originally posted by pitt30
ok retyped the whole thing using edit pad chmod apropriately and restarted. same thing.
i took out the first line and it found the file but other hacks failed.

im totally lost. maybe i should post in the forum

I'm not familiar with EditPad these days and TextPad is powerful too but to be sure I typed just the basic .author straight into .author using TELNET with this linux command that makes the file line by line as you go.

----------------------------------------
cat >> rc.sysinit.author
#!/bin/bash

#Put anything you want next
/hack/tivoftpd
----------------------------------------
Press <CTRL> D to quit and write it out then chmod it. Be careful and type slowly :) if you screw it up as I did several times it will still make the file. Even if you back out in bash you can delete it by rm filename.


Question: Is your 'stuff' directory /hack or /hacks? Also, the entire reason to use the rc.sysinit is that it's called automatically, moving it to another directory and calling it from within rc.sysinit seems to induce many more possibilities for error (chmodding, linking, etc.)

NutKase

PS. I suggest you # (comment) out the stuff you don't use yet to keep the troubleshooting down then uncomment each new hack as required. My order so far was bash, telnet, FTP using tivoftpd, tivoweb, mfs_ftp(not 100% yet.) Hope this helps.

[pitt30]

nutkase,
that was it. i used cat to create the file and the hacks worked!

i would like to thank everyone who helped. now i can continue work on my pearl script which will invoke mplayer get the list and provide a front end for easy streaming of shows and live tv. this will be for windows people. i know theres a java gui, but its flakey and doesn't always work. so hopefully i will be able to give something back to all of you


again, i really apreciate all of your hard work and its people like you all which makes this the best board!!!!

[mkenney5]

here are all the elements:

1. Series 2, 80GB, Sofware Version 4.01
2. Made backup
3. Added second drive of 120GB, no problems
4. Removed all drives from TIVO
5. On a spare 80GB drive I restore the hdvr2-31U5.mfs image
6. Followed instructions on superhero.org/tivo/
7. Installed newly monte'd drive in TIVO
8. Boots up and continuously reboots after seeing just wait a few more seconds

**no typos, cable or jumper problems and drive is in perfect condition**

any ideas of what the problem might be? if more info is needed please let me know

[NutKase]

Originally posted by mkenney5
here are all the elements:

5. On a spare 80GB drive I restore the hdvr2-31U5.mfs image
6. Followed instructions on superhero.org/tivo/

Print his instructions. Don't use the u5 image - it won't work for yours. Print my 2 messages here (which is the beginning of a SA S2 guide.)

NutKase's - How-To Learn and install a 3.0 to 4.01 SA S2 Monte! (http://www.dealdatabase.com/forum/showthread.php?s=&threadid=25743&perpage=40&pagenumber=6)


8. Boots up and continuously reboots after seeing just wait a few more seconds

This is the same symptom I got with U5

Lastly, as you don't have an HDVR2 you need to continue this post in the main thread

Standalone Series 2 People Out of Luck? (http://www.dealdatabase.com/forum/showthread.php?s=&threadid=25743)

You'll get more accurate help there.

NutKase

[mkenney5]

Yes this is a stand alone unit. I made a backup image of the stand alone unit and then upgraded to two drives. When working with the monte, I've been using only the stand alone image and hooking up only one drive.

This was just to see if I can get it to work. This is all testing on spare drives. The actual dual drive tivo is fine and working. If I'm able to get it working on a test drive, I'd like to work on the actual tivo drive. Does this have to have absolutely no recordings on it or can I monte a drive with shows recorded?

Thanks for any help.

[NutKase]

Originally posted by mkenney5
Yes this is a stand alone unit.

See above.

I made a backup image of the stand alone unit and then upgraded to two drives.

Make sure you make a backup of the two-drive setup.

Does this have to have absolutely no recordings on it or can I monte a drive with shows recorded?


You can monte a drive with shows on it. The only thing you're going to be doing is:

1. Placing a hackable 3.0 Kernel and FS onto 2 unused partitions on drive 1 of your 2-drive system.

2. Making and putting the ROMFS on the highest partition.

3. Killing the initrd of your 4 .01 (Please say you have already got the update :)) and placing it back where it was.

4. Lying to your tivo and telling it to boot 3.0 then start up 4.0 after checking is complete (Monte)

5. Flipping the bootpage to boot 3.0.

That said. In order to preserve recordings - backup! You might make an error after all.

You can do anything with your machine after you understand what's happening.

I have redone both of mine recently. I 'forced' my 4.01 partitions to be where I wanted them on 3 and 4, and my 3.0 on 6 and 7 by exporting them out with 'dd' then putting them back where I needed them then 'flipping or not flipping' the bootpage as necessary.

I did this just to ease the support of 2 machines. They're identical in partition placement now although they didn't come that way!

NutKase

[webdeck]

I'd like to give a big THANK YOU to everyone for the instructions. I spent a couple hours doing this last weekend (and adding 2x160GB), and it has worked like a charm. Thanks!

[mkenney5]

ok i got it to boot!!!

however, i forgot my serial adapters at work!!!

is there anyway i can get telnet working without going through the serial port?

thanks nutkase for all your help!!!

[ice5]

So, I'd tried Cobelli's guide earlier with no success.

I then tried a pre-hacked image with success, except I couldn't get one hack working (the hack which we cannot speak of).

So I then tried the monte. And the hack works! YES!

Anyway... with the cobelli hack, everything went in /var/hack/hackinit

With the monte, I've added the bash command, usb drivers, and ip config to /etc/rc.d/rc.sysinit

Should they be going elsewhere? I'm not sure.

Also, I cannot get telnet working. I've tried adding "tnlited 23 /bin/bash -login &" to the rc.sysinit and cannot log in, though I can ping the TiVo

Any help?

Thanks!

[nitrous]

Do you have the URL for downloading a prehacked image? To what extent is this imaged hacked? Has Monte been installed?
Thanks,
N|trous

[ice5]

Originally posted by nitrous
Do you have the URL for downloading a prehacked image? To what extent is this imaged hacked? Has Monte been installed?
Thanks,
N|trous

1) It's in the image begging thread, I believe at the twinbrothers site.

2) I do not know. Bash, telnet, and FTP worked "out of the box"

3) The Pre-Hacked image has not been monte'd to my knowledge.

So, can anyone tell me how to get Telnet running on my freshly monte'd box?

[mkenney5]

Ok...where to begin? Series 2 SA 80GB, 4.0.1, monte'd

Working
1. Loaded good backup to 80GB
2. Updated software from 4.01 to 4.0.1 and made backup
3. Monte'd drive (works beautifully, thanks nutkase!!!)
4. Bash prompt on serial, woohoo!!!
5. Added second 120GB drive
6. Got ftpd, telnet, vim61 to work perfectly
7. Made a rc.sysinit.author and added lines to start telnet and ftpd on boot (no problems)

Problems
9. Working with tivoweb - so so. Telnet slows down and when I started it on boot, tivo froze?? version 1.9.4 beta
10. Working with mfs_ftp - so so. Able to start. Connect via WS_FTP, but when I disconnect and stop on telnet side, I can't start again??
11. Working with TyTools R7V10 - so so. Able to start tserver and download ty file. When I go to make keyfile it doesn't work. Been reading that I have to use kmem and patch tivoapp. Did that and it didn't work

Any ideas on the last three????

If more info is needed, please let me know.

[NutKase]

Originally posted by mkenney5
Ok...where to begin? Series 2 SA 80GB, 4.0.1, monte'd

Please read my answer in the Stand Alone Series 2 thread...

Standalone Series 2 People Out of luck? (http://www.dealdatabase.com/forum/showthread.php?s=&threadid=25743)

I tried to point you there earlier since as you get past the monte things become more S2 specific.

I'll answer this there.

NutKase

[roach]

Originally posted by d7o
28) We need to create a file that will run the monte
vi /mnt/c/img/runmonte

You may need to brush up on vi. If you don't like vi, have never used vi, don't know vi, etc, you can create the file as follows:
cat > /mnt/c/img/runmonte

and then type exactly the following:
#!/bin/bash

export -n BASH_ENV
BASH_ENV=

(
/sbin/insmod -f /mnt/kmonte.o
/mnt/monte /dev/hda6 "root=/dev/hda7 console=2,115200 dsscon=true"
) &


I'm guessing that if my original bootpage showed hda4 I should replace hda6 and hda7 in the above with hda3 and hda4 respectively. Is that correct?

[beck]

Hey all,
I am having a real tough time performing this monte hack on my new S2 HDVR2 from the factory w/3.1.0 etc.
I downloaded a 3U5 image from the twinbrothers site for the two kernels and followed a variety of instructions to the letter but I am concerned that there is something I am missing or a typo or error in what I may be following. I downloaded instructions from this thread and from alt.org, host freedom, and superhero.org and still nothing. I used two different runmontes – the longer and the stripped down one.

#!/bin/bash

export -n BASH_ENV
BASH_ENV=

(
/sbin/insmod -f /mnt/kmonte.o
/mnt/monte /dev/hda6 "root=/dev/hda7 console=2,115200 dsscon=true"
) &

or

#!/bin/bash

# Derive monte settings from the bootparams stored on disk
set -a
newboot=/dev/hda$(/sbin/bootpage -a /dev/hda)
case "$root" in
*4) newroot=/dev/hda7 ;;
*7) newroot=/dev/hda4 ;;
*) handcraft=true
esac

# Unless "handcraft" is set, immediately load the next kernel via monte.
#
# Extra bootparams for that kernel can be set via a comma-separated list
# stuffed into the original "prom" bootparam (e.g. prom=FOO=bar,BAZ=bat)
#
if [ ! "$handcraft" ]; then
/sbin/insmod -f /mnt/kmonte.o
/mnt/monte $newboot root=$newroot dsscon=true console=2,115200 \
upgradesoftware=false ${prom//,/ }
fi


# If this point is reached, either monte failed or was never run.
# Just continually spawn off bash shells on the serial port for debugging.
export -n BASH_ENV
BASH_ENV=
while true; do /bin/bash -login</dev/ttyS2&>/dev/ttyS2; sleep 5; done

with two different bootparams

/cdrom/bootpage -P "root=/dev/hda4 BASH_ENV=\`mount\$IFS-n\$IFS/dev/hda16\$IFS/mnt;echo\$IFS/mnt/runmonte\`" -C /dev/hdc

or

/cdrom/bootpage -P "root=/dev/hdaD console=2,9600 dsscon=true BASH_ENV=\`mount\$IFS-n\$IFS/dev/hda14\$IFS/mnt;echo\$IFS/mnt/runmonte\`" -C /dev/hdc


I am able to watch the tivo boot with my serial cable on window 2000 hyperterminal. But I see no evidence of a load into U5 and then into 3.1. It just seems to boot into the U5 image with n hacks. I can’t get a bash prompt or anything.

If anyone has any thoughts on what I am missing or another version of the instructions it would be much appreciated. I have been loading and reloading various versions for about a week now with no success. Let me know if you have any other questions.

Thanks for all of your hard work in figuring these things out.
:confused:

[NutKase]

Originally posted by beck
with two different bootparams

/cdrom/bootpage -P "root=/dev/hda4 BASH_ENV=\`mount\$IFS-n\$IFS/dev/hda16\$IFS/mnt;echo\$IFS/mnt/runmonte\`" -C /dev/hdc

or

/cdrom/bootpage -P "root=/dev/hdaD console=2,9600 dsscon=true BASH_ENV=\`mount\$IFS-n\$IFS/dev/hda14\$IFS/mnt;echo\$IFS/mnt/runmonte\`" -C /dev/hdc

There are inconsistencies here. What's the D for? Where is your romfs partition, /hda14 or /hda16? Also, tell us where you put u5 and where you put 3.1?

The 3.1 location should be where your hacks are and you have no hacks because, so far, you've mixed up the instructions to tell the tivo where to boot after u5.

If /hda4 is root in the bootpage above then this should be where your u5 root partition was placed and all hack files should then be in partition 7.

NutKase

[geowar]

Does anyone know if the MFS backup/restore tools can be used to backup & restore a monte'ed drive? Will the hacks partition and the boot block be properly restored? Can I do the expansion (-x) to restore to a larger drive at the same time?

Inquiring minds want to know. ;-)

[David Bought]

Originally posted by geowar
Does anyone know if the MFS backup/restore tools can be used to backup & restore a monte'ed drive? Will the hacks partition and the boot block be properly restored? Can I do the expansion (-x) to restore to a larger drive at the same time?

Inquiring minds want to know. ;-)

No. Mfstools typically screws up anything that is not in MFS, /var, or the active root and kernel partitions.

[beck]

.

[beck]

Originally posted by NutKase
There are inconsistencies here. What's the D for? Where is your romfs partition, /hda14 or /hda16? Also, tell us where you put u5 and where you put 3.1?

The 3.1 location should be where your hacks are and you have no hacks because, so far, you've mixed up the instructions to tell the tivo where to boot after u5.

If /hda4 is root in the bootpage above then this should be where your u5 root partition was placed and all hack files should then be in partition 7.

NutKase

Sorry I wasn't more descriptive –

The "D" in hdaD is actually 4 (I just copied this out of the instructions which used an ABCD method to define proper partitions.)

hda4 is my U5 partition and hda7 is my 3.1 with my hacks.

6 has my killinitrd-s2-v3.x of 3.1 and hda3 has what fits of my U5 kernel (3in 2out or something like that)

My Romfs partition is /hda14 - yes I typed in the correct info in the bootpage (like I said I just copied these from the various instructions I found.)

The boot info I got through my serial shows the boot page stuff coming up twice but I don’t see any of the info from my rc.sysinit.author like “Loading Serial Bash”

I guess I can’t tell if the Bash on the U5 kernel ever really works – since I tried to do just a simple Bash hack on the U5 img I downloaded and couldn’t get that going either.

I am new at this but very comfortable with computers and somewhat familiar with linux.

Any tips or any more questions would be greatly appreciated.

[NutKase]

Originally posted by beck
I am missing or a typo or error in what I may be following.

See my post on Updates to 4.0 monte, How-To (http://www.dealdatabase.com/forum/showthread.php?s=&postid=122444#post122444)

They're for the SA S2 but they address a few typo's.

I used two different runmontes – the longer and the stripped down one.

I used the longer one, provided by MuscleNerd, and it works great.

That said, we can't tell if there's a typo in what YOU typed (believe me I typed several typo's (say that ten times)) without seeing it so post exactly what you're doing that's not working. I tried, in the post I pointed you to, to be very clear on spaces and why commands are as they are.

Update us soon.

NutKase

[beck]

Thanks for the reply, I am reading the entire thread now, I know I used a set of directions from http://www.superhero.org/tivo/ from Needo but I will look for your post now but I am pretty sure I already referenced it. I will get back to you after I do a little more work – thanks again.

[beck]

I just did it! That was a good 5 day addiction with no sleep. Well actually I was only semi-succesful. I got the monte up and running, booting from u5 to 3.1 but am getting a "no bash control on this shell" in the boot stream from the serial. (Does anyone have any thoughts on this?) I read in the SAS2 thread NutKase directed me to (Thanks very much BTW) something about :

Check your rc.sysinit to ensure you edited Line 566 (or close to it) to the following.

[ -x /etc/netfilter-disable ] && /etc/netfilter-disable

Did you fix iptables?

If you can't get bash on serial log the boot process to the serial port. Search keyword: dsscon=true

but for the life of me I cannot find that line in my rc.sysinit. Is that maybe just for the SaS2 and not the HDVR2?

BTW I got my system working by incorporating some of NutKase's SAS2 instructions with the new HDVR2 instructions I previously downloaded. This line below is from the HDVR2 monte instructions at www.hostfreedom.com/tivo/ I believe that this is where I was hung up for several days:

/cdrom/bootpage -P "root=/dev/hdaD console=2,9600 dsscon=true BASH_ENV=\`mount\$IFS-n\$IFS/dev/hda14\$IFS/mnt;echo\$IFS/mnt/runmonte\`" -C /dev/hdc

You'll notice Nutkase's boot page is missing the "console=2,9600" above. I thought this line was necessary to watch the Tivo boot stream so I always left it in until last night (well actually early this morning). I decided to copy NutKases's intructions verbatim but obviously updating them to fit the HDVR2 and used his boot page:

/cdrom/bootpage -P "root=/dev/hda7 dsscon=true BASH_ENV=\`mount\$IFS-n\$IFS/dev/hda16\$IFS/mnt;echo\$IFS/mnt/runmonte\`" -C /dev/hdc

Then it started to work (showing monte on the stream) but at the same time this is the first time I used the premade runmonte image without opening it first to change the 2,115200 to 2,9600. So I am not sure if this had something to with it. As I said in an earlier posting - I am still learning (and hopeing to understand more over the next few tinkers).

Let me know if anyone has any thoughts about getting the BASH working. I am expecting my USB adapter to arrive in a few days and will be ready to go.

Will I be able to bypass serial BASH and go right to the USB?

Also, could anyone direct me to kmem instructions for a monte'd HDVR2?

Thanks Everyone (Especially NutKase)!
Beck

[mrblack51]

Originally posted by beck
ICheck your rc.sysinit to ensure you edited Line 566 (or close to it) to the following.

[ -x /etc/netfilter-disable ] && /etc/netfilter-disable

Did you fix iptables?

but for the life of me I cannot find that line in my rc.sysinit. Is that maybe just for the SaS2 and not the HDVR2?

the change for netfilters is only needed for units with software versions where HMO is an option. that means it only is necessary for SA s2 units running 4.x software versions or higher currently

[beck]

Originally posted by mrblack51
the change for netfilters is only needed for units with software versions where HMO is an option. that means it only is necessary for SA s2 units running 4.x software versions or higher currently

Gotcha
Thanks mr. Black. I'll stop scrolling up and down my rc.sysinit now.

[NutKase]

Originally posted by mrblack51
the change for netfilters is only needed for units with software versions where HMO is an option.

Thanks, mrblack51! I learned something today.

NutKase

[nobody]

I read d7o's instructions a few times and thought I knew what to do with my new HDVR2 factory installed with 3.1.0-01-2-151.

However, the partition information displayed on my 40GB is different.

hda1 Apple
hda2 kernel 1
hda3 Root 1
hda4 Bootstrap 2
hda5 Kernel 2
hda6 Root 2
hda7 Linux Swap
hda8 /var
hda9 MFS Application region
hda10 MFS Media region
hda11 Second MFS Application region
hda12 Second media region
hda13 Extra
hda14

bootpage -p /dev/hda returns root=/dev/hda7. How can that be when hda7 is supposed to be Linux Swap from mstools2 boot CD.

Anyway, I completed with d7o instruction as my setup in other aspect is identical to one he described.

Restoring the hard disc in the HDVR2, the system is now stuck on Powering up. On the serial port, it displays service number, then it asks for "What is the password". It would go no further at this point, I suppose that's why it is tuck on Powering Up on the TV.

As I don't have a U5 image, I downloaded a mfs image from one of the referred sites. I then mfsrestore it onto the new disk. Again the partition layout is identical to the above. I use dd to extract the hda6 and hda7.

Note I mount the Tivo disk as Pri Master and DOS disk as Sec Master.

Could someone provide some assistance.

[roach]

Originally posted by nobody
However, the partition information displayed on my 40GB is different.

for whatever reason the partition information printed at boot is shifted by one line so root is at partition 7 and swap is at 8 (try using pdisk after the boot). I've had the similar bootup message:

1. (cross your fingers this is it) your term program has a keyboard buffer and has sent a carriage return as the TiVo boots telling it to go into the service mode, quit and restart the term program (don't press enter!!!) or pull the serial connection for a bit.

2. you've done something wrong with dd'ing the kernel and fs images back to the drives, try again.

Those are the issues that I've run into at least.

[nobody]

Thanks for the tip. pdisk does report the partition info correctly.

I re-did the Monte and it's working fine now.

I waited for a while before plugging the serial cable and it's surprising that it took a while before anything appeared on the screen. It may be that I was pressing Return too many time and thus interrupted the boot causing it to prompt for an administrative password. If so, there was probably nothing wrong with the earlier Monte installation.

[RaysMD]

Thanks for all the informative posts. I just monte'd my system from scratch. Out of box -> monte'd system in 6 hours, but 3 days of reading.


I have the U5 filesystem (hda7) imaged and could upload it somewhere so that some of the initial backup steps can be skipped. Let me know where to put it. It's 16MB.


Second thing. Where are you guys storing your files? I have all of mine in the 3.1 filesystem (hda7). Why would I put a back up copy on the U5 (hda4) partition. That way if anything goes wrong I could go back to it.


Rays

[newbie]

I've gone through Monte twice using montestuff.zip file (da0 instructions). I'm not getting a bash prompt and I don't get past the powering up screen. I checked all my commands at least twice.

The only thing I can think of--my paritions are flipped from the instructions. I used the romfs.img from the zip file (step 30). From reading I thought the current runmonte file checked both partitions.

Is there something obvious I'm missing? I'm either going to try it a third time (in case of a typo) or just use the hacked image and try Monte later.

[David Bought]

Originally posted by newbie
Is there something obvious I'm missing? I'm either going to try it a third time (in case of a typo) or just use the hacked image and try Monte later.

Yes. You should be reading the boot messages and pasting them here if you don't understand them.

Search keywords: console dsscon

[ronnythunder]

Originally posted by RaysMD
I have the U5 filesystem (hda7) imaged and could upload it somewhere so that some of the initial backup steps can be skipped. Let me know where to put it. It's 16MB.

Second thing. Where are you guys storing your files? I have all of mine in the 3.1 filesystem (hda7). Why would I put a back up copy on the U5 (hda4) partition. That way if anything goes wrong I could go back to it. the montestuff.zip on twinbrothers already has this. it's got everything that's needed.

as for your second item, this won't work. that's the fs that gets checked by the u5 kernel, and anything that doesn't pass muster gets deleted. if you make it immutable so it can't be deleted, the box will go into a reboot loop.

ronny

[newbie]

Originally posted by David Bought
Yes. You should be reading the boot messages and pasting them here if you don't understand them.

Search keywords: console dsscon

OK--Set teraterm to 115200 got the boot messages

Partition 0 signature is false, trying alternate
diskboot found partition 0 signature is false.
Ram size = 64

I suspect the problem is a typo (MY ERROR). I used the files from montestuff so killinitrd should have been done right. Before trying for the third time does any one have any ideas as to which part I may have made my mistake.

I am using the pre-canned runmonte--romfs.img file from the archive. Could that be loading the wrong partition? My partitions are reversed but I thought the documentation said it could handle either one.

[newbie]

Originally posted by David Bought
Yes. You should be reading the boot messages and pasting them here if you don't understand them.

Search keywords: console dsscon

OK--Set teraterm to 115200 got the boot messages

Partition 0 signature is false, trying alternate
diskboot found partition 0 signature is false.
Ram size = 64

I suspect the problem is a typo (MY ERROR). I used the files from montestuff so killinitrd should have been done right. Before trying for the third time does any one have any ideas as to which part I may have made my mistake.

I am using the pre-canned runmonte--romfs.img file from the archive. Could that be loading the wrong partition? My partitions are reversed but I thought the documentation said it could handle either one.

[ronnythunder]

the "partition signature is false" messages suggest that something is badly wrong. are you sure you followed the procedure properly?

ronny

[newbie]

Originally posted by ronnythunder
the "partition signature is false" messages suggest that something is badly wrong. are you sure you followed the procedure properly?

ronny

I follwed the instructions from the montestuff.zip. I used the files from archive.
1) backup came from my original 40Gig drive (already upgraded to 311b) I backed up to my HD. Shut down and put my new Seagate 120 in the computer.
2) followed all the steps BUT used the files listed in the manifest.

Possibilities--
1) Could new HD be bad? Follow up I started for the third time I stopped after step 21 and connected Tivo to TV. Got as far as getting satellite signal (I didn't connect the coax). I assume my original backup and probably drive are OK

2) I used romfs.img as in in step 30 skipping steps 27-29 EVEN THOUGH my partitions were reversed. I read that the new runmonte could handle either pairing.

I was very careful to proof my work I don't think I made the same typo in 2 places.

[ronnythunder]

what about steps 32-34? it's easy to mess up the bootpage, and this could cause these problems.

ronny

[newbie]

I'm going to try it a third time. I was VERY careful with step 32 but will triple check next time.

Step 34 I used the kernel in the zip file (311bkern2.img). The manifest fille list says this file can be used in step 34 (already been killinitrd'd). If you think that may be the problem I can use the first 2 steps in 34 and kill it myself.

[NutKase]

I think it's a 'flip' problem. ie

bootpage -f -C /dev/hdX

...but I'll have to admit that I know nothing of the instructions in montestuff.zip. I'll have to read them.

NutKase

[newbie]

After my third attempt my Tivo will boot (initial problem was a typo and/or incorrect jumper on hd) BUT all the changes I made (files added to bin, rc.sysinit chages are gone as is the rc.sysinit.author file).

Yes, I followed step 34 but I used the 311bkern2.img which has already been killinitrd'd)

I ran the bootpage command. My root is hda7 and the command i typed in step 32 is showing. My real files are in hda4. I assume the Monte is working.

1) Do I have to go to the beginning or can I start with step 34? I am guessing the 311bkern2.img may not be killinitrd as stated. or is there another explanation?

I'm starting to use VI much easier than using cat.

[ronnythunder]

you can verify the killinitrd by running it again on the kernel. if it spits out some kind of error about expecting one value but finding another, it's been killinitrd'd.

i agree that this seems like the problem. do you have the serial cable hooked up? i think it shows in the boot messages with the dsscon whether the initrd is being run.

ronny

[newbie]

Thanks for your help. I re-traced my steps. I think the complete failure was caused by mis-setting the jumpers.

Ronny I went back and re-examined bootpage results. I had a typo in step 32 (checking it twice wasn't enough) so the monte wasn't getting executed.

Teraterm is behaving strange. (8N1N) terminal 80 X 25. Not very responsive, and I have to sometimes type a space or the first character of the command gets skipped so the command is not recognized. I'll have to start searching but in the meantime I can get my TV back.

[Sleeper]

I read some post earlier in this thread and on alt.org where embeem and hamsterdam were talking about installing U5 on partitions 2 & 5.

I thought that the S2 DTivos could only boot from the 3,4 or 6,7 partitions. Can this indeed be done?

[vertex]

this is goign to be my first attempt at hacking a hdvr2 but i have been using linux for years. quick question. im going to take my 3.10 HD out of the tivo and set it aside and not use it. im going to take a 3.1u5 image and put it on a new HD. i need to hack that via bash access. i then need to setup 3.1U5 to call monte and the new initrd kernel. also i need to patch the kernel to the partitions etc. but that should be the jist of it ?

[Sleeper]

Originally posted by vertex
quick question

If that's a quick question I'd hate to see a long one.

But yes, that's basically it. However, you're going to make a backup of your original drive and restore it to your new drive.

Then you're going to install the u5 kernel & file system on the alternate partitions, set the bootpage to boot u5 and u5 to monte the null initrd'd real kernel & file system which you are now free to hack.

That's the quick answer.

[captain_video]

I just performed the monte hack for the first time last night using the montestuff.zip files. I hosed it up the first time through because I couldn't remember the locations of all of the files after I booted into Linux :rolleyes: . I modified the tutorial supplied in the montestuff.zip file to make it ridiculously easy to hack your S2 DTivo using the files included in montestuff.zip. I got bash on the very next attempt when using this guide.

If you've already got 3.1.0 on your S2 DTivo you can jump right past the beginning steps and start hacking with step 22 of d7o's guide. If you got updated to 3.1.1b then you'll need to install a backup image of 3.1.0 per the guide before getting into the hacking section.

I take absolutely no credit for developing any of the info in this tutorial so save the kudos for those that really deserve it. It's a first draft but it's entirely functional in its current form. I would like to add steps to perform additional hacks but I'm brushing up on where to put the hacks on my DTivo at the moment. As soon as I get it sorted out I'll post an update. Any and all feedback or suggestions are welcome.

[newbie]

Good guide. My problem (last weekend) was typos. I just used the Cobelli guide for ethernet, kmem, FTP and (thanks to Mr Black) the rcy.sysinit.author file.

[vertex]

Well i thought i monted my system correctly but i dont think i did. i followed the steps in the monstestuff package. my bootpage is correct everything loads up. my question is, i used the 3.1.b kernel when i goto system settings its showing up as 3.1-151. i also added files to the /dev/hda7 3.1 FS and they are not gone so i am also assuming i did the monte correctly. but i cant get nocall to work along with another program that cant be discussed on here. any ideas on howto tell if my monte is 100% working?

[captain_video]

Do you have bash? It seems to me that if you've got bash then you've got a working monte.

[vertex]

thats another problem on its own, i try to see if i have bash but i get some crazy characters on my hyperterminal . i got it set to vt100 9600-8-n-1 or some stuff like that. so i cant confirm. im trying to setup all the hacks via mfstools bootcd and not via serial.

[MuscleNerd]

115200

[jjanicik]

Thanks Captain Video.

After printing out and following your guide I have finally monte 'd my HDVR2/3.

I sincerely appreciate your efforts.

[captain_video]

Thanks, but all I did was edit an existing guide to reflect the use of the montestuff.zip file. I'm still working on getting my other hacks installed so I can update the guide even further. I'm having trouble getting my monte'd HDVR2 to recognize my rc.sysinit.author file as it doesn't seem to be running at bootup. I got telnet working when I input the lines manually but they're not executing at bootup. I also keep getting errors when I try to run tivoftpd. I'll keep playing around with it until I get it right and then report my findings.

[jjanicik]

I thought I saw a thread that mentioned if you chmod the rc.sysinit.author with 755 instead of 777, that this might fix your problem.

[mrblack51]

Originally posted by jjanicik
I thought I saw a thread that mentioned if you chmod the rc.sysinit.author with 755 instead of 777, that this might fix your problem.

um, no

[gms199]

Hey guys,

I have been working on this thing for about 4 days now and i am having some problems.

I have followed the guides and believe i am getting monte to work. This is what happens. In hyper terminal while it starts up. all kinds of stuff scrolls through like i read it was supposed too. Then i see the monte stuff but then where it should so stuff about the next kernel all i get are strange symbols and no bash.

The system does start up fine however.

Any ideas would be appreciated. Thanks.

Greg

[NutKase]

Originally posted by gms199
where it should so stuff about the next kernel all i get are strange symbols and no bash.

Make sure your bootpage command sets the baud rate to

dsscon=true console=2,115200

NutKase

[captain_video]

Make sure you've got your terminal program set up to use a baud rate of 115200 via the serial port or you'll either just get gibberish or nothing at all.

[RUBiK]

EDIT: Figured out the problem and my original post was confusing as hell (hence no replies, perhaps).

I have replaced the entire text of my post since, again, it didn't make much sense and I wasn't even stating what the problem was, I don't think...

Sorry for the waste of a post... Please ignore. I won't post at that late hour on a weekday in the future. :p

[testfish]

I to can't get bash prompt. I have a rcadvr39 i completed monte I am able to boot up and watch tv but thats it. the following is the output from tera term when i am booting. Any thoughts?


bash-2.02# bash-2.02# bash: no job control in this shell
bash: bash-2.02#: command not found
bash-2.02#
bash-2.02# bash: bash:: command not found
bash: bash:: command not found
bash-2.02#
bash-2.02# bash: bashbash-2.02#
bash: bash:: command not found
APGOpen called
APGInitRead called
ttyS: 1 input overrun(s)
bash-2.02#
bash-2.02# bash-2.02# bash: no job control in this shell
bash: bash-2.02#: command not found
APGOpen called
APGInitRead called
DTuneOpen called
ApgLogFrequency( 0 sec )
Finished creating ApgDriverFrameIter
Using a smartSorter!
Creating ApgReader
CamOpen called
ttyS: 14 input overrun(s)
CamInitRead called
CamOpen called
CamInitRead called
oslink: 00 00 00 Access Card: Not Inserted Tuner[0]: Not Locked Tuner[1]:
Not Locked
ttyS: 16 input overrun(s)
d 159 (for 158) waking up
DTuneOpen called
DDLOpen called
DDLInitRead called
INIT to ACQUIRE_MARKER
oslink: 01 00 00 Access Card: Inserted Tuner[0]: Not Locked Tuner[1]:
Not Locked
ttyS: 25 input overrun(s)
Adding ttyS: 23 input overrun(s)
networks of inels
state PREP_SETUP to EXPRESSION_EVALUATION

[mrblack51]

I to can't get bash prompt. I have a rcadvr39 i completed monte I am able to boot up and watch tv but thats it. the following is the output from tera term when i am booting. Any thoughts?


bash-2.02# bash-2.02# bash: no job control in this shell
bash: bash-2.02#: command not found
bash-2.02#
bash-2.02# bash: bash:: command not found
bash: bash:: command not found
bash-2.02#
bash-2.02# bash: bashbash-2.02#
bash: bash:: command not found
APGOpen called
APGInitRead called
ttyS: 1 input overrun(s)
bash-2.02#
bash-2.02# bash-2.02# bash: no job control in this shell
bash: bash-2.02#: command not found
APGOpen called
APGInitRead called
DTuneOpen called
ApgLogFrequency( 0 sec )
Finished creating ApgDriverFrameIter
Using a smartSorter!
Creating ApgReader
CamOpen called
ttyS: 14 input overrun(s)
CamInitRead called
CamOpen called
CamInitRead called
oslink: 00 00 00 Access Card: Not Inserted Tuner[0]: Not Locked Tuner[1]:
Not Locked
ttyS: 16 input overrun(s)
d 159 (for 158) waking up
DTuneOpen called
DDLOpen called
DDLInitRead called
INIT to ACQUIRE_MARKER
oslink: 01 00 00 Access Card: Inserted Tuner[0]: Not Locked Tuner[1]:
Not Locked
ttyS: 25 input overrun(s)
Adding ttyS: 23 input overrun(s)
networks of inels
state PREP_SETUP to EXPRESSION_EVALUATION

sure looks like its working properly. have you tried hitting enter after all that stuff goes by?

[ronnythunder]

testfish, i'd guess either your telnet client is echoing input back to output, or your cable is not plugged into the tivo properly and/or has a problem (intermittant short etc.)

ronny

[testfish]

ok have tested cable with meter seems to be ok and am hitting the enter key still no luck. changed computers and this is the last few lines i receive now.

bash: bash-2.02#: command not found
bash-2.02#
bash-2.02# bash-2.02#
bash: bash-2.02#: command not found
APGOpen called
APGInitRead called
ttyS: 2 input overrun(s)
DTuneOpen called
ApgLogFrequency( 0 sec )
Finished creating ApgDriverFrameIter
Using a smartSorter!
Creating ApgReader
Helper Thread 190 (for 189) wakAtApgBottyS: 20 input overrun(s)
otState:::D

[NutKase]

I agree, it looks like it's working. If you have a home network...

Try logging in with telnet. Just telnet to your tivo ip with

telnet 192.168.x.x

and let us know what happens.

NutKase

[ronnythunder]

i'm *still* betting on either improperly configured terminal emulator or cable not plugged in all the way. i had this exact problem once, and i found out that my plug liked about an eight of an inch being plugged all the way in to the back of the tivo. it worked fine after that.

look at the output again: it's saying "bash-2.02#: command not found". that means that it is somehow receiving the string "bash-2.02#. since the tivo is sending that out, seems only logical that this is the source of it, too.

try using plain old hyperterm or something else, and turn *everything* related to echo off.

ronny

[ronnythunder]

i just thought of something else: the format of the bash command could be wrong. tell us *exactly* what you put, with spaces and all.

ronny

[testfish]

just so you know I have been working on this for a week. I have used several different bash programs but the latest bach is as follows:

cat >> /mnt/tivo/etc/rc.d/rc.sysinit
/bin/bash<dev/ttyS2&>/dev/ttyS2&

[NutKase]

just so you know I have been working on this for a week. I have used several different bash programs but the latest bach is as follows:

cat >> /mnt/tivo/etc/rc.d/rc.sysinit
/bin/bash<dev/ttyS2&>/dev/ttyS2&

You need a space before the </dev and after the /bash and you're missing a / like...

/bin/bash:)</dev/ttyS2&>/dev/ttyS2&

Smiley = space.

NutKase

[testfish]

ok so I added the space but I still am not able to type a command. jsut seems to hang but i am watching tv though.

[NutKase]

ok so I added the space but I still am not able to type a command. jsut seems to hang but i am watching tv though.

How about the /

NutKase

[beck]

EDIT: Figured out the problem and my original post was confusing as hell (hence no replies, perhaps).

I have replaced the entire text of my post since, again, it didn't make much sense and I wasn't even stating what the problem was, I don't think...

Sorry for the waste of a post... Please ignore. I won't post at that late hour on a weekday in the future. :p

Sorry for the delayed response - I have been out of town. If you didn't get it squared away let me know I will be happy to help if I can.

Beck

[RUBiK]

Sorry for the delayed response - I have been out of town. If you didn't get it squared away let me know I will be happy to help if I can.

Beck
Yup, it's all working fine now; in fact, I'm not all too sure what the original problem I had was. I fixed it the following morning, np.

Thanks.

[fginneba]

I will relegate myself to the noob board. Sheesh! No reason to be malicious about it.

No, I hadn't made it through all 25 pages posted across the last year. Thanks to those who actually bothered to answer my question.

[mrblack51]

I admit, I've been away from this for a while, so I haven't been keeping up.

I've had my bash_env working for some time. I know that Monte is the way to go, but what advantages do I get by going that route? I believe I'm still running the 151 image, so maybe the new loads have added some degree of advantage?

Thanks for the info!

this is a newbie question, so it should be posted in that forum. did you even read this thread? all the answers are here.

[geowar]

I've had my bash_env working for some time. I know that Monte is the way to go, but what advantages do I get by going that route?

The v4.0 kernel disabled the BASH_ENV hack. Monte gets around that by booting (w/BASH_ENV) on a v3.0.1 kernel and then switching to the v4.0 kernel.

Plus Spiders scripted install automates the whole process! ;-)

[NutKase]

The v4.0 kernel disabled the BASH_ENV hack. Monte gets around that by booting (w/BASH_ENV) on a v3.0.1 kernel and then switching to the v4.0 kernel.

Plus Spiders scripted install automates the whole process! ;-)

I hope that was tongue in cheek. :)

His question was answered properly by mrBlack51 3 hours earlier. So we wouldn't have to post AGAIN just for one person what can be learned from reading.

Your post was too specific... You can monte many more kernel versions than mentioned.

BTW, who's Spider?

NutKase

[MuscleNerd]

You can monte many more kernel versions than mentioned.
Actually, you can monte into more than kernels. For instance, you can monte directly into a different version of the boot rom prompt (with one or two byte changes). Why would you want to? Who knows. I did it just so I could try out the new v2.5 "color_bars" bootrom command on one of my HDVR2s without the programmable version of the prom installed.

[tivosohn]

I got my HDVR2 properly monte'd with 3.1.u5 and 3.1.0-01-2-151. Is there much benefit to move to either the 3.1.1c or 4.x? I have been reading some posts, but doesn't seems to have much success on the updating without blowing away all the recordings. Anyone know what are the additives since the 3.1.0-01-2-151?

Thx.

[MuscleNerd]

The two biggest benefits I get out of 4.0 are: (a) no more slow entry into Now Showing when you have many recordings...it's very quick; (b) being able to jump to the beginning of the live buffer

[tivosohn]

So, comparing 4.0 to 3.1.1c? which upgrade path do you recommand? Because based on what other has posted, it sounds like each one has their own fixes. Do you know which version people recommand to upgrade to?

[PlainBill]

So, comparing 4.0 to 3.1.1c? which upgrade path do you recommand? Because based on what other has posted, it sounds like each one has their own fixes. Do you know which version people recommand to upgrade to?

It depends on the individual. If you want to quickly upgrade your system, and following Sleeper's TiVoScripts will tax your abilities, I'd suggest 3.1.1b or c. Likewise if you have a RID unit - 4.0 won't work on them yet. If you want a learning experience, or you're already a TiVo expert, go with 4.0.

PlainBill

[tivosohn]

Ahh..cool. I don't have a RID unit. Sounds like 4.0 is the correct path to take for my situation. By the way, I have been doing the Monte manually. Is the Sleeper's tool doing anything different than that? If not, I think I am more comfortable with the manual step better than launching something blindly and cross my fingers hoping the road doesn't have any bumps. :o

[PlainBill]

Ahh..cool. I don't have a RID unit. Sounds like 4.0 is the correct path to take for my situation. By the way, I have been doing the Monte manually. Is the Sleeper's tool doing anything different than that? If not, I think I am more comfortable with the manual step better than launching something blindly and cross my fingers hoping the road doesn't have any bumps. :o

When you Monte manually, you have a number of choices you can make - where you put the romfs image, what hacks are included, where do you put the commands to start your hacks, etc, etc. It will take a while to do this the first time, but you WILL know where everything went.

If your Linux skills are weak, but you CAN follow directions, Sleeper's TiVoScripts will give you a working system in very little time. However, the process will be almost a complete mystery to you. (I didn't even know where rc.sysinit.author was, and didn't have the foggiest idea how to change the ip address).

PlainBill

[tivosohn]

I see. Thanks for your inputs. I think I better off stick with the manual process, otherwise I might end up spending more time trying to debugging or finding things later on. :cool:

[RKone]

By the way, I have been doing the Monte manually. Is the Sleeper's tool doing anything different than that?

If you already know how to do everything, then I'd say sleeper's cd is a good way to go. It doesn't make stupid mistakes and it's fast. You'll still want to update/install some programs and hacks, but the monte itself is perfect everytime.
Sleeper's takes about half an hour including swapping hard drives and testing, and it installs usb, telnet and ftp - you'll be all ready to go to quickly install the custom hacks you prefer.

[NutKase]

I see. Thanks for your inputs. I think I better off stick with the manual process, otherwise I might end up spending more time trying to debugging or finding things later on. :cool:

I disagree with RKone above... Why go backwards when you already know so much? :)

NutKase

[AlphaWolf]

Just for the record, its possible to completely fix the screen corruption on the RID tivos, hence eliminating all current issues with a monte setup. Reference here: http://www.pvrhax0r.com/forum/showthread.php?postid=96#post96

[rc3105]

new times call for a new thread

two card monte with the kernel evolves in 2004 (http://www.dealdatabase.com/forum/showthread.php?t=37226)